2026-01-06_abc_gantt_chart_recalibration_ab60baa7

Meeting Title: ABC Gantt Chart Recalibration Date: 2026-01-06 Meeting participants: Casie Aviles, Samuel Roberts, Mustafa Raja, Amber Lin, Uttam Kumaran

---

WEBVTT

1 00:00:44.860 ⇒ 00:00:45.830 Samuel Roberts: Ayy.

2 00:00:46.820 ⇒ 00:00:47.630 Casie Aviles: Hey, Sham.

3 00:00:49.430 ⇒ 00:00:50.640 Samuel Roberts: How’s it going today?

4 00:00:53.640 ⇒ 00:00:55.420 Casie Aviles: Yeah, doing good so far.

5 00:00:57.140 ⇒ 00:00:57.780 Samuel Roberts: Great.

6 00:00:57.920 ⇒ 00:00:58.929 Samuel Roberts: Excuse me.

7 00:01:01.470 ⇒ 00:01:05.170 Mustafa Raja: I looked into the semantic markdown stuff, and it’s not good.

8 00:01:06.090 ⇒ 00:01:07.360 Samuel Roberts: Oh, really? Okay.

9 00:01:07.360 ⇒ 00:01:15.359 Mustafa Raja: Yeah, so we were thinking of, we were thinking, like, it would do stuff with the metadata, but what it does is it combines…

10 00:01:15.510 ⇒ 00:01:19.640 Mustafa Raja: A lot of the headings and chunks them together, you know.

11 00:01:20.500 ⇒ 00:01:27.770 Mustafa Raja: Yeah, so not really much of control do we get with that. I think…

12 00:01:28.400 ⇒ 00:01:33.960 Mustafa Raja: Our own strategies much better than What this gives us.

13 00:01:34.500 ⇒ 00:01:36.080 Samuel Roberts: Okay, that’s good, good to know.

14 00:01:38.150 ⇒ 00:01:38.720 Mustafa Raja: Yeah.

15 00:01:41.840 ⇒ 00:01:44.859 Samuel Roberts: Alright, sorry, one step, I was just in the middle of something, I’m guessing it on.

16 00:01:48.990 ⇒ 00:01:50.560 Mustafa Raja: I was excited about it, though.

17 00:01:51.080 ⇒ 00:01:52.670 Mustafa Raja: Then I was disappointed.

18 00:01:54.010 ⇒ 00:01:55.840 Samuel Roberts: Yeah, yeah, well, you know.

19 00:01:56.370 ⇒ 00:01:57.340 Samuel Roberts: It happens.

20 00:01:58.480 ⇒ 00:01:59.520 Samuel Roberts: Alright.

21 00:02:02.370 ⇒ 00:02:03.070 Samuel Roberts: Okay.

22 00:02:05.910 ⇒ 00:02:08.650 Samuel Roberts: Where are we at here?

23 00:02:08.650 ⇒ 00:02:13.689 Casie Aviles: Yeah, so I just have this, got jerked up right now.

24 00:02:15.880 ⇒ 00:02:22.359 Casie Aviles: So, right now, I think what we focused on is just the forms, and I’m still aiming to have…

25 00:02:23.000 ⇒ 00:02:28.170 Casie Aviles: Like, the working version, within this week, so…

26 00:02:28.400 ⇒ 00:02:33.809 Casie Aviles: Okay. This should be fine, so I guess what we need to…

27 00:02:36.140 ⇒ 00:02:39.379 Casie Aviles: Recalibrate would be the ones here.

28 00:02:40.460 ⇒ 00:02:50.040 Casie Aviles: That are above… Yeah, and then also, like, take into account maintenance, because that’s also something that…

29 00:02:50.610 ⇒ 00:02:53.230 Casie Aviles: We… I believe we missed, so…

30 00:02:53.360 ⇒ 00:02:58.360 Casie Aviles: That’s also something Amber raised, where we are starting to get more Mmm…

31 00:02:59.030 ⇒ 00:03:01.669 Casie Aviles: what do you call this? Triage tickets again.

32 00:03:02.780 ⇒ 00:03:03.760 Casie Aviles: So…

33 00:03:04.700 ⇒ 00:03:08.790 Uttam Kumaran: Yeah, so I think we should put, like, we should just have, like, a monthly…

34 00:03:09.970 ⇒ 00:03:11.699 Uttam Kumaran: I mean, either we should do…

35 00:03:13.300 ⇒ 00:03:18.859 Uttam Kumaran: Like, weekly, or, like, a basic… but you want to have a through line of, like, maintenance.

36 00:03:19.040 ⇒ 00:03:23.020 Uttam Kumaran: But ideally, we think about, like, what are the types of maintenance that we have to do?

37 00:03:23.400 ⇒ 00:03:25.979 Uttam Kumaran: And we create a line for each of those.

38 00:03:26.600 ⇒ 00:03:31.699 Uttam Kumaran: Because basically, what I want to indicate is now there’s at least X amount of bucket of hours.

39 00:03:31.810 ⇒ 00:03:35.470 Uttam Kumaran: We’re going to just maintain the system, not… not knowing, you know?

40 00:03:35.750 ⇒ 00:03:42.820 Uttam Kumaran: So you can create a maintenance project, or, like, an ongoing maintenance project, and just list the types of maintenance, and just put along in each.

41 00:03:54.300 ⇒ 00:03:56.260 Casie Aviles: Okay… You know.

42 00:04:07.440 ⇒ 00:04:14.739 Casie Aviles: And, so far, for maintenance, how, like, how many hours should we be.

43 00:04:15.040 ⇒ 00:04:18.810 Amber Lin: I’m thinking… in the triage, there’s…

44 00:04:19.000 ⇒ 00:04:27.669 Amber Lin: Are zip code-related issues, and then there’s more of the Central Dock-related issues.

45 00:04:29.820 ⇒ 00:04:43.119 Amber Lin: And then, I think the last category is more of our system issues. So, say if Andy goes down, or, our infrastructure can’t handle it that day, or something goes down, I think that’s another bucket.

46 00:04:43.430 ⇒ 00:04:44.810 Amber Lin: of issues.

47 00:04:49.310 ⇒ 00:04:49.930 Casie Aviles: Okay.

48 00:05:03.610 ⇒ 00:05:06.599 Casie Aviles: Oh… Oh, Central Dublin.

49 00:05:26.430 ⇒ 00:05:28.840 Uttam Kumaran: So I’m just going to share this to everyone.

50 00:05:38.850 ⇒ 00:05:50.270 Casie Aviles: And so I know we also have, like, tasks over here that, we’ll likely… So…

51 00:05:53.910 ⇒ 00:06:00.179 Casie Aviles: for this, so… what’s… I guess, like, what’s the best way to start with this? Should we go, like.

52 00:06:01.710 ⇒ 00:06:09.320 Uttam Kumaran: Yeah, I think we go… well, one, let’s just make sure we have all the sections, that we want. So the first section is Migrate, main, Andy Logic.

53 00:06:09.840 ⇒ 00:06:13.270 Uttam Kumaran: Right, so let’s just go one by one through the core sections.

54 00:06:14.910 ⇒ 00:06:23.219 Uttam Kumaran: And then I think in parallel, Amber, you want to just make sure… I want to make sure our sections are related to the goals. Like…

55 00:06:23.710 ⇒ 00:06:27.659 Uttam Kumaran: So, kind of, like, my take is that clients don’t want to see, like.

56 00:06:28.220 ⇒ 00:06:33.880 Uttam Kumaran: maybe, like, migrate main AND logic, they want to, like, understand, like, what is… how does that affect the system?

57 00:06:34.030 ⇒ 00:06:42.129 Uttam Kumaran: So I don’t mind us having that, but as much as possible, I want the sections to adhere to the… to the list of goals that we have in that Notion.

58 00:06:42.400 ⇒ 00:06:43.749 Uttam Kumaran: So maybe you can…

59 00:06:44.280 ⇒ 00:06:52.409 Uttam Kumaran: You could share that as well, so that everybody can see both screens. We kind of go back and forth, I don’t know what you think. I do think it’d be great to see that.

60 00:06:52.410 ⇒ 00:06:53.140 Samuel Roberts: Yeah.

61 00:06:55.010 ⇒ 00:06:59.410 Amber Lin: See… Where are the goals listed?

62 00:07:00.280 ⇒ 00:07:01.650 Uttam Kumaran: Well, you wrote the goals.

63 00:07:03.250 ⇒ 00:07:05.930 Amber Lin: Oh, I see, I see.

64 00:07:06.130 ⇒ 00:07:07.870 Uttam Kumaran: Like, the Andy 2026?

65 00:07:08.930 ⇒ 00:07:23.079 Amber Lin: Okay, I think that’s separate from this migration. That’s what your vet would like their department to be able to do. I think they’re not as in sync with this migration workflow.

66 00:07:23.080 ⇒ 00:07:27.840 Uttam Kumaran: No, no, no, but they’re all the same. Like, we’re only… there’s no reason to do the migration.

67 00:07:28.620 ⇒ 00:07:30.830 Uttam Kumaran: to achieve nothing, right? So…

68 00:07:31.960 ⇒ 00:07:36.850 Uttam Kumaran: Like, we have to start with what the client goals are, which is the ones that you listed here.

69 00:07:37.690 ⇒ 00:07:44.630 Amber Lin: I see. Some of them sounds like new features or things we need to add.

70 00:07:44.890 ⇒ 00:07:49.289 Amber Lin: at what point should I add in those goals? Because…

71 00:07:49.290 ⇒ 00:07:50.629 Uttam Kumaran: Whatever you think can end up here?

72 00:07:51.920 ⇒ 00:07:53.910 Amber Lin: Okay.

73 00:07:55.400 ⇒ 00:07:56.030 Amber Lin: Stunning.

74 00:07:56.030 ⇒ 00:07:58.230 Uttam Kumaran: Why that makes you… why that makes you nervous.

75 00:07:58.620 ⇒ 00:08:04.550 Amber Lin: No, no, I’m thinking, because the first part is a lot of migration and building the infrastructure.

76 00:08:05.490 ⇒ 00:08:13.969 Uttam Kumaran: But in order… we have to do that in order to do these goals, right? So, what our job is to say, like, you want to achieve your goals.

77 00:08:14.160 ⇒ 00:08:16.800 Uttam Kumaran: Here is, like, everything, and then…

78 00:08:17.110 ⇒ 00:08:21.570 Uttam Kumaran: me and you can start to say, like, okay, what can we paralyze, or if you’re like, look.

79 00:08:22.050 ⇒ 00:08:36.209 Uttam Kumaran: if we put everything, it’s gonna take 9 months, I need to figure out in 6 months, okay, like, let me go… we can go maybe ask for more budget, I can see if we can hire someone else, we can start to move things around. That’s sort of my thought, is that we at least

80 00:08:36.360 ⇒ 00:08:37.230 Uttam Kumaran: halves.

81 00:08:37.600 ⇒ 00:08:39.250 Uttam Kumaran: All of the core…

82 00:08:39.520 ⇒ 00:08:46.749 Uttam Kumaran: goals listed here. Some of these are going to be related to the migration, but the migration… we’re doing the migration in order to

83 00:08:46.880 ⇒ 00:08:48.330 Uttam Kumaran: Execute on those.

84 00:08:49.110 ⇒ 00:08:50.150 Amber Lin: Let’s see…

85 00:08:50.190 ⇒ 00:08:53.420 Uttam Kumaran: And some parts of the migration we may find that aren’t important.

86 00:08:53.620 ⇒ 00:08:55.950 Uttam Kumaran: Short term, right? And we can move them around.

87 00:08:56.990 ⇒ 00:09:15.080 Amber Lin: I see. I would… I will start… I’ll put them at the end first, because I want to know where we’re at and when we’re able to even do these goals, if we’re doing migration first. Because initially, I thought we’re just going to build these in parallel, because I don’t know how long the migration is supposed to take.

88 00:09:15.350 ⇒ 00:09:18.330 Amber Lin: Is there an estimated time when we’ll be able.

89 00:09:18.330 ⇒ 00:09:19.680 Uttam Kumaran: No, that’s what we’re doing now.

90 00:09:19.900 ⇒ 00:09:21.960 Amber Lin: Okay, cool. So I added… I’ll add it…

91 00:09:21.960 ⇒ 00:09:37.140 Uttam Kumaran: We’re doing everything in this meeting. Yeah, we’re doing everything in this… basically, as much as we can get on everything related to Andy in the game chart is the bull here. So I’m not… don’t separate that from my… like, I don’t want us to separate out migration and new feature development.

92 00:09:37.300 ⇒ 00:09:40.339 Uttam Kumaran: Because they don’t think, like, we’re two teams. We’re just one team.

93 00:09:40.340 ⇒ 00:09:40.660 Samuel Roberts: Okay.

94 00:09:40.660 ⇒ 00:09:50.889 Uttam Kumaran: all doing… we’re gonna do as much in parallel as we can. I think what I’ve… what I’ve asked the… what I’ve asked Sam, you, Casey, and stuff in the past is just get all the steps here.

95 00:09:51.140 ⇒ 00:09:56.869 Uttam Kumaran: So for me, I… what I want to know is, like, are all of the… Julie, are we…

96 00:09:57.230 ⇒ 00:10:03.050 Uttam Kumaran: Where all of the… there’s… right now, looking at this, there are things that, like these pieces.

97 00:10:03.630 ⇒ 00:10:10.720 Uttam Kumaran: that we’ve… we’re sort of like, are… is everything here, I guess my question is, updated as done? That is done, basically.

98 00:10:12.500 ⇒ 00:10:15.360 Uttam Kumaran: And technically, anything that’s not done needs to be moved.

99 00:10:15.840 ⇒ 00:10:18.300 Uttam Kumaran: Like, past the due date, past today, right?

100 00:10:18.300 ⇒ 00:10:21.949 Mustafa Raja: Yeah, so the done ones are pretty maintained.

101 00:10:22.210 ⇒ 00:10:24.570 Mustafa Raja: So the ones that are marked and are done.

102 00:10:25.490 ⇒ 00:10:29.040 Uttam Kumaran: So let’s just… let’s just… let’s just go from the top down.

103 00:10:29.720 ⇒ 00:10:33.290 Uttam Kumaran: But we have Migrate the main Andy logged in. We have, so…

104 00:10:33.700 ⇒ 00:10:38.840 Uttam Kumaran: If something is not done, let’s just move it at least to…

105 00:10:39.490 ⇒ 00:10:40.480 Samuel Roberts: Next week.

106 00:10:40.590 ⇒ 00:10:41.440 Uttam Kumaran: For now.

107 00:10:42.400 ⇒ 00:10:46.099 Uttam Kumaran: So don’t change the duration, necessarily, just move it all the next week.

108 00:10:46.470 ⇒ 00:10:47.920 Uttam Kumaran: That way, we can at least

109 00:10:48.600 ⇒ 00:10:56.009 Uttam Kumaran: we can then move things to this speaker, you know, as needed. So that staffing, the integrate structured logging.

110 00:10:56.910 ⇒ 00:10:58.020 Uttam Kumaran: this piece.

111 00:10:59.300 ⇒ 00:11:03.920 Uttam Kumaran: And is there a way to indicate, like, priority at all in Segant?

112 00:11:07.600 ⇒ 00:11:11.999 Samuel Roberts: Priority might just… I mean, I don’t know if there’s a… Field for that, Hammond.

113 00:11:14.240 ⇒ 00:11:15.010 Uttam Kumaran: I mean, I guess…

114 00:11:15.010 ⇒ 00:11:16.370 Samuel Roberts: and costs…

115 00:11:16.370 ⇒ 00:11:18.529 Casie Aviles: It’s a color. Okay.

116 00:11:18.530 ⇒ 00:11:19.490 Uttam Kumaran: Well, it’s fine.

117 00:11:20.080 ⇒ 00:11:23.329 Samuel Roberts: Oh, there’s priority, there’s a priority field in the table, yeah.

118 00:11:23.700 ⇒ 00:11:26.829 Samuel Roberts: It’s hard to see on this view, but if you flip over to the table view.

119 00:11:29.930 ⇒ 00:11:32.850 Samuel Roberts: There’s, priority somewhere there.

120 00:11:34.000 ⇒ 00:11:34.679 Uttam Kumaran: Oh, here.

121 00:11:35.120 ⇒ 00:11:36.100 Samuel Roberts: Yeah.

122 00:11:37.240 ⇒ 00:11:39.259 Uttam Kumaran: To click on, click on one of those?

123 00:11:40.230 ⇒ 00:11:41.929 Uttam Kumaran: What options do we have? Okay.

124 00:11:42.580 ⇒ 00:11:48.590 Uttam Kumaran: So, I guess my… let’s just… start by…

125 00:11:49.120 ⇒ 00:11:54.519 Uttam Kumaran: Like, let’s just assign each of these 1, 3, or 5, like, priority.

126 00:11:57.560 ⇒ 00:11:59.079 Casie Aviles: I can let you guys…

127 00:11:59.080 ⇒ 00:12:02.069 Uttam Kumaran: Go ahead and do that, and then we can move to the timeline view.

128 00:12:04.770 ⇒ 00:12:06.900 Uttam Kumaran: Don’t use two or four, since it’s fine.

129 00:12:07.100 ⇒ 00:12:08.110 Uttam Kumaran: Beautiful.

130 00:12:10.970 ⇒ 00:12:12.979 Casie Aviles: Oh, so 1, 3, or 5, okay.

131 00:12:12.980 ⇒ 00:12:13.450 Samuel Roberts: Yeah.

132 00:12:13.450 ⇒ 00:12:15.189 Uttam Kumaran: Yeah, I would do one for your clients.

133 00:12:16.840 ⇒ 00:12:21.159 Uttam Kumaran: And then while you guys are doing that, maybe… maybe, Amber, what you can do is…

134 00:12:21.270 ⇒ 00:12:30.079 Uttam Kumaran: For your… for the… for the goals that you listed on the Notion doc, if you can just think about them as high-level sections or feature releases.

135 00:12:30.080 ⇒ 00:12:31.359 Amber Lin: Yeah, I’m on…

136 00:12:31.360 ⇒ 00:12:32.669 Uttam Kumaran: I’m not in Wonder Wong.

137 00:12:32.980 ⇒ 00:12:33.700 Uttam Kumaran: Cool. Okay, cool.

138 00:12:33.700 ⇒ 00:12:34.030 Amber Lin: Okay.

139 00:12:34.030 ⇒ 00:12:34.640 Uttam Kumaran: Okay, fair.

140 00:12:35.090 ⇒ 00:12:39.629 Uttam Kumaran: So, and then I’m gonna also do some Instagram cleanup while I sort of cleaning that stuff.

141 00:12:43.010 ⇒ 00:12:43.830 Samuel Roberts: Alright.

142 00:12:43.980 ⇒ 00:12:45.730 Samuel Roberts: lower the monster energy.

143 00:12:46.010 ⇒ 00:12:52.399 Samuel Roberts: endpoint down to 1, because that’s gonna follow after these other things, I think, once it’s a little more stable.

144 00:12:52.840 ⇒ 00:12:53.540 Casie Aviles: Okay.

145 00:12:55.590 ⇒ 00:13:03.969 Mustafa Raja: So, for Master Entry Point, we do have a task for… Selecting our architecture, right?

146 00:13:06.230 ⇒ 00:13:07.970 Samuel Roberts: Selecting our architecture.

147 00:13:08.350 ⇒ 00:13:09.450 Samuel Roberts: Or…

148 00:13:09.640 ⇒ 00:13:13.600 Mustafa Raja: Yeah So, which one we’re committing to, right?

149 00:13:13.840 ⇒ 00:13:17.380 Samuel Roberts: That’s what I mean, like, I’m not as worried about figuring out the entry point until we do that.

150 00:13:17.880 ⇒ 00:13:18.740 Samuel Roberts: I think it’s a lesser.

151 00:13:18.740 ⇒ 00:13:19.220 Mustafa Raja: Priority there.

152 00:13:19.950 ⇒ 00:13:20.890 Samuel Roberts: than that.

153 00:13:21.160 ⇒ 00:13:23.269 Mustafa Raja: But is it mentioned? Yeah, is what I’m.

154 00:13:23.270 ⇒ 00:13:24.950 Samuel Roberts: Oh, is it… I see what you’re saying.

155 00:13:27.760 ⇒ 00:13:29.359 Mustafa Raja: It… I don’t think it is.

156 00:13:29.640 ⇒ 00:13:35.170 Samuel Roberts: No, that’s probably, like, a… A good one to add, maybe even, like, a milestone.

157 00:13:35.170 ⇒ 00:13:35.920 Mustafa Raja: Yeah.

158 00:13:45.030 ⇒ 00:13:54.799 Samuel Roberts: Yeah, I would, what was it to do at Milestone? There’s… A new task,

159 00:13:58.120 ⇒ 00:14:01.370 Samuel Roberts: Determining… Fast.

160 00:14:01.370 ⇒ 00:14:08.830 Mustafa Raja: One more thing, Tim said that he will, give the BigQuery key today.

161 00:14:08.940 ⇒ 00:14:11.429 Mustafa Raja: It’s nearly end of day, should I ping him?

162 00:14:14.290 ⇒ 00:14:16.920 Uttam Kumaran: Oh, I don’t know.

163 00:14:18.000 ⇒ 00:14:21.110 Uttam Kumaran: Why don’t you see, wait till the end, wait till 5 o’clock.

164 00:14:21.110 ⇒ 00:14:24.849 Samuel Roberts: Yeah… Is it… I mean, is that critical to do something with today?

165 00:14:26.110 ⇒ 00:14:26.630 Mustafa Raja: Good luck.

166 00:14:27.070 ⇒ 00:14:28.340 Mustafa Raja: Query to mention that?

167 00:14:29.270 ⇒ 00:14:31.279 Uttam Kumaran: Yeah, big… Oh, it’s all this stuff.

168 00:14:31.280 ⇒ 00:14:33.090 Samuel Roberts: Oh, I see, I see, yeah, I mean…

169 00:14:34.260 ⇒ 00:14:37.140 Samuel Roberts: Or is it already? Yeah, yeah, probably wait till the end of the day.

170 00:14:40.150 ⇒ 00:14:43.659 Uttam Kumaran: So can I add that as a task? What is that? What do we need from him? What do we need?

171 00:14:43.660 ⇒ 00:14:45.380 Mustafa Raja: Service account.

172 00:14:45.680 ⇒ 00:14:48.900 Mustafa Raja: A service account with, BigQuery rules.

173 00:14:50.770 ⇒ 00:14:51.320 Casie Aviles: H.

174 00:14:51.470 ⇒ 00:14:54.860 Casie Aviles: Is it different from just the init bingQuery?

175 00:14:54.860 ⇒ 00:14:55.680 Mustafa Raja: Yes.

176 00:14:56.060 ⇒ 00:15:06.400 Uttam Kumaran: It’s the… it’s… this is the depend… this is the depending task, basically, or the blocking task. So what I’m gonna do, Mustafa, is I’m gonna add this as a milestone.

177 00:15:07.250 ⇒ 00:15:10.679 Uttam Kumaran: And I’m gonna put this as, like, here, and then…

178 00:15:10.940 ⇒ 00:15:18.780 Uttam Kumaran: basically, we can’t do BigQuery until this, right? So basically, I’m gonna… Mark.

179 00:15:19.630 ⇒ 00:15:26.420 Uttam Kumaran: acquire BQ service account as the blocking task for init BigQuery. But it’s gonna take 2 weeks to do BigQuery.

180 00:15:28.670 ⇒ 00:15:29.619 Uttam Kumaran: Probably just need, like.

181 00:15:29.620 ⇒ 00:15:34.810 Mustafa Raja: No, no, no, so, Avish and I were able to work, work…

182 00:15:35.140 ⇒ 00:15:45.739 Mustafa Raja: most of it, so, it’s almost done once, once we get the key, I’ll just, put the data in. So we have the repo and all.

183 00:15:46.050 ⇒ 00:15:47.370 Mustafa Raja: And that isn’t an issue.

184 00:15:47.370 ⇒ 00:15:48.160 Uttam Kumaran: Oh, okay, okay.

185 00:15:48.160 ⇒ 00:15:51.990 Mustafa Raja: Yeah, we worked on it a little bit, just waiting on the key.

186 00:15:53.960 ⇒ 00:15:54.820 Mustafa Raja: Oh, what is a key?

187 00:15:54.820 ⇒ 00:15:55.590 Uttam Kumaran: Yamak.

188 00:15:56.710 ⇒ 00:16:03.150 Mustafa Raja: The key unlocks for us to programmatically access BigQuery.

189 00:16:04.110 ⇒ 00:16:05.329 Uttam Kumaran: Okay, okay.

190 00:16:10.500 ⇒ 00:16:12.370 Samuel Roberts: And this is all for blogging, right?

191 00:16:14.100 ⇒ 00:16:16.240 Uttam Kumaran: And I need it for discovery work.

192 00:16:16.660 ⇒ 00:16:18.070 Samuel Roberts: Oh, okay, okay.

193 00:16:18.560 ⇒ 00:16:19.340 Uttam Kumaran: the pulse.

194 00:16:20.480 ⇒ 00:16:23.379 Uttam Kumaran: Otherwise, I wouldn’t have, like, prioritized it that much.

195 00:16:23.380 ⇒ 00:16:24.630 Samuel Roberts: Yeah, no, that makes sense.

196 00:16:55.360 ⇒ 00:16:56.360 Samuel Roberts: Where are we?

197 00:17:01.290 ⇒ 00:17:04.200 Uttam Kumaran: So, is all… everything has, like, priorities on it, basically?

198 00:17:04.730 ⇒ 00:17:09.299 Samuel Roberts: And… Yeah, in the first.

199 00:17:09.430 ⇒ 00:17:11.270 Casie Aviles: For the first section.

200 00:17:11.849 ⇒ 00:17:13.869 Uttam Kumaran: I would rip through all of those.

201 00:17:13.869 ⇒ 00:17:15.109 Samuel Roberts: Yeah, we’ll just keep going.

202 00:17:16.000 ⇒ 00:17:28.639 Uttam Kumaran: Yeah, through all of them, because what this is going to allow us to do is see the low priority, we should move as far as possible, as long as it doesn’t have dependencies, you know what I mean? So, just to kind of do some optimization here.

203 00:17:39.420 ⇒ 00:17:45.860 Samuel Roberts: Okay, so for… Yeah, I mean, I think this looks good for the first section.

204 00:17:46.070 ⇒ 00:17:49.730 Samuel Roberts: For routing and RAG migrations.

205 00:17:50.270 ⇒ 00:18:00.899 Samuel Roberts: Yeah, that’s a good one. Load testing and ride under concurrent usage is Lower, I think.

206 00:18:04.250 ⇒ 00:18:05.930 Casie Aviles: This is a bit vague.

207 00:18:06.190 ⇒ 00:18:07.490 Casie Aviles: Yeah.

208 00:18:07.490 ⇒ 00:18:12.199 Samuel Roberts: I think we need to push that out a little bit, but I think it’s just making sure that it can handle

209 00:18:13.780 ⇒ 00:18:14.630 Samuel Roberts: you know.

210 00:18:14.810 ⇒ 00:18:18.880 Samuel Roberts: more concurrent users, then, you know, we’re just testing one-offs at a time, but I think that’s.

211 00:18:18.880 ⇒ 00:18:19.640 Casie Aviles: Amazing.

212 00:18:21.020 ⇒ 00:18:25.609 Samuel Roberts: And fix performance or correctness issues. Yeah, that’s a good one.

213 00:18:29.100 ⇒ 00:18:34.800 Samuel Roberts: Okay, by the way, text SQL… User feedback mapping.

214 00:18:35.090 ⇒ 00:18:36.420 Samuel Roberts: Yeah, okay.

215 00:18:37.430 ⇒ 00:18:38.310 Samuel Roberts: Yeah.

216 00:18:38.770 ⇒ 00:18:40.809 Casie Aviles: I think these are lower priority.

217 00:18:41.090 ⇒ 00:18:47.970 Samuel Roberts: Yeah… I mean, they’re… they’re lower in that, like, monster’s gotta get, you know, it’s not important until that’s…

218 00:18:51.150 ⇒ 00:18:52.310 Casie Aviles: It’s working.

219 00:18:54.460 ⇒ 00:18:58.759 Samuel Roberts: So yeah, I would say user feedback, caption logging is like a… yeah.

220 00:18:59.270 ⇒ 00:19:01.670 Samuel Roberts: We can bump that down.

221 00:19:01.920 ⇒ 00:19:03.359 Casie Aviles: 10 segmento.

222 00:19:04.070 ⇒ 00:19:06.029 Samuel Roberts: Oh, by the way, Generation.

223 00:19:06.820 ⇒ 00:19:08.030 Samuel Roberts: Yeah.

224 00:19:08.940 ⇒ 00:19:12.980 Samuel Roberts: Yeah, that’s fine. And then end-to-end testing of enriched responses.

225 00:19:14.020 ⇒ 00:19:17.380 Samuel Roberts: Yeah, 3 sounds fine.

226 00:19:20.010 ⇒ 00:19:22.540 Samuel Roberts: Legacy and Independencies for 9AM.

227 00:19:24.020 ⇒ 00:19:27.580 Samuel Roberts: Is that just making sure we’re not using any of the other flows there?

228 00:19:29.280 ⇒ 00:19:35.040 Casie Aviles: Yeah, I think… This might be, like, you know, making sure we’re not using anything on N8N anymore.

229 00:19:35.040 ⇒ 00:19:38.680 Samuel Roberts: Right, right, so yeah, we’ve got the main… yeah, okay, that’s fine, that’s… that’s a good one.

230 00:19:39.490 ⇒ 00:19:40.750 Samuel Roberts: Yeah, okay.

231 00:19:41.930 ⇒ 00:19:45.030 Uttam Kumaran: You can’t hide weekends on this, or no?

232 00:19:46.130 ⇒ 00:19:47.630 Samuel Roberts: Oh, good question.

233 00:19:51.880 ⇒ 00:19:53.380 Samuel Roberts: Let’s go ahead and share, and…

234 00:19:53.380 ⇒ 00:19:54.609 Uttam Kumaran: That’s… that’s common.

235 00:19:54.610 ⇒ 00:19:55.660 Samuel Roberts: accustomed to,

236 00:20:02.300 ⇒ 00:20:03.619 Samuel Roberts: Huh, yeah, that’s crazy.

237 00:20:10.480 ⇒ 00:20:12.370 Samuel Roberts: A little trash can up there.

238 00:20:12.590 ⇒ 00:20:15.080 Samuel Roberts: That’s the holidays, I think, the block.

239 00:20:27.400 ⇒ 00:20:34.029 Casie Aviles: A lot of these feel a bit higher prio, but I’m trying to… Understand that approach. Yeah.

240 00:20:34.250 ⇒ 00:20:35.730 Casie Aviles: Can be bumped down, or…

241 00:20:36.490 ⇒ 00:20:36.970 Samuel Roberts: Right.

242 00:20:36.970 ⇒ 00:20:41.439 Uttam Kumaran: It’s okay, you don’t have to… you don’t necessarily have to have a quota of low or high. They’re all high.

243 00:20:41.440 ⇒ 00:20:44.829 Samuel Roberts: Yeah, yeah. If they’re… yeah, if they’re important, they’re important. You know, it’s not like.

244 00:20:44.830 ⇒ 00:20:49.070 Uttam Kumaran: Yeah, market as such. There’s no quota on whether it is or isn’t, but…

245 00:20:49.260 ⇒ 00:20:55.119 Uttam Kumaran: Ideally, there are some that… the ones that aren’t, those are the ones we’ll kick as far as we can.

246 00:20:59.950 ⇒ 00:21:07.320 Samuel Roberts: Figure alerting channels… Lrl, enter… alright, yeah.

247 00:21:09.320 ⇒ 00:21:15.850 Samuel Roberts: Yeah, I… Yeah, alert… how do alerts are risk-related scenarios.

248 00:21:18.180 ⇒ 00:21:21.969 Samuel Roberts: Tune thresholds to reduce… what thresholds is this referring to?

249 00:21:22.180 ⇒ 00:21:23.899 Casie Aviles: Yeah, this one’s vague as well.

250 00:21:28.250 ⇒ 00:21:31.200 Casie Aviles: I think I can remove this task.

251 00:21:32.240 ⇒ 00:21:35.380 Casie Aviles: If it doesn’t… it’s not very helpful right now.

252 00:21:35.380 ⇒ 00:21:38.089 Samuel Roberts: Yeah, I don’t know what that’s referring to. I mean, unless it’s just…

253 00:21:40.690 ⇒ 00:21:45.849 Samuel Roberts: I mean, I can make guesses, but… I don’t think it’s…

254 00:21:47.570 ⇒ 00:21:49.349 Casie Aviles: Not sure how we can get rid of this.

255 00:21:49.350 ⇒ 00:21:55.650 Samuel Roberts: Yeah, I was gonna… I, I, right-click… no, okay.

256 00:21:57.340 ⇒ 00:21:58.030 Samuel Roberts: Huh.

257 00:21:58.460 ⇒ 00:22:00.849 Samuel Roberts: Oh, I think if you click on it,

258 00:22:01.640 ⇒ 00:22:05.259 Samuel Roberts: It should open up on the side. Oh yeah, close out the since you’ve been gone.

259 00:22:07.750 ⇒ 00:22:09.250 Samuel Roberts: And then they should be like, yeah.

260 00:22:13.670 ⇒ 00:22:17.239 Samuel Roberts: validated alerts by a forced failure scenario. I think that’s a…

261 00:22:17.810 ⇒ 00:22:21.689 Samuel Roberts: Still high-ish, but not as high as my thought. Like.

262 00:22:22.570 ⇒ 00:22:25.130 Samuel Roberts: Priority-wise, everything else is more important.

263 00:22:25.530 ⇒ 00:22:28.130 Samuel Roberts: Validating that is gonna come after anyway, so…

264 00:22:31.990 ⇒ 00:22:35.030 Samuel Roberts: I guess I could… we could add dependencies there, but…

265 00:22:36.670 ⇒ 00:22:40.980 Samuel Roberts: Let’s just keep doing the prior. Configure Cloud Moon… oh, okay.

266 00:22:44.920 ⇒ 00:22:46.880 Mustafa Raja: 45 is duplicate, no?

267 00:22:48.360 ⇒ 00:22:54.630 Samuel Roberts: You commission meeting and workflows? Yeah, I think we have that up with the… that’ll be part of the one we just talked about, right?

268 00:22:54.630 ⇒ 00:22:57.730 Mustafa Raja: Yeah… I’m going to delete this one, though.

269 00:22:57.730 ⇒ 00:22:58.420 Samuel Roberts: hands.

270 00:22:59.640 ⇒ 00:23:00.799 Samuel Roberts: I’m assuming you…

271 00:23:20.570 ⇒ 00:23:21.240 Samuel Roberts: Hmm.

272 00:23:23.390 ⇒ 00:23:28.089 Samuel Roberts: Where are we at now? Embedding pipeline, replicating deploy, embedding pipeline tomorrow, so…

273 00:23:37.920 ⇒ 00:23:42.410 Samuel Roberts: Where the embedding pipeline is… a…

274 00:23:43.460 ⇒ 00:23:44.000 Mustafa Raja: And it didn’t.

275 00:23:44.000 ⇒ 00:23:50.330 Samuel Roberts: Asna and flow, okay, I wasn’t sure if we were doing that somewhere else, because that’s why I was hoping to maybe bump that down if it’s not part of any info.

276 00:23:50.510 ⇒ 00:23:53.399 Samuel Roberts: Couldn’t remember where that all happens. Alright, yeah, definitely keep that up.

277 00:23:53.900 ⇒ 00:23:58.350 Samuel Roberts: The maintenance is ongoing, I don’t know if we need to…

278 00:24:00.030 ⇒ 00:24:03.609 Samuel Roberts: probably all that. And then fixes…

279 00:24:07.620 ⇒ 00:24:13.210 Samuel Roberts: Yeah, that’s good. Link service areas, that’s service names to service names.

280 00:24:15.010 ⇒ 00:24:18.710 Mustafa Raja: And the references aren’t created between the two.

281 00:24:19.780 ⇒ 00:24:21.060 Samuel Roberts: Hmm, okay.

282 00:24:21.410 ⇒ 00:24:25.790 Mustafa Raja: Yeah, so same thing might have multiple names is what happens.

283 00:24:25.790 ⇒ 00:24:26.980 Samuel Roberts: Got it, okay.

284 00:24:31.110 ⇒ 00:24:32.649 Casie Aviles: Let me rename this one.

285 00:24:33.170 ⇒ 00:24:33.960 Samuel Roberts: Which one?

286 00:24:34.930 ⇒ 00:24:36.360 Casie Aviles: Yeah, it’s good.

287 00:24:37.100 ⇒ 00:24:40.060 Samuel Roberts: Yeah, it’s Admin UI for Zips DB or something.

288 00:24:44.820 ⇒ 00:24:45.510 Samuel Roberts: Yeah.

289 00:24:52.990 ⇒ 00:24:54.150 Samuel Roberts: Oops.

290 00:24:54.680 ⇒ 00:24:55.400 Samuel Roberts: The damage?

291 00:24:55.400 ⇒ 00:24:56.110 Casie Aviles: this…

292 00:24:56.110 ⇒ 00:24:56.730 Samuel Roberts: Oh, okay.

293 00:24:57.810 ⇒ 00:24:58.560 Samuel Roberts: I was like, what?

294 00:25:05.810 ⇒ 00:25:06.420 Samuel Roberts: Is that…

295 00:25:06.700 ⇒ 00:25:10.249 Casie Aviles: Oh, okay, yeah, I understand this one.

296 00:25:10.250 ⇒ 00:25:11.030 Samuel Roberts: Okay.

297 00:25:11.030 ⇒ 00:25:16.890 Casie Aviles: Well, I actually did… I did this already, but it’s in, like, a… in a test.

298 00:25:17.270 ⇒ 00:25:18.579 Casie Aviles: Table that is not…

299 00:25:18.580 ⇒ 00:25:24.350 Samuel Roberts: Yeah, yeah, don’t check it yet, then. Yeah, yeah, when we move all that to… when… I think that’s part of, like, the admin UI stuff.

300 00:25:25.700 ⇒ 00:25:27.720 Samuel Roberts: You had with the dev… the dev tables?

301 00:25:28.720 ⇒ 00:25:39.219 Amber Lin: Did we end up adding the different aliases that people use? Because there’s a lot of different names that… it’s not even the same as the service names. Like, did we ever end up…

302 00:25:39.470 ⇒ 00:25:44.120 Casie Aviles: We did… Yeah, we do constantly, like, update the prompt.

303 00:25:44.400 ⇒ 00:25:48.860 Casie Aviles: That handles the querying. That’s where I put, like, the aliases.

304 00:25:48.970 ⇒ 00:25:49.830 Casie Aviles: To hand them.

305 00:25:50.430 ⇒ 00:25:51.000 Amber Lin: Yeah.

306 00:25:51.640 ⇒ 00:25:54.009 Samuel Roberts: Do we have, like, a master list somewhere?

307 00:25:54.890 ⇒ 00:25:56.399 Samuel Roberts: That lives outside the prompt.

308 00:25:57.520 ⇒ 00:25:59.889 Casie Aviles: No, it’s just within the prompt.

309 00:26:00.170 ⇒ 00:26:01.010 Samuel Roberts: Okay.

310 00:26:02.880 ⇒ 00:26:13.349 Mustafa Raja: And do we know they… do we have documentation of the responses that they… that the CSRs need exact response from, Central Doc?

311 00:26:15.410 ⇒ 00:26:16.979 Mustafa Raja: Like, the cancellation flows.

312 00:26:16.980 ⇒ 00:26:22.320 Samuel Roberts: I was gonna say, I know at least the cancellation flow. Are there other ones? I mean, I know there are other ones, I don’t know which ones they are, specifically.

313 00:26:22.320 ⇒ 00:26:23.120 Mustafa Raja: I’m like…

314 00:26:26.080 ⇒ 00:26:27.549 Samuel Roberts: Amber, do you know that offhand?

315 00:26:27.820 ⇒ 00:26:29.640 Amber Lin: Sorry, what was that?

316 00:26:29.870 ⇒ 00:26:37.380 Samuel Roberts: The, there’s certain questions to Andy that they want, like, the verbatim response from the central doc.

317 00:26:38.100 ⇒ 00:26:38.450 Amber Lin: Cancelation.

318 00:26:38.450 ⇒ 00:26:39.920 Samuel Roberts: Low script and stuff?

319 00:26:40.810 ⇒ 00:26:41.170 Amber Lin: Yeah.

320 00:26:41.170 ⇒ 00:26:44.670 Samuel Roberts: Are there… or what are… do you know what the other ones are for that?

321 00:26:44.670 ⇒ 00:26:58.360 Amber Lin: I would say the other ones are the templates, which we also made a custom solution in putting that in a spreadsheet and returning the cell answers, which I don’t know if it’s sustainable. So, templates.

322 00:26:58.360 ⇒ 00:26:58.950 Samuel Roberts: What’d you say?

323 00:26:58.950 ⇒ 00:26:59.530 Amber Lin: everyone.

324 00:26:59.560 ⇒ 00:27:00.150 Mustafa Raja: I’m singing…

325 00:27:00.150 ⇒ 00:27:00.660 Samuel Roberts: Meh.

326 00:27:01.070 ⇒ 00:27:06.780 Mustafa Raja: Yeah, I think… Andy… And it does not have that anymore.

327 00:27:08.480 ⇒ 00:27:09.220 Amber Lin: Hmm.

328 00:27:10.040 ⇒ 00:27:13.070 Samuel Roberts: But there is a list of those that we can… those are… so those aren’t in a central document.

329 00:27:13.070 ⇒ 00:27:14.819 Mustafa Raja: Those are just a… Yeah.

330 00:27:14.820 ⇒ 00:27:15.510 Samuel Roberts: Spreadsheet.

331 00:27:16.620 ⇒ 00:27:17.290 Mustafa Raja: Hmm.

332 00:27:17.490 ⇒ 00:27:18.290 Mustafa Raja: Yeah.

333 00:27:18.440 ⇒ 00:27:19.010 Mustafa Raja: Probably.

334 00:27:19.010 ⇒ 00:27:26.710 Samuel Roberts: So… Okay, so, but we can include that as part of the sub-agent for the verbatim responses.

335 00:27:26.710 ⇒ 00:27:29.359 Mustafa Raja: Oh yeah, Andy doesn’t have the templates, sorry.

336 00:27:29.360 ⇒ 00:27:30.000 Samuel Roberts: Okay.

337 00:27:30.580 ⇒ 00:27:31.150 Mustafa Raja: Yeah.

338 00:27:33.300 ⇒ 00:27:34.620 Samuel Roberts: Okay.

339 00:27:34.620 ⇒ 00:27:37.430 Amber Lin: Where’s this section for the verbatim responses?

340 00:27:37.660 ⇒ 00:27:39.639 Samuel Roberts: Well, that’s why I was just about to ask the same thing.

341 00:27:39.640 ⇒ 00:27:42.750 Amber Lin: Oh, I see. I don’t think we have one.

342 00:27:44.350 ⇒ 00:27:45.150 Amber Lin: And here…

343 00:27:45.150 ⇒ 00:27:51.500 Samuel Roberts: Yeah, it’s… Part of the… The migration itself was part of the determination

344 00:27:52.040 ⇒ 00:28:01.830 Samuel Roberts: architecture, and one of the considerations is definitely, like, we don’t need to rag the things, we just need to look up and return for certain things.

345 00:28:02.970 ⇒ 00:28:10.330 Samuel Roberts: So that might be… Maybe a task in the routing and rag migration section?

346 00:28:10.870 ⇒ 00:28:12.339 Samuel Roberts: Would that be a good place for it?

347 00:28:16.510 ⇒ 00:28:21.449 Amber Lin: Cool, yeah, that’s part of it, because part of routing, we do route to the templates.

348 00:28:21.840 ⇒ 00:28:22.440 Samuel Roberts: Yeah.

349 00:28:22.440 ⇒ 00:28:23.100 Amber Lin: I’m blue.

350 00:28:23.100 ⇒ 00:28:28.249 Samuel Roberts: Yeah, it’s just trying to make Andy a little more robust with the responses that, like the… like the…

351 00:28:28.900 ⇒ 00:28:32.059 Samuel Roberts: Mustafa and I were talking about this yesterday, it’s really…

352 00:28:32.200 ⇒ 00:28:38.210 Samuel Roberts: it’s not retrieval augmented generation, really, it’s just retrieval, and so it can kind of bypass the whole rag step.

353 00:28:38.830 ⇒ 00:28:41.080 Samuel Roberts: I can just look up the right sections.

354 00:28:42.350 ⇒ 00:28:43.240 Mustafa Raja: Yeah.

355 00:28:46.360 ⇒ 00:28:48.979 Samuel Roberts: Alright, what else, what else do we have done here?

356 00:28:49.510 ⇒ 00:28:58.660 Mustafa Raja: Can we have CSRs, if they want to… want exact response from Central Doc, ask a certain way?

357 00:29:02.000 ⇒ 00:29:03.109 Samuel Roberts: I don’t…

358 00:29:03.110 ⇒ 00:29:03.660 Mustafa Raja: Oops.

359 00:29:04.250 ⇒ 00:29:05.640 Samuel Roberts: I… I mean, I think…

360 00:29:06.140 ⇒ 00:29:13.009 Samuel Roberts: the AI step here is going to be having to interpret that, given a list of the ones that we know they want.

361 00:29:14.090 ⇒ 00:29:29.940 Amber Lin: Yeah, and I think part of it, of the new features, they want to add, and we want to add, especially this year, is better input validation from Andy asking probing questions based on what it receives to tell them, hey, we only got these in our database, what are you…

362 00:29:29.940 ⇒ 00:29:37.630 Amber Lin: what are you really asking? So, I think that’s the additional step of what we’re gonna build, more than just what we have right now.

363 00:29:38.090 ⇒ 00:29:41.340 Samuel Roberts: Yeah, I think the new… the new agents up is gonna be better at…

364 00:29:41.790 ⇒ 00:29:44.599 Samuel Roberts: Asking questions, like, clarifying questions.

365 00:29:45.600 ⇒ 00:29:47.490 Samuel Roberts: And then potentially follow-ups.

366 00:29:48.440 ⇒ 00:29:59.529 Amber Lin: Cool. I put that at the, second last section of Andy Improvement Goals, because we gotta finish the main migration first before we can do that, but let’s put all their

367 00:30:00.150 ⇒ 00:30:00.730 Amber Lin: want to do there.

368 00:30:00.730 ⇒ 00:30:01.890 Samuel Roberts: Yeah, definitely, definitely.

369 00:30:01.890 ⇒ 00:30:02.470 Uttam Kumaran: Yeah.

370 00:30:03.240 ⇒ 00:30:05.420 Uttam Kumaran: The other thing I’m gonna suggest is, like.

371 00:30:05.760 ⇒ 00:30:11.800 Uttam Kumaran: anything… I would probably… like, my rule of thumb for Gantt is to, like.

372 00:30:12.570 ⇒ 00:30:17.699 Uttam Kumaran: Have everything at least, like, A week, typically.

373 00:30:18.110 ⇒ 00:30:24.260 Uttam Kumaran: I don’t know whether that makes sense in some cases, because some of these, if they are very small, then they can be done in a few days.

374 00:30:24.370 ⇒ 00:30:26.460 Samuel Roberts: But I always…

375 00:30:27.080 ⇒ 00:30:30.219 Uttam Kumaran: Like to just say, like, what’s going on within the week.

376 00:30:31.170 ⇒ 00:30:37.370 Uttam Kumaran: Versus, like, you may not get, like, oh, this is gonna happen on the 12th and the 13th, versus… I see what you’re saying. I know it’s gonna happen, like…

377 00:30:38.200 ⇒ 00:30:42.380 Uttam Kumaran: Yeah, so at a Gantt level, I think about the… I think about weak units of time.

378 00:30:42.910 ⇒ 00:30:46.549 Samuel Roberts: smart, because we can just put multiple things in that week, and then it doesn’t matter…

379 00:30:46.550 ⇒ 00:30:47.090 Uttam Kumaran: names.

380 00:30:47.090 ⇒ 00:30:49.810 Samuel Roberts: As long as we know, basically, the effort required, we can.

381 00:30:49.810 ⇒ 00:31:00.469 Uttam Kumaran: And we… and we are… we’re, like, we’re kind of moving on weak units, right? So, like, thinking about things as, like, 2 days or three days is, like, sort of irrelevant at this… at this level. At the linear.

382 00:31:00.470 ⇒ 00:31:00.870 Samuel Roberts: level.

383 00:31:01.270 ⇒ 00:31:06.419 Uttam Kumaran: we may… we may do that. So, that would be my suggestion, unless something is certainly, like.

384 00:31:06.950 ⇒ 00:31:09.059 Uttam Kumaran: I’ll, like, click a button or something.

385 00:31:09.280 ⇒ 00:31:10.709 Uttam Kumaran: I don’t know, I’m gonna blush.

386 00:31:10.710 ⇒ 00:31:13.660 Samuel Roberts: Yeah, I would… the only other thought is if they’re… if they’re clearly.

387 00:31:13.660 ⇒ 00:31:19.229 Uttam Kumaran: Clarence might have another opinion, so it may be worth asking him what he thinks, but, like, that’s usually what I like to.

388 00:31:19.460 ⇒ 00:31:30.520 Samuel Roberts: No, I like… I like that for… for this. The only thought is that if there are a couple short ones that are dependent on each other, it’s nice to see that flow. And if that’s, like, across a week or two… Yeah, yeah, that’s true, that’s true. …that could be good.

389 00:31:31.000 ⇒ 00:31:34.129 Samuel Roberts: Depending on… on that. But if they’re… if they’re…

390 00:31:34.490 ⇒ 00:31:37.580 Samuel Roberts: Not strictly dependent, it’s less of a worry.

391 00:31:38.310 ⇒ 00:31:38.920 Uttam Kumaran: Yeah.

392 00:31:39.160 ⇒ 00:31:44.480 Uttam Kumaran: Okay, so we have… we have priorities for everything, so maybe let’s also work… let’s… let’s work our way,

393 00:31:44.630 ⇒ 00:31:47.029 Uttam Kumaran: Down again, now that we have that, so…

394 00:31:47.620 ⇒ 00:31:50.009 Uttam Kumaran: We have priorities for everything.

395 00:31:52.890 ⇒ 00:31:59.129 Uttam Kumaran: So, in my lens, like, if we work down from… let me just refresh, I’ll make sure I don’t have the latest.

396 00:31:59.360 ⇒ 00:32:02.780 Uttam Kumaran: And maybe we can also go back to the,

397 00:32:03.190 ⇒ 00:32:08.750 Uttam Kumaran: Or are you guys still working on the priorities for these? Oh, these are fine. These are just maintenance cases.

398 00:32:08.750 ⇒ 00:32:09.920 Samuel Roberts: It’s the maintenance ones don’t need it, yeah.

399 00:32:09.920 ⇒ 00:32:11.580 Casie Aviles: Oh, okay, perfect.

400 00:32:11.790 ⇒ 00:32:13.109 Uttam Kumaran: Yeah, we can go back.

401 00:32:19.460 ⇒ 00:32:27.629 Uttam Kumaran: Cool, so… If we’re looking at this now, let’s start from 7, so acquired BT service account.

402 00:32:27.970 ⇒ 00:32:33.450 Uttam Kumaran: I mean, this one, we need to happen. Establish programmatic access to BQ.

403 00:32:33.710 ⇒ 00:32:36.449 Uttam Kumaran: This one, you know, we need to happen.

404 00:32:36.610 ⇒ 00:32:40.369 Uttam Kumaran: Implement MOSFRA endpoint for main request handling needs to happen.

405 00:32:40.490 ⇒ 00:32:45.359 Uttam Kumaran: I’m gonna extend that to a week. Integrate structure long if you’re master at a data warehouse.

406 00:32:45.580 ⇒ 00:32:47.110 Uttam Kumaran: I’m gonna extend that to a week.

407 00:32:47.330 ⇒ 00:32:52.069 Uttam Kumaran: I’m just gonna… yeah, so then, basically, I’m gonna put,

408 00:32:52.340 ⇒ 00:32:54.890 Uttam Kumaran: I’m gonna change this also, because it doesn’t look like,

409 00:32:55.390 ⇒ 00:33:02.600 Uttam Kumaran: the task titles wrap very well, so I’m just gonna put, master logs to BPU.

410 00:33:04.990 ⇒ 00:33:08.780 Samuel Roberts: The second piece is I validate response correctness.

411 00:33:09.040 ⇒ 00:33:17.009 Uttam Kumaran: So I’m gonna say, Yeah, like, validation of Mastra versus an event.

412 00:33:17.420 ⇒ 00:33:18.180 Samuel Roberts: Yep.

413 00:33:19.530 ⇒ 00:33:21.940 Uttam Kumaran: And I’m gonna extend that to a week.

414 00:33:22.690 ⇒ 00:33:31.079 Uttam Kumaran: Initialize dbt, this, yeah, this one we’re gonna need. Gonna put that as a week.

415 00:33:31.710 ⇒ 00:33:41.140 Uttam Kumaran: Such edge cases and regressions identified during testing. So this is going to be based on the validation of math versus NAM, right?

416 00:33:41.880 ⇒ 00:33:44.280 Samuel Roberts: It’s hard to tell there, but yeah, I think so.

417 00:33:45.600 ⇒ 00:33:48.339 Uttam Kumaran: So, I’m gonna just make sure that this has a…

418 00:33:48.790 ⇒ 00:33:51.399 Uttam Kumaran: Okay, yep, the dependencies are correct here.

419 00:33:52.080 ⇒ 00:33:59.029 Uttam Kumaran: Migrate logging from Snowflake to BigQuery and migrate Rail from pulling the sunset to BigQuery.

420 00:33:59.300 ⇒ 00:34:05.100 Uttam Kumaran: I think both of these are probably, like, a week’s worth of… Work.

421 00:34:05.900 ⇒ 00:34:10.149 Uttam Kumaran: and then, this one is actually…

422 00:34:11.880 ⇒ 00:34:15.470 Uttam Kumaran: Number 15 is actually dependent on,

423 00:34:15.909 ⇒ 00:34:20.270 Uttam Kumaran: not initializing dbt, but on moving from stuff, like, to BQ.

424 00:34:20.719 ⇒ 00:34:26.519 Samuel Roberts: Yeah. Okay, great, and then what does determine the best MASTRA architecture?

425 00:34:27.460 ⇒ 00:34:30.329 Samuel Roberts: We were testing out a few different comparisons to NADM.

426 00:34:30.960 ⇒ 00:34:33.259 Samuel Roberts: The different, like, agents…

427 00:34:33.260 ⇒ 00:34:35.600 Uttam Kumaran: Does that need to happen before 12?

428 00:34:36.500 ⇒ 00:34:40.900 Samuel Roberts: Well, we’re kind of…

429 00:34:41.440 ⇒ 00:34:44.509 Samuel Roberts: That’s… the validation is kind of doing some of this determination.

430 00:34:44.780 ⇒ 00:34:47.550 Samuel Roberts: That makes sense. Like, the testing against… yeah.

431 00:34:48.540 ⇒ 00:34:50.919 Mustafa Raja: It needs to happen before 9.

432 00:34:53.409 ⇒ 00:34:53.839 Samuel Roberts: before.

433 00:34:53.840 ⇒ 00:34:54.819 Uttam Kumaran: before implementing.

434 00:34:54.820 ⇒ 00:34:55.479 Samuel Roberts: Yes, before noon.

435 00:34:55.489 ⇒ 00:34:56.259 Uttam Kumaran: coin?

436 00:34:56.260 ⇒ 00:34:56.810 Samuel Roberts: Yeah.

437 00:34:57.050 ⇒ 00:34:57.790 Mustafa Raja: Yeah.

438 00:35:00.160 ⇒ 00:35:00.590 Samuel Roberts: Honestly, I’m.

439 00:35:00.590 ⇒ 00:35:00.990 Uttam Kumaran: Yeah, that’s…

440 00:35:00.990 ⇒ 00:35:06.730 Samuel Roberts: My master entry endpoint is… is as pressing anyway, while we’re figuring… while we’re still finishing the…

441 00:35:07.120 ⇒ 00:35:07.790 Mustafa Raja: Yeah, yeah, yeah.

442 00:35:07.790 ⇒ 00:35:11.309 Samuel Roberts: of Mazra, so that can bump back a little bit.

443 00:35:11.750 ⇒ 00:35:12.370 Mustafa Raja: Yeah.

444 00:35:12.790 ⇒ 00:35:19.410 Uttam Kumaran: So what… so then what we’re gonna do is we’re gonna move validation of Maestra versus… NNN.

445 00:35:19.740 ⇒ 00:35:24.549 Uttam Kumaran: That’s gonna lead to us determining the best monster architecture, right?

446 00:35:24.550 ⇒ 00:35:25.070 Samuel Roberts: Yeah.

447 00:35:25.520 ⇒ 00:35:27.379 Uttam Kumaran: So 13 moves up.

448 00:35:27.730 ⇒ 00:35:32.649 Uttam Kumaran: And then, implementing the Mosh for endpoint, is actually…

449 00:35:33.310 ⇒ 00:35:36.150 Uttam Kumaran: More determined on determining the best.

450 00:35:36.350 ⇒ 00:35:37.599 Uttam Kumaran: I’m sure. Yeah.

451 00:35:38.130 ⇒ 00:35:38.700 Samuel Roberts: Yes.

452 00:35:38.700 ⇒ 00:35:41.900 Uttam Kumaran: So on determining… so then I’m gonna move this here…

453 00:35:42.230 ⇒ 00:35:47.609 Uttam Kumaran: Okay, great. In the meantime, you can still finish the logs to BQ, initializing dbt.

454 00:35:48.910 ⇒ 00:35:55.940 Uttam Kumaran: Edge cases… So the edge cases and regressions, does that… that happens after the implementation of the Nostra endpoint.

455 00:35:57.480 ⇒ 00:35:58.200 Uttam Kumaran: Right.

456 00:36:00.070 ⇒ 00:36:08.579 Samuel Roberts: Well, we can still do testing in the short term, but there will be others… I mean, that’s gonna be… there’s always gonna be that sort of stuff, so yeah, I would add that afterwards, probably, and… to make that clearer.

457 00:36:09.390 ⇒ 00:36:11.509 Uttam Kumaran: Okay, so I’m just gonna put that’s, like, 2 weeks.

458 00:36:12.110 ⇒ 00:36:13.430 Samuel Roberts: Yes, that’s fine.

459 00:36:13.580 ⇒ 00:36:18.340 Samuel Roberts: Like, we’re kind of doing that as we’re testing out the different architectures, but yeah.

460 00:36:18.340 ⇒ 00:36:19.060 Uttam Kumaran: Yes.

461 00:36:21.730 ⇒ 00:36:22.280 Samuel Roberts: Alright.

462 00:36:22.280 ⇒ 00:36:23.650 Uttam Kumaran: Okay, thanks.

463 00:36:25.160 ⇒ 00:36:32.620 Uttam Kumaran: And then, so there’s, like, this… there is this, like, data warehouse work stream. So now you guys can see how I put them in parallel, right? So…

464 00:36:32.750 ⇒ 00:36:34.809 Uttam Kumaran: Take a look at, like,

465 00:36:35.140 ⇒ 00:36:37.670 Uttam Kumaran: So now, see how we can run these in parallel?

466 00:36:38.490 ⇒ 00:36:39.120 Samuel Roberts: Yeah.

467 00:36:40.410 ⇒ 00:36:44.469 Uttam Kumaran: Because by this point, you’ll have masters set up, and we can move with this, we can move with this.

468 00:36:44.680 ⇒ 00:36:45.650 Samuel Roberts: Right, right.

469 00:36:46.450 ⇒ 00:36:52.329 Uttam Kumaran: So, that’s how we’ll parallelize this. Okay, alright, so take a look at Section 1. How do you guys feel?

470 00:36:52.930 ⇒ 00:36:53.520 Uttam Kumaran: Don’t worry.

471 00:36:53.520 ⇒ 00:36:53.980 Mustafa Raja: They didn’t.

472 00:36:53.980 ⇒ 00:36:54.710 Uttam Kumaran: It’s,

473 00:36:54.990 ⇒ 00:37:02.460 Uttam Kumaran: We’re gonna look through everything at the end, but let’s just… let’s just nail section by section, and then we’ll look at all the sections.

474 00:37:05.690 ⇒ 00:37:06.759 Uttam Kumaran: I’m feeling good.

475 00:37:08.010 ⇒ 00:37:09.330 Uttam Kumaran: Yeah, let’s move on to…

476 00:37:09.610 ⇒ 00:37:12.690 Samuel Roberts: Let’s move on to routing and RAG, so we have…

477 00:37:13.100 ⇒ 00:37:20.349 Uttam Kumaran: Validate document relevance, load testing, So all these, I’m gonna move a little bit to a week.

478 00:37:20.720 ⇒ 00:37:25.269 Uttam Kumaran: For validate document relevance, do we have a plan? Like.

479 00:37:25.380 ⇒ 00:37:28.579 Uttam Kumaran: For this, are we going to, yeah, like.

480 00:37:28.830 ⇒ 00:37:31.850 Uttam Kumaran: I’m kinda curious how we’re gonna do this, but yeah.

481 00:37:33.450 ⇒ 00:37:42.119 Mustafa Raja: Yeah, so, this sort of relates to how many triage tickets we would be getting for central docks.

482 00:37:42.730 ⇒ 00:37:53.549 Mustafa Raja: I think we do not get very much, but the ones that we do get, sometimes related to the way they have structured the central dogs.

483 00:37:53.710 ⇒ 00:37:58.259 Mustafa Raja: Casey, me, and Patricia had a meeting about that.

484 00:37:58.880 ⇒ 00:38:00.069 Mustafa Raja: Last week.

485 00:38:00.070 ⇒ 00:38:06.690 Uttam Kumaran: Well, I guess I’m almost like… I’m almost like, do we need to do… Do you need to, like…

486 00:38:08.980 ⇒ 00:38:12.130 Uttam Kumaran: Set up evals in order to do this.

487 00:38:14.330 ⇒ 00:38:18.549 Mustafa Raja: We do have a pretty good, what’s it called, dataset now.

488 00:38:18.890 ⇒ 00:38:22.269 Mustafa Raja: And we could use that to validate,

489 00:38:22.560 ⇒ 00:38:30.350 Mustafa Raja: this stuff, and we are actually using that to validate, the… the new master architectures.

490 00:38:31.950 ⇒ 00:38:36.980 Samuel Roberts: Yeah, I think… That might be a little ongoing with triage kind of stuff.

491 00:38:36.980 ⇒ 00:38:37.440 Mustafa Raja: Yeah.

492 00:38:37.440 ⇒ 00:38:39.729 Samuel Roberts: But there’s also an element here of making sure that

493 00:38:40.130 ⇒ 00:38:48.639 Samuel Roberts: you know, we have the documents in a good place for them to be able to edit and use, but also for the RAG to use. So, I don’t know if that…

494 00:38:48.840 ⇒ 00:38:49.929 Samuel Roberts: If we want to change this.

495 00:38:49.930 ⇒ 00:38:53.820 Uttam Kumaran: I’m, like, trying to understand, like, how you define accuracy here, like…

496 00:38:54.090 ⇒ 00:38:58.550 Uttam Kumaran: Are you gonna have a list of questions, expected answers, and then, like, just test?

497 00:38:58.790 ⇒ 00:39:00.619 Uttam Kumaran: Whether it’s getting it right, or what.

498 00:39:00.620 ⇒ 00:39:08.840 Mustafa Raja: Yeah, so, they do have… we do have thumbs up, thumbs ups for, central dog questions, right?

499 00:39:09.170 ⇒ 00:39:14.400 Mustafa Raja: And we could always backtest our… Our architectures on that, right?

500 00:39:16.890 ⇒ 00:39:20.129 Uttam Kumaran: Okay. Yeah, I mean, I’m asking you, I’m asking you guys, yeah.

501 00:39:20.370 ⇒ 00:39:23.639 Samuel Roberts: Yeah, yeah, I think that’s a good start, and I think that as we get more…

502 00:39:24.000 ⇒ 00:39:26.599 Samuel Roberts: Thumbs ups and thumbs down, like, with the…

503 00:39:27.820 ⇒ 00:39:31.349 Samuel Roberts: Like, we can add those evals in and just make it more robust, but…

504 00:39:31.950 ⇒ 00:39:33.569 Samuel Roberts: For now, we have a pretty solid set.

505 00:39:35.050 ⇒ 00:39:39.860 Uttam Kumaran: So I’m gonna say, like, establish… evaluation data sets.

506 00:39:40.680 ⇒ 00:39:44.700 Uttam Kumaran: I don’t know if we’ve done that in a while, or… because that’s something that we’re gonna need for the…

507 00:39:44.920 ⇒ 00:39:51.489 Uttam Kumaran: Evals work later anyways, which is, like, a clear set of, like, This question and this answer.

508 00:39:52.590 ⇒ 00:39:53.250 Mustafa Raja: Yeah.

509 00:39:56.540 ⇒ 00:40:04.490 Casie Aviles: You know, we’ll likely need to update that as well, because there are, like, new kinds of questions since when we first created that set.

510 00:40:04.800 ⇒ 00:40:05.340 Samuel Roberts: Yeah.

511 00:40:05.340 ⇒ 00:40:06.070 Uttam Kumaran: Yeah.

512 00:40:11.340 ⇒ 00:40:14.209 Uttam Kumaran: Okay? So we’re established evaluation data set.

513 00:40:14.880 ⇒ 00:40:19.780 Uttam Kumaran: Validate document relevance. That’s probably, like, 2 weeks of work, right?

514 00:40:24.140 ⇒ 00:40:24.980 Samuel Roberts: Yeah.

515 00:40:28.380 ⇒ 00:40:30.770 Uttam Kumaran: This is where I actually want you guys to go slower.

516 00:40:31.450 ⇒ 00:40:32.450 Samuel Roberts: I’m like…

517 00:40:33.050 ⇒ 00:40:35.209 Uttam Kumaran: Nail retrieval, basically.

518 00:40:40.960 ⇒ 00:40:45.599 Uttam Kumaran: So validate document… I mean, dude, yeah, validate document relevance.

519 00:40:47.260 ⇒ 00:40:48.130 Samuel Roberts: Yeah. Okay.

520 00:40:48.180 ⇒ 00:40:52.229 Uttam Kumaran: Load testing, routing, fixed performance or correctness issues.

521 00:40:55.780 ⇒ 00:40:56.230 Samuel Roberts: I think…

522 00:40:56.230 ⇒ 00:40:59.549 Uttam Kumaran: So, the fifth fixed performance and…

523 00:41:00.150 ⇒ 00:41:05.549 Uttam Kumaran: So it’s probably, like, 3 weeks of validate document relevance and retrieval.

524 00:41:06.640 ⇒ 00:41:10.229 Uttam Kumaran: And I actually think this is, like, happens in parallel, like…

525 00:41:12.050 ⇒ 00:41:17.729 Uttam Kumaran: Yeah. Load testing and the fixed performance or correctness issues happen when parents.

526 00:41:18.310 ⇒ 00:41:22.069 Samuel Roberts: Yeah, I mean, validating the fixing is gonna be…

527 00:41:22.070 ⇒ 00:41:28.789 Uttam Kumaran: I guess, like, my point is, like, why… wouldn’t you… aren’t you guys gonna do a lot of the eval work in order to make sure this is right?

528 00:41:29.860 ⇒ 00:41:30.420 Uttam Kumaran: like, how.

529 00:41:30.420 ⇒ 00:41:30.930 Samuel Roberts: Yeah, you’re.

530 00:41:30.930 ⇒ 00:41:33.159 Uttam Kumaran: Basically evaluate the solution.

531 00:41:33.490 ⇒ 00:41:38.279 Samuel Roberts: No, I agree, I agree, that’s what we’re kind of doing already. So we have some, but, you know, I think there’s…

532 00:41:38.890 ⇒ 00:41:41.570 Samuel Roberts: Fixes that can be made to the central dock as well.

533 00:41:41.880 ⇒ 00:41:46.120 Samuel Roberts: That can be… like, I’m not sure if that’s in here, but the…

534 00:41:46.490 ⇒ 00:41:51.760 Samuel Roberts: The evals we have can get bigger from the testing, or the… the thumbs up, thumbs downs that we have.

535 00:41:52.260 ⇒ 00:41:52.950 Samuel Roberts: Hmm.

536 00:41:52.950 ⇒ 00:41:59.749 Uttam Kumaran: I guess my question is, like, as part of validating document relevance, do you have to just start on the eval section anyways?

537 00:42:00.090 ⇒ 00:42:00.950 Mustafa Raja: Yep.

538 00:42:00.950 ⇒ 00:42:03.610 Uttam Kumaran: Which is defining eval metrics, implement links.

539 00:42:03.610 ⇒ 00:42:04.270 Mustafa Raja: Yeah.

540 00:42:04.270 ⇒ 00:42:05.320 Uttam Kumaran: Integrate.

541 00:42:05.600 ⇒ 00:42:06.370 Uttam Kumaran: Okay.

542 00:42:06.630 ⇒ 00:42:08.799 Uttam Kumaran: So then let’s move some of these back.

543 00:42:08.990 ⇒ 00:42:15.979 Casie Aviles: We’ll have to think about whether… because I think we did try before, like, similarity scores.

544 00:42:16.840 ⇒ 00:42:17.360 Uttam Kumaran: Yes.

545 00:42:17.360 ⇒ 00:42:18.040 Casie Aviles: Fair enough.

546 00:42:18.310 ⇒ 00:42:22.159 Casie Aviles: It’d be something that we should look into again, if there’s…

547 00:42:22.160 ⇒ 00:42:27.149 Uttam Kumaran: But basically now, I think the benefit you guys have for Nav and Gabe So you can all, like.

548 00:42:27.780 ⇒ 00:42:28.969 Uttam Kumaran: Everyone could kind of…

549 00:42:30.410 ⇒ 00:42:33.710 Uttam Kumaran: Kind of, like, brainstorm on what the best path is here.

550 00:42:41.190 ⇒ 00:42:49.560 Uttam Kumaran: Yeah, and then I would basically… yeah, so… so I would say this is fine. Implement basic evals, so…

551 00:42:49.960 ⇒ 00:42:54.380 Uttam Kumaran: I don’t know how long the Mastra eval logging stuff is gonna take, so…

552 00:42:55.710 ⇒ 00:43:01.199 Uttam Kumaran: Okay, so let’s go through this one. Let’s go through… so I’m kind of linking both of these sections.

553 00:43:01.360 ⇒ 00:43:07.350 Uttam Kumaran: which is, like, routing and rag migration. I think what I’m gonna change this to is, like,

554 00:43:08.270 ⇒ 00:43:16.380 Uttam Kumaran: I’m gonna say routing, and… Retrieval… accuracy.

555 00:43:16.960 ⇒ 00:43:18.359 Uttam Kumaran: Improvements.

556 00:43:18.470 ⇒ 00:43:19.210 Uttam Kumaran: Right?

557 00:43:19.720 ⇒ 00:43:23.740 Uttam Kumaran: Do you see how I’m, like, I’m trying to label to the client, like, what.

558 00:43:23.740 ⇒ 00:43:24.270 Samuel Roberts: interactions.

559 00:43:24.270 ⇒ 00:43:29.799 Uttam Kumaran: we’re achieving for them. So, think about for Amber, right? I’m trying to tell her that if this gets done.

560 00:43:29.920 ⇒ 00:43:31.530 Uttam Kumaran: We’re gonna have more accurate.

561 00:43:31.860 ⇒ 00:43:34.010 Samuel Roberts: We’re gonna have more accurate retrievals.

562 00:43:34.010 ⇒ 00:43:34.610 Uttam Kumaran: Right.

563 00:43:36.950 ⇒ 00:43:47.230 Uttam Kumaran: You see what I mean? Which is something that is, like, an ongoing issue. So, if we’re able to get both of these done, we should have more HQ. Evals and alerting, what I’m gonna change this into is…

564 00:43:47.640 ⇒ 00:43:55.930 Uttam Kumaran: evaluation… Evaluation… Alerting and monitoring.

565 00:43:56.470 ⇒ 00:43:58.759 Uttam Kumaran: So this is, like, system stability.

566 00:43:59.280 ⇒ 00:43:59.840 Samuel Roberts: Yeah.

567 00:44:05.220 ⇒ 00:44:10.979 Uttam Kumaran: Cool, so we have… so let’s work from the top again, so… Establish the evaluation data set.

568 00:44:11.200 ⇒ 00:44:16.800 Uttam Kumaran: Validate document relevance and retrieval accuracy. Load testing, 6?

569 00:44:17.330 ⇒ 00:44:20.290 Uttam Kumaran: Performance or correctness issues.

570 00:44:20.640 ⇒ 00:44:23.989 Uttam Kumaran: Tell me about migrate templates and migrate cancellations.

571 00:44:26.810 ⇒ 00:44:33.439 Samuel Roberts: This is what we were just talking about, the things that need to be exact, like, verbatim from the central docs.

572 00:44:33.770 ⇒ 00:44:34.939 Samuel Roberts: Rather than…

573 00:44:34.940 ⇒ 00:44:35.460 Uttam Kumaran: Oh, okay.

574 00:44:35.460 ⇒ 00:44:44.609 Samuel Roberts: something taken passed from the central docs and the embeddings passed to the agent and then responded. The idea is that if they ask about a few things, it’ll go to a sub-agent that just…

575 00:44:44.750 ⇒ 00:44:45.700 Samuel Roberts: pulls that.

576 00:44:45.990 ⇒ 00:44:48.290 Samuel Roberts: That stuff, and returns it.

577 00:44:48.830 ⇒ 00:44:54.700 Samuel Roberts: Without… We, you know, these issues with the way it’s chunked, and it’s pulling different pieces, and…

578 00:44:54.990 ⇒ 00:44:55.530 Mustafa Raja: two units.

579 00:44:55.530 ⇒ 00:45:00.430 Samuel Roberts: If they’re looking for verbatim, it’s… as I was saying earlier, it’s not retrieval augmented, it’s just retrieval.

580 00:45:02.000 ⇒ 00:45:02.540 Mustafa Raja: Hmm.

581 00:45:03.530 ⇒ 00:45:05.550 Samuel Roberts: Just trying to… there’s a few things that they want.

582 00:45:05.800 ⇒ 00:45:08.140 Samuel Roberts: Exactly as it is in the central batch.

583 00:45:08.350 ⇒ 00:45:09.010 Mustafa Raja: Yeah.

584 00:45:11.150 ⇒ 00:45:13.590 Samuel Roberts: So where… where’s that here?

585 00:45:14.010 ⇒ 00:45:18.150 Samuel Roberts: Migrate templates, migrate cancellations, yeah, that’s…

586 00:45:21.300 ⇒ 00:45:24.890 Mustafa Raja: Yeah, so I’m working… Working on this this week.

587 00:45:27.540 ⇒ 00:45:30.139 Samuel Roberts: Right, right, that’s what we talked about yesterday.

588 00:45:33.250 ⇒ 00:45:35.360 Samuel Roberts: So that’s probably fine where they are, then.

589 00:45:35.360 ⇒ 00:45:36.180 Mustafa Raja: Yeah, yeah.

590 00:45:46.570 ⇒ 00:45:48.749 Samuel Roberts: So what else do we have, then? .

591 00:45:49.750 ⇒ 00:45:52.399 Uttam Kumaran: I just need to send a message. Okay, so…

592 00:45:53.430 ⇒ 00:45:57.839 Uttam Kumaran: So, sorry, so to migrate templates cancellations, should we leave that, like…

593 00:45:58.260 ⇒ 00:46:03.110 Samuel Roberts: I think they’re good where they are. Yes, it’s kind of… it’s already ongoing a little bit, so…

594 00:46:03.380 ⇒ 00:46:07.110 Samuel Roberts: Okay, okay. Oh, well, so you’ve already started it, so then why don’t we move it back?

595 00:46:07.500 ⇒ 00:46:09.599 Uttam Kumaran: Or why don’t we move the start thing back?

596 00:46:09.600 ⇒ 00:46:10.959 Samuel Roberts: Yeah, yeah, we can do that.

597 00:46:20.060 ⇒ 00:46:24.739 Uttam Kumaran: So, what I’m gonna put here is templates. I’m gonna put, like, fixed copy templates.

598 00:46:24.920 ⇒ 00:46:28.340 Uttam Kumaran: Well, actually, I’ll put, like… Thanks, Copy.

599 00:46:28.920 ⇒ 00:46:33.309 Uttam Kumaran: Or something. I just wanted… I just want it to be really obvious, I’m gonna, because I’m gonna forget, yeah.

600 00:46:33.310 ⇒ 00:46:34.619 Samuel Roberts: Yeah, yeah, definitely.

601 00:46:37.940 ⇒ 00:46:44.640 Uttam Kumaran: Great, okay, so let’s copy of my great templates. Let’s talk about eval. So, define for eval metrics.

602 00:46:45.330 ⇒ 00:46:50.420 Uttam Kumaran: I’m gonna… so that’s next week… implement basic evals using LaneFuse.

603 00:46:50.970 ⇒ 00:46:56.180 Uttam Kumaran: I’m gonna remove basic, and I’m just gonna say implement lane views.

604 00:46:56.560 ⇒ 00:46:57.130 Samuel Roberts: Sure.

605 00:46:59.090 ⇒ 00:47:08.120 Uttam Kumaran: Integrate eval logging into Moctra, so I’m gonna say… Eva, logging… Mostra?

606 00:47:09.350 ⇒ 00:47:10.990 Samuel Roberts: in Austria.

607 00:47:13.610 ⇒ 00:47:14.380 Uttam Kumaran: Okay.

608 00:47:14.510 ⇒ 00:47:17.739 Uttam Kumaran: add error and latency monitoring, so I’m just gonna put…

609 00:47:18.220 ⇒ 00:47:22.759 Uttam Kumaran: Error, latency, monitoring… what is hooks?

610 00:47:24.800 ⇒ 00:47:25.500 Casie Aviles: Hmm.

611 00:47:25.710 ⇒ 00:47:27.910 Uttam Kumaran: Like, if it goes above us for an amount?

612 00:47:31.940 ⇒ 00:47:34.049 Uttam Kumaran: Like, why was the word… what’s the word hoax?

613 00:47:34.780 ⇒ 00:47:39.489 Samuel Roberts: Is it just hooks, like, the code that it needs to do that, so it’s probably just… you can get rid of it.

614 00:47:41.010 ⇒ 00:47:41.760 Samuel Roberts: Error latency.

615 00:47:41.760 ⇒ 00:47:45.359 Uttam Kumaran: Well, you tell me, is it, like… is this, like… I mean, is it.

616 00:47:45.360 ⇒ 00:47:46.160 Casie Aviles: It’s like…

617 00:47:47.050 ⇒ 00:47:47.590 Uttam Kumaran: Yeah.

618 00:47:47.920 ⇒ 00:47:52.320 Casie Aviles: Yeah, I think, I think this is to, like, send, like, alerts.

619 00:47:52.660 ⇒ 00:47:53.290 Mustafa Raja: I’m able to cut.

620 00:47:54.720 ⇒ 00:47:56.230 Casie Aviles: However, there’s an error, or…

621 00:47:56.230 ⇒ 00:47:56.720 Uttam Kumaran: I don’t so much.

622 00:47:56.720 ⇒ 00:47:58.810 Samuel Roberts: Oh, like a, like a web… okay, okay.

623 00:47:59.100 ⇒ 00:48:03.160 Uttam Kumaran: Yeah, so monitoring… Yeah, Monterey Hooks is fine, man.

624 00:48:03.380 ⇒ 00:48:06.080 Uttam Kumaran: Yeah. So, error latency threshold…

625 00:48:09.270 ⇒ 00:48:10.170 Uttam Kumaran: Stop.

626 00:48:11.220 ⇒ 00:48:15.349 Uttam Kumaran: Validate alerts via full-scale integrative. That’s great, I love that.

627 00:48:18.270 ⇒ 00:48:19.510 Uttam Kumaran: Cool.

628 00:48:19.700 ⇒ 00:48:25.320 Uttam Kumaran: So… I’m gonna go ahead and move those back a week.

629 00:48:26.000 ⇒ 00:48:26.650 Uttam Kumaran: Okay.

630 00:48:28.940 ⇒ 00:48:30.539 Uttam Kumaran: How do we feel about this section?

631 00:48:35.980 ⇒ 00:48:37.529 Casie Aviles: Yeah, I think, yeah.

632 00:48:38.040 ⇒ 00:48:38.980 Casie Aviles: This makes sense.

633 00:48:39.780 ⇒ 00:48:43.579 Samuel Roberts: Just making sure everything happening in parallel here next week is still, like.

634 00:48:43.580 ⇒ 00:48:43.980 Casie Aviles: Yeah.

635 00:48:43.980 ⇒ 00:48:47.210 Uttam Kumaran: Yeah, so we’re gonna go, we’re gonna, we’re gonna come… we’re gonna come back, don’t…

636 00:48:47.830 ⇒ 00:48:49.100 Samuel Roberts: Okay, that’s fine. Yeah, in terms of.

637 00:48:49.100 ⇒ 00:48:51.449 Uttam Kumaran: The structure of it looks good, yeah, you’re good.

638 00:48:52.240 ⇒ 00:48:55.250 Uttam Kumaran: Yeah, so, like, let’s say if we were to just take this section.

639 00:48:55.930 ⇒ 00:48:56.330 Samuel Roberts: Yeah.

640 00:48:56.330 ⇒ 00:48:59.929 Uttam Kumaran: this… we could write it like this, right? So don’t think about the other sections just now.

641 00:48:59.930 ⇒ 00:49:01.230 Samuel Roberts: Okay, cool, yeah.

642 00:49:02.270 ⇒ 00:49:03.240 Uttam Kumaran: Okay, cool.

643 00:49:05.260 ⇒ 00:49:10.060 Uttam Kumaran: Great, so let’s move on to the remaining migration.

644 00:49:17.230 ⇒ 00:49:23.149 Uttam Kumaran: So… Validates SQL accuracy. Oh, so this is, like, Texas SQL?

645 00:49:24.610 ⇒ 00:49:27.259 Casie Aviles: Yeah, the SQL that’s being generated.

646 00:49:29.370 ⇒ 00:49:31.210 Samuel Roberts: Yeah. Yeah, Texas people.

647 00:49:31.210 ⇒ 00:49:34.840 Uttam Kumaran: And so… but that, we have to do… we need the evals for that as well, right?

648 00:49:35.600 ⇒ 00:49:39.289 Casie Aviles: Yeah, that should… that would be related to the EVAS as well.

649 00:49:46.980 ⇒ 00:49:50.440 Uttam Kumaran: And then, what is the end-to-end testing of enriched responses?

650 00:49:55.000 ⇒ 00:50:03.469 Casie Aviles: I think that that means being able to, like, test the question… the answers with the returned

651 00:50:04.000 ⇒ 00:50:07.860 Casie Aviles: output of the SQL, or the querying.

652 00:50:08.380 ⇒ 00:50:09.460 Casie Aviles: Engine.

653 00:50:11.300 ⇒ 00:50:13.560 Casie Aviles: But yeah, that’s probably…

654 00:50:15.080 ⇒ 00:50:18.310 Samuel Roberts: Was enriched, like… I guess that is enriched, yeah.

655 00:50:19.140 ⇒ 00:50:23.960 Uttam Kumaran: Well, I would just change this to validate, well, but that’s all part of Validate Texas.

656 00:50:23.960 ⇒ 00:50:31.270 Samuel Roberts: Oh, no, this is the user feedback stuff, so this is, like, enriched responses, meaning the thumbs up, thumbs down, all that stuff that I think is, like.

657 00:50:31.470 ⇒ 00:50:33.840 Samuel Roberts: On top of the response.

658 00:50:34.800 ⇒ 00:50:41.890 Samuel Roberts: Because that’s the other part of this remaining migration. Oh, by the way, Texas SQL user feedback login. So I’m assuming enriched here means…

659 00:50:42.260 ⇒ 00:50:43.469 Samuel Roberts: the… Okay.

660 00:50:43.660 ⇒ 00:50:46.740 Samuel Roberts: the Google Chat side of it, and making sure that that all is…

661 00:50:46.740 ⇒ 00:50:47.310 Uttam Kumaran: Okay.

662 00:50:47.800 ⇒ 00:50:48.480 Samuel Roberts: Fortune.

663 00:50:48.790 ⇒ 00:50:52.359 Uttam Kumaran: But then we have also migrate user feedback, capture, and login.

664 00:50:52.590 ⇒ 00:50:53.510 Uttam Kumaran: It was the same.

665 00:50:53.510 ⇒ 00:50:56.500 Samuel Roberts: Yeah, I think this is just, like, the testing of everything at the end.

666 00:50:58.110 ⇒ 00:51:04.390 Samuel Roberts: Is my thought, maybe I’m wrong here, but the user feedback capture and logging is… Yeah, the testing could.

667 00:51:04.390 ⇒ 00:51:08.719 Uttam Kumaran: Well, you already have implement master endpoint for main request handling.

668 00:51:09.550 ⇒ 00:51:11.769 Uttam Kumaran: On top of that, you’re gonna migrate

669 00:51:11.920 ⇒ 00:51:16.369 Uttam Kumaran: The user feedback, by the way, and then test all the enriched response types.

670 00:51:17.070 ⇒ 00:51:18.090 Uttam Kumaran: Is that right?

671 00:51:18.850 ⇒ 00:51:19.590 Samuel Roberts: I think so.

672 00:51:19.960 ⇒ 00:51:21.390 Samuel Roberts: That’s how I read Enriched.

673 00:51:33.680 ⇒ 00:51:41.210 Uttam Kumaran: And then, okay, so validate text SQL accuracy, migrate, oh, by the way, generation, migrate user feedback, logging.

674 00:51:41.710 ⇒ 00:51:45.219 Uttam Kumaran: Clean up legacy and dependencies for migrated features.

675 00:51:46.790 ⇒ 00:51:49.290 Uttam Kumaran: What is this one?

676 00:51:50.930 ⇒ 00:51:51.790 Samuel Roberts: Which one? I’m sorry?

677 00:51:51.790 ⇒ 00:51:54.680 Uttam Kumaran: This is at the end of everything, so this is 37.

678 00:51:58.010 ⇒ 00:52:00.230 Uttam Kumaran: Or, sorry, can you repeat? Like, I’m gonna…

679 00:52:00.230 ⇒ 00:52:02.520 Samuel Roberts: Yeah, hold on, I’m refreshing mine, yeah.

680 00:52:07.540 ⇒ 00:52:08.920 Uttam Kumaran: Yeah, this is.

681 00:52:11.140 ⇒ 00:52:12.120 Casie Aviles: 37?

682 00:52:13.290 ⇒ 00:52:14.420 Uttam Kumaran: 37, yeah.

683 00:52:17.100 ⇒ 00:52:20.619 Casie Aviles: I think it’s just making sure that we’re offending.

684 00:52:21.620 ⇒ 00:52:22.890 Samuel Roberts: Yeah, so there’s, like.

685 00:52:22.890 ⇒ 00:52:24.600 Uttam Kumaran: Kind of, like, across everything.

686 00:52:25.340 ⇒ 00:52:30.019 Samuel Roberts: Yeah, so this is, like, there’s the main flow on any of them, but then there’s other flows that have been tapped into.

687 00:52:30.660 ⇒ 00:52:32.660 Samuel Roberts: for Andy, and for…

688 00:52:33.000 ⇒ 00:52:40.860 Samuel Roberts: The embedding and, like, there’s a few other things that it does, and so making sure that all that functionality that might not be the core ANDI stuff is…

689 00:52:41.100 ⇒ 00:52:44.900 Samuel Roberts: But I guess, can we… why don’t we just list out all of that functionality.

690 00:52:45.150 ⇒ 00:52:51.380 Uttam Kumaran: fear, I mean, basically, I’m gonna change this to a milestone, which is, like, N8M deprecation.

691 00:52:53.020 ⇒ 00:52:55.709 Uttam Kumaran: Is there other functionality that isn’t listed here?

692 00:53:00.600 ⇒ 00:53:04.479 Samuel Roberts: I… I’m not sure specifically if there’s, like, other… like, I don’t know what all the flows…

693 00:53:05.070 ⇒ 00:53:05.870 Casie Aviles: Yeah.

694 00:53:05.980 ⇒ 00:53:10.889 Casie Aviles: So there’s the whole AND piece, that’s one workflow.

695 00:53:12.280 ⇒ 00:53:17.720 Casie Aviles: The, the oh-by-the-way generation and feedback capture and logging are…

696 00:53:17.720 ⇒ 00:53:18.150 Samuel Roberts: Okay.

697 00:53:18.150 ⇒ 00:53:19.190 Casie Aviles: Brit ones.

698 00:53:20.750 ⇒ 00:53:22.000 Casie Aviles: So… Okay.

699 00:53:23.290 ⇒ 00:53:28.309 Casie Aviles: Then, there’s also the forms, which are all in NH10.

700 00:53:29.630 ⇒ 00:53:31.590 Samuel Roberts: Sure, that’s already up there, too.

701 00:53:33.240 ⇒ 00:53:36.100 Casie Aviles: So the… Those are, like, the entire…

702 00:53:37.350 ⇒ 00:53:39.629 Casie Aviles: workflows that we have on LinkedIn. Cool.

703 00:53:41.380 ⇒ 00:53:41.910 Uttam Kumaran: Okay.

704 00:53:42.960 ⇒ 00:53:44.340 Uttam Kumaran: And then.

705 00:53:44.340 ⇒ 00:53:44.565 Samuel Roberts: Hmm.

706 00:53:44.790 ⇒ 00:53:47.619 Uttam Kumaran: Now let’s move to… let’s move to client GCP.

707 00:53:47.730 ⇒ 00:53:53.289 Uttam Kumaran: and retire NAM, so… Configure Cloud Run Service Account Secrets.

708 00:53:54.290 ⇒ 00:53:55.490 Uttam Kumaran: We’re in progress.

709 00:53:56.750 ⇒ 00:53:58.239 Uttam Kumaran: Are all these in progress?

710 00:53:59.940 ⇒ 00:54:02.580 Casie Aviles: Let’s see…

711 00:54:03.110 ⇒ 00:54:05.020 Samuel Roberts: Oh, the embedding pipeline is the other thing, right?

712 00:54:05.870 ⇒ 00:54:12.299 Casie Aviles: Yeah, so… I would say, like, at least configure Cloud Run.

713 00:54:12.480 ⇒ 00:54:14.929 Casie Aviles: Was some… was in… is in progress.

714 00:54:16.260 ⇒ 00:54:21.509 Casie Aviles: Other ones are not yet… like, I haven’t started these ones yet.

715 00:54:22.290 ⇒ 00:54:25.569 Samuel Roberts: Yeah, I think we can bump those to a few weeks down, even.

716 00:54:25.930 ⇒ 00:54:28.939 Uttam Kumaran: Do you think we’ll start any of these, any of these other ones this week?

717 00:54:31.050 ⇒ 00:54:32.190 Samuel Roberts: Hmm…

718 00:54:34.650 ⇒ 00:54:40.319 Samuel Roberts: I mean, we don’t want to update the Google Chat till it’s ready to go. Like, that’s the… that’s the switchover.

719 00:54:41.950 ⇒ 00:54:42.340 Uttam Kumaran: Oh.

720 00:54:42.340 ⇒ 00:54:42.830 Samuel Roberts: The West Coast.

721 00:54:42.830 ⇒ 00:54:44.360 Uttam Kumaran: I’ll be, like, really at the end.

722 00:54:44.360 ⇒ 00:54:47.810 Samuel Roberts: Yeah, or at least much later, yeah, like…

723 00:54:48.150 ⇒ 00:54:52.279 Samuel Roberts: We’re gonna want all the other features there. Full production validation could happen

724 00:54:52.400 ⇒ 00:54:56.760 Samuel Roberts: Well, not full production validation until we switch that, but… We could have a dev.

725 00:54:56.970 ⇒ 00:54:57.870 Samuel Roberts: Chat.

726 00:54:58.670 ⇒ 00:54:59.840 Samuel Roberts: Test that.

727 00:55:00.100 ⇒ 00:55:03.619 Samuel Roberts: And then deploy to GCP, yeah. That’s probably later, too.

728 00:55:05.470 ⇒ 00:55:07.080 Uttam Kumaran: Did we end up getting a job?

729 00:55:10.810 ⇒ 00:55:11.790 Casie Aviles: What was that?

730 00:55:12.990 ⇒ 00:55:15.690 Uttam Kumaran: Did we end up getting, like, a dev environment for chat?

731 00:55:17.010 ⇒ 00:55:23.809 Casie Aviles: Oh, yeah, we do have, like, separate… Environments are broad, yeah.

732 00:55:24.140 ⇒ 00:55:25.590 Samuel Roberts: Yeah, we haven’t needed to use it yet.

733 00:55:25.590 ⇒ 00:55:27.049 Uttam Kumaran: You could set up, like, a fake AMD.

734 00:55:27.510 ⇒ 00:55:28.439 Samuel Roberts: Yeah, that’s the other thing I’m.

735 00:55:28.440 ⇒ 00:55:30.440 Uttam Kumaran: We can get it there for…

736 00:55:30.620 ⇒ 00:55:31.260 Samuel Roberts: Yeah.

737 00:55:31.260 ⇒ 00:55:35.430 Uttam Kumaran: Yeah, established, like… Andy Saging.

738 00:55:35.630 ⇒ 00:55:36.620 Uttam Kumaran: Basically.

739 00:55:37.580 ⇒ 00:55:40.520 Uttam Kumaran: So, do we have 3, or we only have 2 now?

740 00:55:43.160 ⇒ 00:55:45.139 Samuel Roberts: To, like, environments?

741 00:55:45.900 ⇒ 00:55:46.520 Uttam Kumaran: Yeah.

742 00:55:46.780 ⇒ 00:55:48.449 Casie Aviles: Yeah, we have 3 implants.

743 00:55:49.500 ⇒ 00:55:50.060 Uttam Kumaran: Okay.

744 00:55:54.680 ⇒ 00:55:56.439 Samuel Roberts: So, Deb Staging and prom?

745 00:55:57.510 ⇒ 00:55:58.590 Uttam Kumaran: Yeah.

746 00:56:00.500 ⇒ 00:56:05.920 Samuel Roberts: Okay. But yeah, we can probably even switch over to Dev at some point soon, because the monster is…

747 00:56:06.180 ⇒ 00:56:09.910 Samuel Roberts: you know, studio lets you chat.

748 00:56:10.030 ⇒ 00:56:12.340 Samuel Roberts: So we didn’t have to stand anything up for that yet.

749 00:56:12.560 ⇒ 00:56:14.220 Samuel Roberts: But we could probably get that.

750 00:56:15.970 ⇒ 00:56:21.929 Samuel Roberts: That way we have it, like, see how bad it is to switch over and everything, and test it end-to-end, too.

751 00:56:22.230 ⇒ 00:56:28.720 Uttam Kumaran: Well, I want to give every… I want to give this whole team and client some access to… anti-staging.

752 00:56:30.050 ⇒ 00:56:37.809 Uttam Kumaran: Right? So, I’m gonna say, so provide… Andy, staging access to team…

753 00:56:38.610 ⇒ 00:56:41.139 Uttam Kumaran: And this is gonna be, like, the milestone.

754 00:56:49.360 ⇒ 00:56:54.610 Uttam Kumaran: So then, okay, so we have Configure Cloud Run, set up Cloud SQL,

755 00:56:55.020 ⇒ 00:56:59.489 Uttam Kumaran: Ultras and PGVector. Let’s plug in if they can build on DevChat.

756 00:56:59.900 ⇒ 00:57:05.790 Uttam Kumaran: validation, so we need to establish ANDI development, Do the dev chat.

757 00:57:07.440 ⇒ 00:57:08.710 Uttam Kumaran: validation.

758 00:57:09.780 ⇒ 00:57:13.300 Uttam Kumaran: Update Google Chat webhook, so this is gonna be all the way at the end.

759 00:57:13.590 ⇒ 00:57:17.509 Samuel Roberts: Run full production, and spoke to us as an engineer.

760 00:57:18.160 ⇒ 00:57:19.510 Uttam Kumaran: Something like that.

761 00:57:20.570 ⇒ 00:57:24.590 Uttam Kumaran: Establish anti-staging, I’m gonna need, like.

762 00:57:25.360 ⇒ 00:57:29.069 Uttam Kumaran: Deploy a master service to client, it’s a P and Blackline.

763 00:57:29.070 ⇒ 00:57:34.099 Samuel Roberts: Yeah, I was gonna say, update, Google Chat webhook deployment is basically setting it up in Prod.

764 00:57:34.220 ⇒ 00:57:35.569 Samuel Roberts: So, you might want to rewrite.

765 00:57:35.570 ⇒ 00:57:36.160 Uttam Kumaran: Yeah.

766 00:57:36.160 ⇒ 00:57:36.960 Samuel Roberts: naming that.

767 00:57:36.960 ⇒ 00:57:37.840 Uttam Kumaran: Oh, okay.

768 00:57:39.260 ⇒ 00:57:44.149 Uttam Kumaran: Well, that’s gonna be, like… is that just, like, a flick of… that’s basically, like, a flick of a switch, right?

769 00:57:44.300 ⇒ 00:57:49.900 Samuel Roberts: Well, I mean, once we get the staging set up, we’ll know if it’s… if it’s a one-to-one switch or not, you know?

770 00:57:50.060 ⇒ 00:57:51.149 Uttam Kumaran: Okay, that’s all.

771 00:57:54.940 ⇒ 00:58:04.990 Samuel Roberts: Yeah, I mean, we could map it so that it’s hopefully just the same as the current one, but we’re also… we were limited by N8M when we first did that, so it might not be the best model to fit.

772 00:58:18.960 ⇒ 00:58:19.670 Samuel Roberts: Excuse me.

773 00:58:21.810 ⇒ 00:58:22.450 Uttam Kumaran: Okay.

774 00:58:25.170 ⇒ 00:58:30.300 Uttam Kumaran: So this is, like, okay, let’s look at this section, how do we feel about… Nice.

775 00:58:30.300 ⇒ 00:58:33.970 Samuel Roberts: Let’s set up Cloud SQL.

776 00:58:34.670 ⇒ 00:58:41.480 Samuel Roberts: Set up Cloud SQL, We… is that just…

777 00:58:42.010 ⇒ 00:58:45.660 Samuel Roberts: Like, are we talking… I mean, it looks like a bunch of it, so that’s, like, moving…

778 00:58:46.030 ⇒ 00:58:47.190 Samuel Roberts: Everything over, too.

779 00:58:47.190 ⇒ 00:58:54.550 Casie Aviles: This is… yeah, this is going to be moving Superbase into the Google environment.

780 00:58:54.550 ⇒ 00:58:58.570 Samuel Roberts: Okay, I’m just making sure setup meant not just, you know, init kind of thing.

781 00:58:59.570 ⇒ 00:59:01.580 Samuel Roberts: I may want to add a note there or something.

782 00:59:11.470 ⇒ 00:59:12.700 Uttam Kumaran: Yeah, conclude.

783 00:59:17.660 ⇒ 00:59:18.609 Casie Aviles: you know, discussion.

784 00:59:18.610 ⇒ 00:59:21.980 Uttam Kumaran: And then, anything on… anything on fixes?

785 00:59:23.930 ⇒ 00:59:27.259 Uttam Kumaran: I mean, I’m not gonna look at the new features just yet, but let’s…

786 00:59:27.260 ⇒ 00:59:27.650 Samuel Roberts: Yeah.

787 00:59:27.650 ⇒ 00:59:28.690 Uttam Kumaran: prefixes.

788 00:59:34.000 ⇒ 00:59:36.330 Samuel Roberts: Yeah, Casey, how are you feeling about the admin UI?

789 00:59:36.790 ⇒ 00:59:38.949 Casie Aviles: Yeah, this, this one’s fine.

790 00:59:39.290 ⇒ 00:59:42.120 Casie Aviles: Or within this week, yeah, I’m…

791 00:59:42.570 ⇒ 00:59:44.900 Casie Aviles: Okay. I’m going to close this out this week.

792 00:59:48.250 ⇒ 00:59:51.910 Casie Aviles: And then these other ones, yeah, these are done standardized service.

793 00:59:52.880 ⇒ 00:59:57.149 Samuel Roberts: That’ll be… that’ll happen with the admin UI being done, right?

794 00:59:57.410 ⇒ 00:59:58.229 Casie Aviles: Yeah, yeah.

795 00:59:58.400 ⇒ 00:59:59.130 Casie Aviles: Wow.

796 01:00:00.300 ⇒ 01:00:03.520 Casie Aviles: I’m not sure about this central dog structure.

797 01:00:03.520 ⇒ 01:00:10.429 Amber Lin: I think the last one is sort of like the earlier one we have, validate the document.

798 01:00:10.430 ⇒ 01:00:14.419 Uttam Kumaran: Yeah, so I’m gonna probably remove this, and it’s gonna get covered there.

799 01:00:14.700 ⇒ 01:00:24.890 Samuel Roberts: Yeah, I think so. The only thing I’ll add is that we want to make sure to loop them in, or get their input as needed, or make, you know… if we’re changing the central dock, it’s the two-way thing.

800 01:00:26.600 ⇒ 01:00:27.390 Samuel Roberts: I don’t know if this was.

801 01:00:27.390 ⇒ 01:00:33.509 Uttam Kumaran: Yeah, so what I’m… what I’m gonna do is I’m gonna change this to confirm… In a central dock.

802 01:00:33.910 ⇒ 01:00:35.649 Uttam Kumaran: structure your client.

803 01:00:36.140 ⇒ 01:00:37.910 Uttam Kumaran: And I’m gonna move this up.

804 01:00:39.780 ⇒ 01:00:40.440 Samuel Roberts: Yeah.

805 01:00:40.440 ⇒ 01:00:46.110 Uttam Kumaran: I’m gonna move this up to… .

806 01:00:56.790 ⇒ 01:01:00.769 Amber Lin: Do we just call the fixes CIPCO-related issues?

807 01:01:06.050 ⇒ 01:01:08.699 Uttam Kumaran: Yeah, maybe we can just say, like, zip.

808 01:01:09.660 ⇒ 01:01:10.969 Amber Lin: Yeah, zip code.

809 01:01:11.910 ⇒ 01:01:15.880 Uttam Kumaran: I’m gonna say zip code, management features.

810 01:01:23.390 ⇒ 01:01:27.179 Samuel Roberts: Yeah, the only other one was in bed. Okay. But… that’s done.

811 01:01:29.170 ⇒ 01:01:37.600 Uttam Kumaran: So now… If we look at this…

812 01:01:38.160 ⇒ 01:01:45.440 Uttam Kumaran: And my suggestion, Casey, is to just zoom out, but don’t use the Zoom thing, just zoom out on your browser.

813 01:01:49.280 ⇒ 01:01:57.880 Uttam Kumaran: And yeah, so now, or whatever’s comfortable, but now we’re gonna… we can look at across all of the four sections.

814 01:01:59.000 ⇒ 01:02:02.410 Uttam Kumaran: Right, so…

815 01:02:06.020 ⇒ 01:02:07.919 Uttam Kumaran: My question is…

816 01:02:08.310 ⇒ 01:02:16.099 Uttam Kumaran: How… my question is, how many tasks can we work on in parallel? And that’s gonna allow us to kind of figure out, like, what this actually looks like.

817 01:02:21.690 ⇒ 01:02:25.630 Samuel Roberts: One thing Clarence mentioned earlier is, color coding these things.

818 01:02:26.240 ⇒ 01:02:31.479 Samuel Roberts: Which could be helpful for, like, visualizing, you know, who’s on what.

819 01:02:33.880 ⇒ 01:02:39.599 Samuel Roberts: So it might be a good way to see, like, have we overloaded someone with a certain number of things in a week or something?

820 01:02:40.520 ⇒ 01:02:45.040 Uttam Kumaran: Yeah, I guess low… Let’s assume… let’s just assume, like, we have two people.

821 01:02:45.590 ⇒ 01:02:46.200 Samuel Roberts: Okay.

822 01:02:46.690 ⇒ 01:02:47.320 Uttam Kumaran: like…

823 01:02:47.410 ⇒ 01:02:49.960 Samuel Roberts: I kinda just wanna know what our…

824 01:02:50.100 ⇒ 01:02:53.380 Uttam Kumaran: What our team’s collective max capacity is.

825 01:02:54.000 ⇒ 01:02:54.420 Samuel Roberts: Okay.

826 01:02:54.420 ⇒ 01:02:56.780 Uttam Kumaran: Like, so looking at this, I feel like…

827 01:02:57.420 ⇒ 01:02:59.670 Samuel Roberts: Migrate ANDILogic.

828 01:03:03.490 ⇒ 01:03:06.830 Uttam Kumaran: like, for example, migrate main ANDI logic

829 01:03:09.100 ⇒ 01:03:13.760 Uttam Kumaran: How much of the evaluation can we do Before that’s done.

830 01:03:16.300 ⇒ 01:03:17.110 Uttam Kumaran: Right?

831 01:03:18.180 ⇒ 01:03:24.850 Uttam Kumaran: And I’m actually gonna move the routing thing to the top section, because… so that there’s just, like, it’s a little bit divided. Yeah.

832 01:03:24.850 ⇒ 01:03:25.850 Samuel Roberts: Yeah. Like, how…

833 01:03:26.280 ⇒ 01:03:30.640 Uttam Kumaran: What, what, what from, what from Section 2 is dependent on Section 1?

834 01:03:32.730 ⇒ 01:03:41.380 Samuel Roberts: I would say the dataset is not the… Load… Load test routing is…

835 01:03:42.430 ⇒ 01:03:46.610 Samuel Roberts: Fixed performance or correctness issues is…

836 01:03:47.990 ⇒ 01:03:54.139 Samuel Roberts: And, oh, where am I looking? Define the core valuation metrics is… is not, so you could bump that up.

837 01:03:54.970 ⇒ 01:04:03.030 Samuel Roberts: Implement length views, eval logging, and MOSTRA, that one…

838 01:04:05.220 ⇒ 01:04:08.850 Samuel Roberts: We could probably do that,

839 01:04:09.930 ⇒ 01:04:15.340 Samuel Roberts: Until… well, no, I think the eval logging in Mosra is kind of dependent on having the right architecture done.

840 01:04:21.370 ⇒ 01:04:22.760 Samuel Roberts: Define quarter, yeah.

841 01:04:23.440 ⇒ 01:04:24.720 Samuel Roberts: Excuse me.

842 01:04:26.560 ⇒ 01:04:32.180 Casie Aviles: Yeah, it’s… it’s a bit dependent on… Yeah.

843 01:04:32.180 ⇒ 01:04:32.820 Samuel Roberts: The evaluation…

844 01:04:32.820 ⇒ 01:04:34.460 Casie Aviles: Architecture.

845 01:04:34.750 ⇒ 01:04:36.590 Samuel Roberts: Yeah, so we could probably move up

846 01:04:37.420 ⇒ 01:04:41.600 Samuel Roberts: What is it? 29 is not dependent on being finished.

847 01:04:41.900 ⇒ 01:04:43.170 Samuel Roberts: the…

848 01:04:43.620 ⇒ 01:04:47.309 Samuel Roberts: what are we looking at here? I hope I’m saying the right ones on your screen, too, but…

849 01:04:48.530 ⇒ 01:04:53.850 Samuel Roberts: 30, 31… Need to kind of happen after.

850 01:04:55.670 ⇒ 01:04:59.180 Samuel Roberts: At least, the end of…

851 01:04:59.420 ⇒ 01:05:03.339 Samuel Roberts: the… the German Best Mastra architecture.

852 01:05:04.320 ⇒ 01:05:07.639 Samuel Roberts: End of next week, we can bump those back then.

853 01:05:13.290 ⇒ 01:05:14.420 Samuel Roberts: Does that sound good?

854 01:05:16.200 ⇒ 01:05:22.100 Samuel Roberts: But the metrics can be bumped, where… yeah, metrics can stay there, implement length use can be bump to here.

855 01:05:23.070 ⇒ 01:05:24.720 Samuel Roberts: Logging can bump to heat.

856 01:05:24.720 ⇒ 01:05:26.450 Uttam Kumaran: It’s from the right, Casey?

857 01:05:28.140 ⇒ 01:05:28.840 Casie Aviles: Okay.

858 01:05:29.260 ⇒ 01:05:35.639 Samuel Roberts: So I just, lid, or ice cream.

859 01:05:40.540 ⇒ 01:05:44.849 Samuel Roberts: 30… This slid these two, or not these two here.

860 01:05:47.180 ⇒ 01:05:51.169 Samuel Roberts: Because that’s kind of dependent on this milestone.

861 01:05:54.270 ⇒ 01:05:56.379 Samuel Roberts: So as we can even add that dependency, right?

862 01:05:56.760 ⇒ 01:05:57.780 Casie Aviles: Yeah.

863 01:05:59.490 ⇒ 01:06:00.380 Casie Aviles: This one.

864 01:06:04.530 ⇒ 01:06:05.095 Samuel Roberts: Oh.

865 01:06:19.250 ⇒ 01:06:20.930 Samuel Roberts: Zoomed out too much.

866 01:06:28.770 ⇒ 01:06:32.759 Uttam Kumaran: Yeah, the main thing I’m trying to understand is what from routing and retrieval

867 01:06:34.530 ⇒ 01:06:37.680 Uttam Kumaran: needs to wait for anything from Section 1.

868 01:06:38.190 ⇒ 01:06:44.430 Samuel Roberts: Yeah, I was… yeah, I’m trying to see, like, what’s dependent on, you know, 11 determined best monster architecture.

869 01:06:49.830 ⇒ 01:06:51.929 Samuel Roberts: What do we add dependency is?

870 01:06:51.930 ⇒ 01:06:56.999 Uttam Kumaran: Because for me, it also seems like we should just crush the GCP stuff, like, early.

871 01:07:03.410 ⇒ 01:07:07.260 Uttam Kumaran: We can’t really do anything until the GCP stuff is, like, set up, right?

872 01:07:09.970 ⇒ 01:07:17.449 Samuel Roberts: Can’t do anything… well… Yeah, I mean, it’s not necessarily dependent on the best monster architecture and settling that.

873 01:07:18.210 ⇒ 01:07:21.729 Samuel Roberts: at least the… set up the class QL,

874 01:07:22.090 ⇒ 01:07:29.459 Samuel Roberts: We could set up the dev environment to at least point to something, whether or not, you know, and then we can bounce that around.

875 01:07:30.200 ⇒ 01:07:32.470 Samuel Roberts: If we change the architecture or not.

876 01:07:36.160 ⇒ 01:07:45.130 Samuel Roberts: Yeah, I think that could be happening. Establish anti-staging, that’s fine too. Where does this one? Provide anti-staging access to the team, we might want to bump that back a little bit.

877 01:07:45.650 ⇒ 01:07:47.549 Samuel Roberts: Until we have the architecture settled.

878 01:07:48.340 ⇒ 01:08:00.310 Samuel Roberts: Yeah, I think I’m gonna slide that back to… Here… Because…

879 01:08:01.390 ⇒ 01:08:04.150 Samuel Roberts: Any staging could even happen there.

880 01:08:05.090 ⇒ 01:08:06.889 Samuel Roberts: I think that’s fine, actually, you’re right.

881 01:08:10.560 ⇒ 01:08:16.750 Casie Aviles: This is final. There are… I don’t see any more dependents from the first section.

882 01:08:17.090 ⇒ 01:08:20.260 Uttam Kumaran: So now, so purely right now, we’re just gonna work on, like.

883 01:08:20.399 ⇒ 01:08:22.569 Uttam Kumaran: How much we can take on concurrently, right?

884 01:08:23.540 ⇒ 01:08:24.520 Samuel Roberts: Yeah.

885 01:08:24.520 ⇒ 01:08:29.469 Uttam Kumaran: So… I mean, it’s… for me, it seems like…

886 01:08:31.870 ⇒ 01:08:35.629 Uttam Kumaran: Bulk of the work is going to be on…

887 01:08:51.430 ⇒ 01:08:53.140 Samuel Roberts: I don’t want to create a picture.

888 01:08:53.149 ⇒ 01:08:59.249 Uttam Kumaran: If I was to ask a question, would you rather start on the retrieval accuracy or finish up GCPs?

889 01:09:00.299 ⇒ 01:09:02.029 Uttam Kumaran: Set up, like, what would you say?

890 01:09:05.220 ⇒ 01:09:08.689 Samuel Roberts: My thought would be retrieval accuracy, because we’re kind of…

891 01:09:09.189 ⇒ 01:09:11.599 Samuel Roberts: Doing some of that with the…

892 01:09:11.609 ⇒ 01:09:12.789 Uttam Kumaran: I guess to put another way.

893 01:09:13.200 ⇒ 01:09:15.949 Uttam Kumaran: Any findings on retrieval accuracy help us?

894 01:09:16.439 ⇒ 01:09:18.389 Uttam Kumaran: Even in the current NADM setup?

895 01:09:24.310 ⇒ 01:09:29.890 Casie Aviles: Yeah, I think… It should help us with, you know, addressing some of the…

896 01:09:30.600 ⇒ 01:09:34.359 Casie Aviles: The concerns that they’ve reported to us.

897 01:09:36.290 ⇒ 01:09:41.500 Uttam Kumaran: Okay, then what I’m gonna do is I’m gonna say we kick GCP Till the end.

898 01:09:42.710 ⇒ 01:09:46.059 Samuel Roberts: Yeah, I would at least save maybe the…

899 01:09:46.830 ⇒ 01:09:51.080 Samuel Roberts: Service accounts and secrets is something we still want, so that we’re not dependent on them.

900 01:09:54.850 ⇒ 01:09:57.290 Samuel Roberts: But everything else can probably be slid back.

901 01:09:58.170 ⇒ 01:09:58.720 Uttam Kumaran: Gotcha.

902 01:10:02.600 ⇒ 01:10:11.409 Uttam Kumaran: So I’m gonna move… the GCP stuff until the end of January.

903 01:10:13.260 ⇒ 01:10:17.039 Uttam Kumaran: Do you want to go ahead and finish up the Cloud Run thing anyways, or no?

904 01:10:18.550 ⇒ 01:10:20.410 Samuel Roberts: Yeah, I’d like to…

905 01:10:20.410 ⇒ 01:10:22.430 Uttam Kumaran: I haven’t done any of it, then I would… okay.

906 01:10:22.590 ⇒ 01:10:26.199 Samuel Roberts: No, I mean, we have… yeah, I would say it’s not… yeah, it’s probably fine to push back, too.

907 01:10:30.110 ⇒ 01:10:33.939 Samuel Roberts: The other stuff is, like, much more… critical, I think, to…

908 01:10:34.470 ⇒ 01:10:37.059 Samuel Roberts: A lot of the other things. That’s fine.

909 01:10:37.260 ⇒ 01:10:37.860 Uttam Kumaran: Okay.

910 01:10:37.970 ⇒ 01:10:39.190 Uttam Kumaran: So…

911 01:10:48.680 ⇒ 01:10:51.989 Uttam Kumaran: The snowflake, the BQ stuff is not important.

912 01:10:52.500 ⇒ 01:10:54.960 Samuel Roberts: Oh, there we go, that’s how this works. Okay.

913 01:10:58.820 ⇒ 01:11:01.009 Uttam Kumaran: Oh, I’m gonna kick initial as dbt.

914 01:11:06.920 ⇒ 01:11:07.740 Uttam Kumaran: Because I don’t think.

915 01:11:07.740 ⇒ 01:11:08.070 Samuel Roberts: Okay.

916 01:11:08.070 ⇒ 01:11:08.580 Uttam Kumaran: that important.

917 01:11:09.560 ⇒ 01:11:12.299 Samuel Roberts: Which one, sorry.

918 01:11:12.630 ⇒ 01:11:13.480 Samuel Roberts: dbt.

919 01:11:13.480 ⇒ 01:11:14.290 Uttam Kumaran: It was 15.

920 01:11:16.850 ⇒ 01:11:18.329 Samuel Roberts: Yeah, that’s fine.

921 01:11:20.940 ⇒ 01:11:26.970 Samuel Roberts: Because, Snowflake… I mean, Snowflake is working for us, so, you know, for the evals we need to do, we have all the data there.

922 01:11:27.860 ⇒ 01:11:28.720 Uttam Kumaran: And…

923 01:11:38.210 ⇒ 01:11:41.980 Uttam Kumaran: Okay, so take a look at this, what do you think about this? The way it’s out?

924 01:11:44.700 ⇒ 01:11:47.420 Samuel Roberts: Forgot.

925 01:11:56.040 ⇒ 01:11:57.660 Samuel Roberts: This looks pretty good, I think.

926 01:11:59.800 ⇒ 01:12:01.420 Casie Aviles: I think this is still there.

927 01:12:01.750 ⇒ 01:12:05.969 Casie Aviles: So we will be doing… zip code.

928 01:12:05.970 ⇒ 01:12:06.420 Samuel Roberts: Oh, yeah.

929 01:12:06.730 ⇒ 01:12:10.810 Casie Aviles: And then… Rag… okay.

930 01:12:11.000 ⇒ 01:12:12.020 Casie Aviles: Evos.

931 01:12:12.020 ⇒ 01:12:15.710 Uttam Kumaran: So, go ahead and go to the Overview tab on the top left.

932 01:12:16.080 ⇒ 01:12:18.339 Uttam Kumaran: It’ll give you a little bit of, like, a…

933 01:12:18.890 ⇒ 01:12:19.850 Samuel Roberts: Oh, whoa.

934 01:12:19.850 ⇒ 01:12:21.519 Uttam Kumaran: Higher level viewed sections.

935 01:12:23.350 ⇒ 01:12:29.200 Uttam Kumaran: So we’re going… we’re gonna do… migrate main AND Logic.

936 01:12:29.330 ⇒ 01:12:38.230 Uttam Kumaran: and logging to the Astra. So I… maybe what we do is we split out, like, the logging.

937 01:12:38.570 ⇒ 01:12:39.899 Samuel Roberts: Logging can be split, yeah.

938 01:12:41.980 ⇒ 01:12:44.049 Uttam Kumaran: I don’t even… smiling.

939 01:12:44.050 ⇒ 01:12:47.999 Samuel Roberts: Drop a section somewhere, right? I think you can just drag the sections around.

940 01:12:58.970 ⇒ 01:13:01.570 Uttam Kumaran: Okay, and then Fox will do that.

941 01:13:04.480 ⇒ 01:13:08.670 Uttam Kumaran: Oh, okay, so I can just do, like… login version.

942 01:13:10.050 ⇒ 01:13:11.460 Samuel Roberts: Yeah, and then drag that.

943 01:13:12.100 ⇒ 01:13:14.129 Samuel Roberts: Or to start moving things into it. Yeah.

944 01:13:14.400 ⇒ 01:13:17.299 Samuel Roberts: Because the… what do we need to stay there?

945 01:13:18.370 ⇒ 01:13:22.329 Samuel Roberts: Implement monster entry point, determine best MASTRA architecture, validate Mastra.

946 01:13:23.060 ⇒ 01:13:26.099 Samuel Roberts: programmatic access to BQ, that’s logging…

947 01:13:27.140 ⇒ 01:13:30.310 Samuel Roberts: required VQ service account that’s logging. Yeah, so those are split now.

948 01:13:31.900 ⇒ 01:13:33.510 Samuel Roberts: So, if we go to Overview…

949 01:13:33.510 ⇒ 01:13:37.560 Uttam Kumaran: I still need… I still need the programmatic access for Discovery.

950 01:13:37.560 ⇒ 01:13:40.780 Samuel Roberts: I’m just saying, in terms of section, you can keep it up.

951 01:13:40.880 ⇒ 01:13:45.829 Samuel Roberts: But in terms of where it’s gonna go, it’s more of a logging thing than a monster migration thing.

952 01:13:47.720 ⇒ 01:13:50.709 Samuel Roberts: Yeah, I wouldn’t… I wouldn’t slide it back, I was just reorganizing the…

953 01:13:51.100 ⇒ 01:13:51.610 Uttam Kumaran: Yeah.

954 01:13:51.610 ⇒ 01:13:55.340 Samuel Roberts: What’s where… Alright.

955 01:14:13.020 ⇒ 01:14:18.990 Uttam Kumaran: Yeah, I mean, it’s gonna… I think it’s really, like, this routing, retrieval, eval… And…

956 01:14:19.530 ⇒ 01:14:25.199 Uttam Kumaran: other features is really what I’m gonna ask, like, you guys need a month for that, and, like.

957 01:14:25.700 ⇒ 01:14:28.100 Uttam Kumaran: If so, then you have it, you know?

958 01:14:33.540 ⇒ 01:14:35.090 Samuel Roberts: That we’re coming to? Yeah.

959 01:14:36.150 ⇒ 01:14:37.819 Samuel Roberts: It’s edge cases…

960 01:14:44.880 ⇒ 01:14:51.010 Samuel Roberts: Excuse me. Yeah. I mean, I think… I think this makes sense if we… if we’re settled on the architecture by the end of next week.

961 01:14:54.000 ⇒ 01:14:57.070 Samuel Roberts: Because, like, what’s going into that is, like, some of the, like.

962 01:14:57.410 ⇒ 01:15:01.380 Samuel Roberts: like, the basic evals we have right now, but we… as we make that more robust.

963 01:15:02.640 ⇒ 01:15:08.520 Samuel Roberts: We’ll know what’s better there, and then there’s some of the changes to the…

964 01:15:09.560 ⇒ 01:15:13.310 Samuel Roberts: The retrieval of the verbatim stuff, that’ll…

965 01:15:13.620 ⇒ 01:15:16.670 Samuel Roberts: affect that, too. So I think… I think this makes sense.

966 01:15:18.680 ⇒ 01:15:21.690 Uttam Kumaran: Do you guys feel like this is enough to start, like, ticketing things out?

967 01:15:26.260 ⇒ 01:15:29.229 Uttam Kumaran: I guess, like, for me, I think you… we probably need to.

968 01:15:29.230 ⇒ 01:15:29.779 Samuel Roberts: this Sunday.

969 01:15:29.780 ⇒ 01:15:34.130 Uttam Kumaran: meetings around… All the eval work, and, like, loop in.

970 01:15:34.550 ⇒ 01:15:35.110 Uttam Kumaran: of…

971 01:15:38.840 ⇒ 01:15:39.980 Samuel Roberts: Yeah.

972 01:15:40.700 ⇒ 01:15:44.850 Uttam Kumaran: And then, basically, I think one thing that we need to do… one thing maybe I’ll ask

973 01:15:45.270 ⇒ 01:15:48.190 Uttam Kumaran: You guys should do is to assign… do assignments?

974 01:15:49.510 ⇒ 01:15:51.330 Uttam Kumaran: And then that way I can…

975 01:15:51.530 ⇒ 01:15:53.149 Samuel Roberts: I can look at.

976 01:15:53.270 ⇒ 01:15:54.430 Uttam Kumaran: Workload.

977 01:15:58.650 ⇒ 01:15:59.760 Samuel Roberts: Okay, yeah.

978 01:16:01.270 ⇒ 01:16:06.599 Uttam Kumaran: If you guys are able to do assignments, then I can look at workload and see, like, what we can split off to who.

979 01:16:06.930 ⇒ 01:16:11.320 Uttam Kumaran: Because I want to loop in for Nav for, like, maybe, like, 5 to 10 hours a week on Andy’s stuff.

980 01:16:11.580 ⇒ 01:16:15.490 Uttam Kumaran: I have to… I wanna… I need to jump to go call Zoran.

981 01:16:16.590 ⇒ 01:16:18.290 Samuel Roberts: Oh, yeah, I didn’t realize how late it was.

982 01:16:18.640 ⇒ 01:16:24.120 Uttam Kumaran: Yeah, but maybe I’ll let you guys continue. We haven’t gotten to new AND features, but I think this is good progress today.

983 01:16:25.400 ⇒ 01:16:26.270 Samuel Roberts: Definitely.

984 01:16:27.050 ⇒ 01:16:32.429 Uttam Kumaran: What I don’t want to do is have some people working on news, some people also, like, basically working on too much.

985 01:16:32.860 ⇒ 01:16:33.220 Samuel Roberts: Yeah.

986 01:16:33.220 ⇒ 01:16:43.319 Uttam Kumaran: let’s just see how much we can split, logically, between Casey and Mustasa, and then the new features, we can decide, like, once we start to scope that out, maybe tomorrow we’ll do another session like this.

987 01:16:44.290 ⇒ 01:16:44.860 Samuel Roberts: Boom.

988 01:16:46.570 ⇒ 01:16:48.140 Uttam Kumaran: Alright, I’m gonna go jump to talk to Zero.

989 01:16:48.690 ⇒ 01:16:49.830 Samuel Roberts: Alright.

990 01:16:49.990 ⇒ 01:16:50.530 Samuel Roberts: How’s it going?

991 01:16:50.530 ⇒ 01:16:51.200 Casie Aviles: Thank you.

992 01:16:52.970 ⇒ 01:16:57.400 Samuel Roberts: Okay, so, I think some of this is probably…

993 01:16:58.330 ⇒ 01:17:02.280 Samuel Roberts: Pretty easy to do assignments, because we’re kind of already working on some things, right?

994 01:17:02.890 ⇒ 01:17:03.610 Casie Aviles: Hmm.

995 01:17:08.630 ⇒ 01:17:14.980 Samuel Roberts: So… Like, yeah, I mean… It’s clear…

996 01:17:16.980 ⇒ 01:17:19.999 Samuel Roberts: I wish it showed them more clearly in the Gantt.

997 01:17:20.220 ⇒ 01:17:22.680 Samuel Roberts: We could maybe do the color assignments if we need.

998 01:17:25.300 ⇒ 01:17:27.720 Casie Aviles: Color based on who’s assigned?

999 01:17:28.160 ⇒ 01:17:33.809 Samuel Roberts: We could, if that’s helpful for us right now, visualizing, but if we’re just gonna do the assignments and then he can run with that, I think that’s fine.

1000 01:17:34.770 ⇒ 01:17:35.330 Casie Aviles: Yeah, common.

1001 01:17:35.330 ⇒ 01:17:40.459 Samuel Roberts: Let’s just… let’s just walk down, anything that’s not really assigned in the first few things.

1002 01:17:41.210 ⇒ 01:17:42.820 Samuel Roberts: So… Great.

1003 01:17:42.820 ⇒ 01:17:46.180 Casie Aviles: Now I’m mostly handling these.

1004 01:17:46.940 ⇒ 01:17:51.340 Casie Aviles: But… I think… We can split it up.

1005 01:17:52.480 ⇒ 01:17:54.009 Samuel Roberts: The ongoing maintenance, yeah.

1006 01:17:54.010 ⇒ 01:17:55.190 Casie Aviles: Yeah, it’s mostly me.

1007 01:17:55.190 ⇒ 01:18:03.120 Samuel Roberts: Yeah, ongoing maintenance, yeah… That one, yeah.

1008 01:18:03.650 ⇒ 01:18:07.260 Samuel Roberts: I would say, go ahead… I mean…

1009 01:18:08.090 ⇒ 01:18:11.129 Samuel Roberts: however… I guess however you guys feel like splitting that up.

1010 01:18:11.620 ⇒ 01:18:16.720 Samuel Roberts: I’m not, you know, I think you definitely have, like, the zip code stuff better.

1011 01:18:17.430 ⇒ 01:18:27.689 Samuel Roberts: Does that mean… is that, like, that, and then… Central Doc, go to… Mustafa, does that work?

1012 01:18:28.700 ⇒ 01:18:33.030 Samuel Roberts: System outages is kind of a… you know…

1013 01:18:33.680 ⇒ 01:18:36.159 Samuel Roberts: Catch-all, I want to say.

1014 01:18:36.650 ⇒ 01:18:41.229 Samuel Roberts: And update aliases is definitely a good one to make sure we do.

1015 01:18:42.250 ⇒ 01:18:44.879 Samuel Roberts: regularly, so I don’t know who wants that.

1016 01:18:44.980 ⇒ 01:18:53.229 Casie Aviles: Yeah, I can take that. That’s also involved with, you know, updating the prompt that we have for the DBQuery tool.

1017 01:18:53.450 ⇒ 01:18:55.090 Casie Aviles: So, I can take that.

1018 01:18:55.670 ⇒ 01:18:56.340 Samuel Roberts: Okay, cool.

1019 01:18:56.340 ⇒ 01:19:00.619 Casie Aviles: Do I need to… does this require me putting, like, ours?

1020 01:19:00.970 ⇒ 01:19:05.470 Samuel Roberts: I think just do the assignments, and, like, we’ll… we’ll go from there. We can always come back and…

1021 01:19:06.150 ⇒ 01:19:09.289 Samuel Roberts: Because, like, the maintenance hours aren’t going to make any sense anyway.

1022 01:19:09.800 ⇒ 01:19:11.180 Casie Aviles: Oh, okay, yeah.

1023 01:19:11.810 ⇒ 01:19:14.289 Samuel Roberts: At least, it’s not the way the Kant represents them.

1024 01:19:23.350 ⇒ 01:19:28.749 Samuel Roberts: Yeah, I would say that’s… you know, we could even do two, if you want to both be on that, maybe, just because, like, it’s a…

1025 01:19:31.700 ⇒ 01:19:37.359 Samuel Roberts: it’s a weird one. I’m less worried about the assignments for those than I am the other tasks.

1026 01:19:38.090 ⇒ 01:19:46.660 Samuel Roberts: But important to make sure we’re not overlooking the zip and the central dock, but the system outages is a… that’s a different fire to put out, you know?

1027 01:19:47.700 ⇒ 01:20:01.950 Samuel Roberts: Okay, let’s jump back up. What else is unassigned in the top one? So those, that’s all pretty good. Determine best monster architecture, that’s just a, what we’ll call it, milestone, so that’s fine, we’ll all be in on that.

1028 01:20:02.450 ⇒ 01:20:10.755 Samuel Roberts: This all kind of got pushed… Master logs to be queue…

1029 01:20:15.360 ⇒ 01:20:25.020 Samuel Roberts: I’m thinking… Well, I don’t know. You guys tell me who feels more comfortable with, the BigQuery stuff.

1030 01:20:28.250 ⇒ 01:20:29.759 Casie Aviles: Our homepage, yeah.

1031 01:20:29.760 ⇒ 01:20:36.949 Mustafa Raja: Yeah, so, for BigQuery, I’ve already started working on some of the stuff with Damesh.

1032 01:20:37.400 ⇒ 01:20:44.039 Samuel Roberts: Okay, then I would say go ahead, and probably take… most of those are, like, chained, so it’s not, like, a lot of overlap, but yeah.

1033 01:20:44.040 ⇒ 01:20:44.480 Mustafa Raja: Yeah.

1034 01:20:44.480 ⇒ 01:20:46.379 Samuel Roberts: But those two must stop, though, yeah.

1035 01:20:47.040 ⇒ 01:20:50.490 Casie Aviles: So these ones as well, 14, 15, I don’t know.

1036 01:20:50.490 ⇒ 01:20:54.910 Samuel Roberts: Yeah, that one’s a milestone, so it’s kind of a wait.

1037 01:20:54.910 ⇒ 01:20:55.260 Casie Aviles: Oh.

1038 01:20:55.260 ⇒ 01:20:59.110 Samuel Roberts: But yeah, I would say… I would say both of those. Just assign them for now, that way we know who’.

1039 01:20:59.110 ⇒ 01:20:59.960 Mustafa Raja: is kind of…

1040 01:21:00.110 ⇒ 01:21:00.940 Samuel Roberts: The point person.

1041 01:21:00.940 ⇒ 01:21:05.939 Mustafa Raja: Yeah, programmatic access is just configuring the… what’s it called?

1042 01:21:06.460 ⇒ 01:21:10.620 Mustafa Raja: Service, service account with, you know.

1043 01:21:11.380 ⇒ 01:21:15.379 Mustafa Raja: So we, we do have access to it, we are able to write and write and read.

1044 01:21:15.380 ⇒ 01:21:16.050 Samuel Roberts: Okay.

1045 01:21:17.180 ⇒ 01:21:19.480 Mustafa Raja: And so easy, pretty easy.

1046 01:21:19.640 ⇒ 01:21:25.229 Casie Aviles: Okay. One thing… I forgot is,

1047 01:21:25.420 ⇒ 01:21:29.550 Casie Aviles: Is there, like, a difference between the master logs and the migrate logging?

1048 01:21:29.700 ⇒ 01:21:31.490 Casie Aviles: 16 and 18.

1049 01:21:32.390 ⇒ 01:21:33.400 Mustafa Raja: Hmm…

1050 01:21:33.400 ⇒ 01:21:38.289 Samuel Roberts: Monster logs, I think, might… Be more of the, like.

1051 01:21:39.190 ⇒ 01:21:46.750 Samuel Roberts: logging of the requests and stuff. The other logging that’s going to Snowflake is the, like, thumbs up, thumbs down stuff, right?

1052 01:21:48.540 ⇒ 01:21:49.690 Casie Aviles: Oh, okay.

1053 01:21:50.020 ⇒ 01:21:54.030 Samuel Roberts: So I’m thinking, like, Mastra, like, the thing we’ll get out of this is a little better…

1054 01:21:55.170 ⇒ 01:22:03.340 Samuel Roberts: observability and traceability for some requests. That is harder to do. We had, like, we figured out a system a little bit for the last 30 days, at least, in NIDM.

1055 01:22:03.570 ⇒ 01:22:09.670 Samuel Roberts: But we want to make sure we don’t lose that evaluation Logging… log stuff.

1056 01:22:10.200 ⇒ 01:22:11.060 Casie Aviles: Right.

1057 01:22:12.420 ⇒ 01:22:17.169 Casie Aviles: I… I can assist here for… Okay. …for these bottom two.

1058 01:22:18.580 ⇒ 01:22:19.309 Samuel Roberts: Okay, cool.

1059 01:22:21.420 ⇒ 01:22:28.529 Samuel Roberts: And then… let’s see, evaluation data set, we have something, right?

1060 01:22:30.700 ⇒ 01:22:40.719 Casie Aviles: Yeah, we have, like, a very old one, when we started this, the work for ABC, but that’s very outdated now.

1061 01:22:40.720 ⇒ 01:22:44.830 Samuel Roberts: Yeah, well, we’ve also pulled some of the thumbs up, thumbs-down stuff for some testing,

1062 01:22:45.840 ⇒ 01:22:50.259 Samuel Roberts: of the Mastra, to see how well it compares, and if it’s improved. So, I think…

1063 01:22:51.230 ⇒ 01:22:54.990 Samuel Roberts: Like, Mustafa’s done some of that for the master stuff.

1064 01:22:56.160 ⇒ 01:23:01.830 Samuel Roberts: So I would… I don’t know, whoever… yeah, I would say that’s probably fine for Mustafa.

1065 01:23:02.740 ⇒ 01:23:08.700 Samuel Roberts: Some of this is the routing and stuff that we’ve done in the…

1066 01:23:09.650 ⇒ 01:23:11.690 Samuel Roberts: Yeah, he’s kind of working on the…

1067 01:23:12.540 ⇒ 01:23:13.819 Samuel Roberts: I’m gonna stop at you here.

1068 01:23:14.140 ⇒ 01:23:15.220 Mustafa Raja: Yo, yes.

1069 01:23:15.500 ⇒ 01:23:19.040 Samuel Roberts: Okay, sorry, I thought you were muted, I wasn’t sure. You had stepped away for a second.

1070 01:23:19.430 ⇒ 01:23:23.590 Samuel Roberts: I would say, like, the fixed copy stuff you’ve also been working on, right?

1071 01:23:24.370 ⇒ 01:23:32.160 Mustafa Raja: Yeah, yeah, fixed copy one. Yeah, today I was just a little disappointed in… Yeah, well, yeah, yeah.

1072 01:23:32.160 ⇒ 01:23:33.289 Samuel Roberts: That’s too bad, but we’ll…

1073 01:23:33.290 ⇒ 01:23:39.460 Mustafa Raja: But, I’ll work on something custom. I think that’s our best bet.

1074 01:23:39.630 ⇒ 01:23:41.460 Mustafa Raja: What do you think about that?

1075 01:23:41.840 ⇒ 01:23:47.800 Samuel Roberts: Yeah, I think… I think something… yeah, we’re gonna have to do something custom. Yeah.

1076 01:23:47.800 ⇒ 01:23:56.130 Mustafa Raja: I think I had, because, the output that I received, the thing that we have, built ourselves is much more better.

1077 01:23:56.260 ⇒ 01:23:57.390 Mustafa Raja: Anyways… Okay.

1078 01:23:58.270 ⇒ 01:24:02.660 Samuel Roberts: Alright, cool. And then what else is unassigned at this point? Fixed performance or correctness issues?

1079 01:24:06.640 ⇒ 01:24:08.410 Samuel Roberts: I mean, I would probably do.

1080 01:24:09.460 ⇒ 01:24:11.479 Samuel Roberts: It’s always kind of in the same…

1081 01:24:12.170 ⇒ 01:24:18.230 Mustafa Raja: Yeah, this most likely comes under, what’s it called? Under triageies, right?

1082 01:24:19.860 ⇒ 01:24:20.440 Casie Aviles: Hmm.

1083 01:24:20.440 ⇒ 01:24:22.860 Samuel Roberts: I suppose, I suppose.

1084 01:24:27.020 ⇒ 01:24:33.669 Samuel Roberts: Okay, what else do we have here? Validate document relevance and retrieval accuracy. That’s also part of the evaluation stuff, so…

1085 01:24:34.370 ⇒ 01:24:38.849 Samuel Roberts: Either… I suppose either one of you can run with that, though, at this point, once we have the data set, so…

1086 01:24:42.100 ⇒ 01:24:46.849 Samuel Roberts: Cool. Confirm new central doc structure with client.

1087 01:24:48.210 ⇒ 01:24:50.890 Samuel Roberts: How, that’s a milestone, I wouldn’t worry about it.

1088 01:24:53.650 ⇒ 01:24:54.959 Samuel Roberts: Alright, and then…

1089 01:24:54.960 ⇒ 01:24:59.990 Mustafa Raja: Yeah, I’m, I’m also wondering how we want to take over this, like…

1090 01:25:00.800 ⇒ 01:25:05.029 Mustafa Raja: We need to identify the issues within the central dock.

1091 01:25:05.290 ⇒ 01:25:09.999 Mustafa Raja: We did look into one last month with Patricia.

1092 01:25:10.180 ⇒ 01:25:13.479 Mustafa Raja: I’m wondering, if there are more…

1093 01:25:13.600 ⇒ 01:25:15.679 Mustafa Raja: More sort of issues like that.

1094 01:25:16.200 ⇒ 01:25:18.960 Mustafa Raja: Yeah.

1095 01:25:18.960 ⇒ 01:25:20.830 Samuel Roberts: Yeah, I mean, we may want to do…

1096 01:25:21.830 ⇒ 01:25:29.130 Samuel Roberts: as part of the validate document relevance and retrieval accuracy, it could be identifying, like…

1097 01:25:30.000 ⇒ 01:25:38.290 Samuel Roberts: When there’s issues with it, is it a central dock issue, and how can we fix that at the organization of the central dock level?

1098 01:25:39.270 ⇒ 01:25:40.310 Mustafa Raja: Yep, yeah.

1099 01:25:40.710 ⇒ 01:25:51.139 Samuel Roberts: So maybe over the next couple weeks, we can make sure to keep a running list of, like, issues that the central doc has that we think we have solutions for, and present that to them as a way to reorganize the central doc.

1100 01:25:51.690 ⇒ 01:25:52.360 Mustafa Raja: Yep.

1101 01:25:53.160 ⇒ 01:25:53.770 Samuel Roberts: Okay.

1102 01:25:54.620 ⇒ 01:25:57.400 Samuel Roberts: Let’s see, define core valuation metrics…

1103 01:26:03.750 ⇒ 01:26:04.420 Casie Aviles: Oh, shit.

1104 01:26:04.900 ⇒ 01:26:05.360 Mustafa Raja: condition.

1105 01:26:06.830 ⇒ 01:26:17.700 Casie Aviles: Yeah, I’m interested here, but I think Utama’s also mentioning to loop in. Should we loop in, like, front of here already, or…

1106 01:26:18.340 ⇒ 01:26:18.889 Casie Aviles: Just,

1107 01:26:19.450 ⇒ 01:26:20.230 Casie Aviles: That’s for now.

1108 01:26:20.660 ⇒ 01:26:24.670 Samuel Roberts: I think he’s trying to figure out where to put Pernel, so I would just say let’s keep assigning

1109 01:26:25.100 ⇒ 01:26:32.440 Samuel Roberts: you guys, and let him make that determination of, like, yeah, this would be a good area. Because I think it… it does make sense for…

1110 01:26:32.620 ⇒ 01:26:34.660 Samuel Roberts: That, cause it’s a little bit separate.

1111 01:26:34.900 ⇒ 01:26:37.699 Samuel Roberts: then… The monster stuff right now.

1112 01:26:39.580 ⇒ 01:26:43.379 Samuel Roberts: Yeah, I would say, I mean…

1113 01:26:44.270 ⇒ 01:26:47.000 Samuel Roberts: However you guys are feeling about who takes what here.

1114 01:26:49.750 ⇒ 01:26:52.450 Samuel Roberts: It’s kind of as evenly as possible, maybe, I don’t know.

1115 01:26:53.640 ⇒ 01:26:58.719 Mustafa Raja: Yeah, so… Yeah, that’s good.

1116 01:27:01.390 ⇒ 01:27:03.070 Samuel Roberts: you know, logging in Astra.

1117 01:27:08.270 ⇒ 01:27:11.529 Samuel Roberts: I mean, who has the most experience with blank views at this point?

1118 01:27:14.670 ⇒ 01:27:16.889 Mustafa Raja: What are we implementing languages for?

1119 01:27:17.380 ⇒ 01:27:19.749 Mustafa Raja: The observability only, right?

1120 01:27:19.750 ⇒ 01:27:20.530 Casie Aviles: Yay.

1121 01:27:20.660 ⇒ 01:27:21.910 Casie Aviles: For observability is.

1122 01:27:21.910 ⇒ 01:27:22.490 Mustafa Raja: Please implement.

1123 01:27:22.490 ⇒ 01:27:23.030 Samuel Roberts: Yeah.

1124 01:27:25.110 ⇒ 01:27:28.640 Mustafa Raja: Yeah, I think either of us have pretty good…

1125 01:27:28.980 ⇒ 01:27:31.449 Samuel Roberts: Okay, so then go ahead, either one of you.

1126 01:27:31.450 ⇒ 01:27:31.870 Mustafa Raja: Yeah.

1127 01:27:31.870 ⇒ 01:27:37.660 Samuel Roberts: And make sure we’re kind of doing this, because this is the time when the different weeks matter, I guess, is overloaded or something, but…

1128 01:27:37.870 ⇒ 01:27:41.730 Samuel Roberts: Okay, yeah, both of these there is fine. Eval logging in Mastra…

1129 01:27:43.300 ⇒ 01:27:44.730 Mustafa Raja: What’s event logging, though?

1130 01:27:46.820 ⇒ 01:27:50.410 Samuel Roberts: eval logging is distinct from the master logs, right?

1131 01:27:51.460 ⇒ 01:27:56.170 Samuel Roberts: eval… this is, like, the eval logging? Right, why does that make sense?

1132 01:27:56.560 ⇒ 01:28:04.300 Casie Aviles: I think that’s… Like, producing cores for each ex… like…

1133 01:28:04.300 ⇒ 01:28:09.140 Samuel Roberts: Yeah, I think that that would be… yeah, logging’s a little bit of a weird word there, but I see what you’re saying.

1134 01:28:15.410 ⇒ 01:28:19.850 Casie Aviles: Yeah, because, like, after every trace, I believe.

1135 01:28:20.050 ⇒ 01:28:24.220 Casie Aviles: It’s going to output, like, a score if we’ve set it up that way.

1136 01:28:24.350 ⇒ 01:28:27.340 Casie Aviles: Right. That’s how it will be working.

1137 01:28:28.060 ⇒ 01:28:28.660 Samuel Roberts: Okay.

1138 01:28:30.660 ⇒ 01:28:35.880 Samuel Roberts: Okay, it’s… I may have to run. Do you guys think you’re able to assign the rest of these?

1139 01:28:36.420 ⇒ 01:28:37.889 Casie Aviles: Oh, sure, sure.

1140 01:28:37.890 ⇒ 01:28:44.240 Samuel Roberts: Okay, sorry, I just didn’t realize what time it was getting to, and I gotta go relieve my wife with the baby, so she can get ready for work.

1141 01:28:44.470 ⇒ 01:28:45.020 Samuel Roberts: Okay.

1142 01:28:45.480 ⇒ 01:28:48.400 Samuel Roberts: Yeah, I guess finish this up, and then… then share it.

1143 01:28:49.030 ⇒ 01:28:51.260 Casie Aviles: We’ll just share it in the ABC channel.

1144 01:28:51.570 ⇒ 01:28:52.430 Samuel Roberts: Yeah, that’s good.

1145 01:28:53.620 ⇒ 01:28:54.530 Samuel Roberts: Alright, thank you guys.

1146 01:28:55.020 ⇒ 01:28:55.680 Casie Aviles: Thank you.

1147 01:28:55.680 ⇒ 01:28:57.790 Samuel Roberts: Let me know if there are any questions, I’ll be on Slack, but…

1148 01:28:58.000 ⇒ 01:29:01.690 Samuel Roberts: I may have a baby with me, so… Okay. Bye.

1149 01:29:01.950 ⇒ 01:29:02.930 Mustafa Raja: Thank you. Bye.

1150 01:29:09.440 ⇒ 01:29:10.790 Casie Aviles: Completely…

1151 01:29:13.850 ⇒ 01:29:23.280 Casie Aviles: So, are… you’re still, like, the capacity, in terms of, like, I was just thinking of capacity, you’re still, like, at how many hours per… per…

1152 01:29:23.770 ⇒ 01:29:26.240 Casie Aviles: per week, on ABC.

1153 01:29:26.770 ⇒ 01:29:28.860 Casie Aviles: Because I think they gave me more…

1154 01:29:28.860 ⇒ 01:29:30.760 Mustafa Raja: Best, I don’t know.

1155 01:29:31.080 ⇒ 01:29:33.439 Mustafa Raja: Let me, let me check in operating.

1156 01:29:33.780 ⇒ 01:29:37.960 Casie Aviles: Yeah, I think they gave me more, so I might have to take more.

1157 01:29:37.970 ⇒ 01:29:39.970 Mustafa Raja: Oh, how do you check it, though?

1158 01:29:41.160 ⇒ 01:29:45.030 Casie Aviles: It’s just based on what I talked with Awash last month, I…

1159 01:29:46.800 ⇒ 01:29:48.130 Mustafa Raja: Yeah.

1160 01:29:48.500 ⇒ 01:29:52.730 Casie Aviles: As far as I know, I have 20 hours per week for this month.

1161 01:29:52.730 ⇒ 01:29:55.050 Mustafa Raja: To be honest, I’m working on only two clients.

1162 01:29:55.640 ⇒ 01:29:56.650 Casie Aviles: right now.

1163 01:29:56.650 ⇒ 01:29:57.250 Mustafa Raja: 11.

1164 01:29:58.360 ⇒ 01:30:02.419 Mustafa Raja: It’s defaulted on this, so we should be good to go.

1165 01:30:03.600 ⇒ 01:30:04.450 Casie Aviles: Okay.

1166 01:30:04.800 ⇒ 01:30:10.660 Casie Aviles: Yeah, I only have… Let me remove this.

1167 01:30:11.040 ⇒ 01:30:16.699 Casie Aviles: Yeah, I only have ABC at the moment. I think Insomnia is done for now.

1168 01:30:17.840 ⇒ 01:30:20.319 Casie Aviles: And I think I’ll be moving to…

1169 01:30:20.670 ⇒ 01:30:25.629 Casie Aviles: Eden Martech. I don’t know the allocations for that skill, but…

1170 01:30:27.200 ⇒ 01:30:31.819 Casie Aviles: For now, I have the majority of my time will still be ABC, so…

1171 01:30:32.820 ⇒ 01:30:37.660 Mustafa Raja: Yeah, I think I’ll be… I’ll be working some… some on Hitra.

1172 01:30:38.160 ⇒ 01:30:42.289 Mustafa Raja: So yeah, allocations… allocations are no question mark.

1173 01:30:42.890 ⇒ 01:30:43.930 Casie Aviles: Okay, so…

1174 01:30:43.930 ⇒ 01:30:44.490 Mustafa Raja: Yeah.

1175 01:30:44.490 ⇒ 01:30:48.749 Casie Aviles: Yeah, so let’s just… I guess we can… we can just assign for now, based on…

1176 01:30:49.610 ⇒ 01:30:59.739 Casie Aviles: whether we want to take it on or not, and then later on, probably, we can probably switch out the people anyways, depending on, like, capacity, I guess.

1177 01:31:00.060 ⇒ 01:31:01.059 Mustafa Raja: Yeah, yeah, no.

1178 01:31:02.400 ⇒ 01:31:03.530 Mustafa Raja: Alright, so…

1179 01:31:03.620 ⇒ 01:31:11.459 Casie Aviles: Because I feel like it’s a little early right now to know if we can take this on or not, because we have other clients, too.

1180 01:31:11.990 ⇒ 01:31:13.379 Mustafa Raja: Yeah, yeah, you’re right.

1181 01:31:14.170 ⇒ 01:31:14.850 Mustafa Raja: Excuse me.

1182 01:31:14.850 ⇒ 01:31:20.129 Casie Aviles: And we don’t really have any estimates in terms of, like, hours, or… it’s just…

1183 01:31:20.360 ⇒ 01:31:23.399 Casie Aviles: A duration for weeks, but yeah, okay.

1184 01:31:24.280 ⇒ 01:31:28.500 Casie Aviles: So this will be tentative anyway.

1185 01:31:29.730 ⇒ 01:31:30.790 Casie Aviles: What else?

1186 01:31:32.360 ⇒ 01:31:39.199 Mustafa Raja: I think… I think these ones we will be talking tomorrow, right? So only these ones we need to…

1187 01:31:40.170 ⇒ 01:31:41.510 Mustafa Raja: Assigned aid.

1188 01:31:43.340 ⇒ 01:31:48.540 Mustafa Raja: The evaluation? So we need to only assign evaluation and GCB, right?

1189 01:31:48.910 ⇒ 01:31:51.499 Mustafa Raja: No, not the ones at the bottom, right?

1190 01:31:54.060 ⇒ 01:31:55.929 Casie Aviles: Which bucket on? This one?

1191 01:31:55.930 ⇒ 01:31:56.619 Mustafa Raja: Yeah, they’re…

1192 01:31:56.620 ⇒ 01:31:57.370 Casie Aviles: from reporting.

1193 01:31:57.370 ⇒ 01:32:07.600 Mustafa Raja: Yeah, the new stuff, yeah, yeah. So, let’s go… I can take embedding pipeline. I worked on it a little today, the embedding pipeline, yeah.

1194 01:32:07.600 ⇒ 01:32:08.039 Casie Aviles: I bet.

1195 01:32:08.040 ⇒ 01:32:12.119 Mustafa Raja: I worked on analytics today, I have a good sense of how it should go.

1196 01:32:12.560 ⇒ 01:32:18.810 Mustafa Raja: Okay. So I should be good with it. And then…

1197 01:32:19.990 ⇒ 01:32:33.880 Mustafa Raja: What else is there? I think… So validating, text to SQL, would you be doing that, with, the dataset that, that is in the Google Sheets?

1198 01:32:36.060 ⇒ 01:32:40.700 Casie Aviles: Hmm… yeah, yeah, I mean, that’s worth doing as well.

1199 01:32:40.950 ⇒ 01:32:41.740 Mustafa Raja: Yeah.

1200 01:32:42.740 ⇒ 01:32:46.960 Mustafa Raja: Any application, I guess you could assign that to me.

1201 01:32:48.670 ⇒ 01:32:50.420 Casie Aviles: No, it’s more like… it’s more like a…

1202 01:32:50.420 ⇒ 01:32:51.170 Mustafa Raja: Oh, yes.

1203 01:32:51.170 ⇒ 01:32:51.660 Casie Aviles: on.

1204 01:32:51.660 ⇒ 01:32:53.839 Mustafa Raja: So this, this doesn’t need to be assigned, yeah.

1205 01:32:53.970 ⇒ 01:32:54.800 Mustafa Raja: Okay.

1206 01:32:55.870 ⇒ 01:33:05.190 Casie Aviles: Let’s see… Well, it’s spaced out, but… Oh, yeah.

1207 01:33:05.190 ⇒ 01:33:06.769 Mustafa Raja: I guess I could take a ratings.

1208 01:33:07.460 ⇒ 01:33:11.200 Casie Aviles: Yeah, we could… we could probably… Split.

1209 01:33:11.200 ⇒ 01:33:12.720 Mustafa Raja: Give me the slack also.

1210 01:33:15.380 ⇒ 01:33:16.090 Casie Aviles: K…

1211 01:33:16.090 ⇒ 01:33:20.409 Mustafa Raja: Validate alerts, we are… what’s forced failure?

1212 01:33:21.130 ⇒ 01:33:31.699 Casie Aviles: I think what that just means is, we… we want to deliberately… Have, like, errors or failures.

1213 01:33:32.370 ⇒ 01:33:34.219 Casie Aviles: So we can test if…

1214 01:33:34.360 ⇒ 01:33:41.490 Casie Aviles: the alerts are working as intended, but I’m not sure if that’s something we can force At the moment.

1215 01:33:42.000 ⇒ 01:33:44.570 Casie Aviles: So…

1216 01:33:44.570 ⇒ 01:33:49.950 Mustafa Raja: And we want to, we want to… so this is more like validating if alerts are working.

1217 01:33:50.520 ⇒ 01:33:55.070 Casie Aviles: Yeah, yeah, mostly. Yeah, that’s… I think that’s the point of this one.

1218 01:33:56.470 ⇒ 01:34:00.960 Mustafa Raja: So that would, okay, error or latency, right? So, okay.

1219 01:34:01.670 ⇒ 01:34:04.609 Mustafa Raja: Yeah, you could assign that to me if you feel good about that.

1220 01:34:05.860 ⇒ 01:34:08.650 Casie Aviles: Okay, well, for now, just a sec, please.

1221 01:34:09.000 ⇒ 01:34:13.660 Casie Aviles: It’s setting… it’s setting the R’s, sorry. Let me get rid of the R’s.

1222 01:34:14.200 ⇒ 01:34:14.880 Mustafa Raja: Right.

1223 01:34:16.100 ⇒ 01:34:21.699 Mustafa Raja: Yeah, Instagram sends us an email every time a task is assigned, it’s an.

1224 01:34:21.700 ⇒ 01:34:22.450 Casie Aviles: Oh, really?

1225 01:34:22.900 ⇒ 01:34:26.070 Mustafa Raja: Yeah, my phone has been vibrating.

1226 01:34:29.670 ⇒ 01:34:34.539 Mustafa Raja: Yeah, so for… and for Heroku, I looked into it.

1227 01:34:34.920 ⇒ 01:34:40.720 Mustafa Raja: I remember setting up a… Profile and stuff.

1228 01:34:41.250 ⇒ 01:34:43.230 Casie Aviles: I see, I see.

1229 01:34:43.230 ⇒ 01:34:49.830 Mustafa Raja: So, so, A profile, and then, what’s it called?

1230 01:34:50.150 ⇒ 01:35:04.390 Mustafa Raja: package JSON also, which tells it that, we want to deploy this… this project, and then in that project, we’d have our proc file and stuff.

1231 01:35:04.580 ⇒ 01:35:11.310 Mustafa Raja: So that is how… it navigated. I… but that… that… that whole thing was, like, pushed.

1232 01:35:11.690 ⇒ 01:35:15.279 Mustafa Raja: in the repo, And I cannot find it anymore.

1233 01:35:16.070 ⇒ 01:35:17.330 Casie Aviles: Oh…

1234 01:35:17.980 ⇒ 01:35:19.630 Mustafa Raja: Which is weird, right?

1235 01:35:20.360 ⇒ 01:35:21.230 Casie Aviles: Yeah…

1236 01:35:21.470 ⇒ 01:35:26.540 Mustafa Raja: Even if the commits were reverted, the rewards… revert commits should show up.

1237 01:35:26.690 ⇒ 01:35:28.790 Mustafa Raja: I don’t know what’s going on there.

1238 01:35:29.580 ⇒ 01:35:34.620 Casie Aviles: Oh, but you did send me, like, oh, I thought… yeah, no, never mind, it’s fine, it’s fine.

1239 01:35:36.200 ⇒ 01:35:41.590 Mustafa Raja: Yeah, so… so how it’s going to be is we’ll have to set up,

1240 01:35:42.800 ⇒ 01:35:53.370 Mustafa Raja: what’s it called? PackageSon, to tell Heroku that we want to deploy, some…

1241 01:35:54.570 ⇒ 01:36:02.400 Mustafa Raja: Application, and then after that, we should have a profile, To start building and deploying.

1242 01:36:03.400 ⇒ 01:36:04.610 Mustafa Raja: in the repo.

1243 01:36:07.100 ⇒ 01:36:08.890 Casie Aviles: Okay, I see, I see.

1244 01:36:09.620 ⇒ 01:36:12.850 Casie Aviles: Yeah, I’ll probably take a look.

1245 01:36:13.070 ⇒ 01:36:14.320 Mustafa Raja: Yeah, take a look.

1246 01:36:16.320 ⇒ 01:36:21.309 Mustafa Raja: And let me know if it works out. If not, I’ll take a look in the morning.

1247 01:36:21.310 ⇒ 01:36:27.809 Casie Aviles: Yeah, do you… do you at least… do you have, like, the references you… you were… or was that just, you know, Google?

1248 01:36:27.810 ⇒ 01:36:33.240 Mustafa Raja: Yeah, that was… ChatGPT. Yeah, that was ChatGPT and Google, and I…

1249 01:36:33.970 ⇒ 01:36:37.159 Mustafa Raja: And I have been trying to look into that.

1250 01:36:37.900 ⇒ 01:36:44.070 Mustafa Raja: I cannot really find the… What’s it called? The chat.

1251 01:36:46.240 ⇒ 01:36:53.889 Casie Aviles: Okay, yeah, no worries, no worries. I’ll try to just do the same thing. I mean, I’ve been asking ChatGPT, but…

1252 01:36:54.210 ⇒ 01:36:59.769 Casie Aviles: I just wanted to make sure, like, if you already did something, then might as well just.

1253 01:36:59.770 ⇒ 01:37:04.780 Mustafa Raja: Yeah, I did, and it was in the repo, I don’t know how or what happened, actually.

1254 01:37:04.900 ⇒ 01:37:08.440 Mustafa Raja: I’ll, this is weird, so I’ll take a look at…

1255 01:37:08.540 ⇒ 01:37:10.509 Mustafa Raja: Add that, but in the morning.

1256 01:37:10.760 ⇒ 01:37:18.620 Casie Aviles: Okay, no problem. Yeah, it might have been my… it might have been my fault, like, something… I… I changed…

1257 01:37:19.370 ⇒ 01:37:27.480 Casie Aviles: I did, like, a change to the dashboard, but it got… the master app was missing, and I’m not sure why.

1258 01:37:28.970 ⇒ 01:37:32.859 Casie Aviles: But I… so I put it back now. I put it back.

1259 01:37:32.860 ⇒ 01:37:37.069 Mustafa Raja: So, oh, that is… that is really weird.

1260 01:37:37.450 ⇒ 01:37:43.789 Casie Aviles: I’m not sure why, I mean, I think it’s because of the git ignore, because when I merged it, it didn’t show any…

1261 01:37:43.790 ⇒ 01:37:45.510 Mustafa Raja: Any changes?

1262 01:37:45.510 ⇒ 01:37:46.420 Casie Aviles: So it’s here now.

1263 01:37:46.530 ⇒ 01:37:47.590 Mustafa Raja: Oh…

1264 01:37:48.180 ⇒ 01:37:50.750 Casie Aviles: I restored it, but…

1265 01:37:50.750 ⇒ 01:37:51.790 Mustafa Raja: Okay…

1266 01:37:52.100 ⇒ 01:37:56.410 Casie Aviles: It’s weird, because… Where is it?

1267 01:37:57.580 ⇒ 01:38:02.499 Casie Aviles: Yeah, this one. I mean, I’m not sure why it disappeared, because…

1268 01:38:04.270 ⇒ 01:38:09.900 Casie Aviles: It’s just one line that I updated, and then I merged it. See? It’s just this, right?

1269 01:38:10.550 ⇒ 01:38:11.400 Mustafa Raja: Yeah…

1270 01:38:11.650 ⇒ 01:38:17.350 Casie Aviles: So I think it’s the gitignore. It probably… for some reason, it got into there.

1271 01:38:17.940 ⇒ 01:38:21.240 Mustafa Raja: Yeah, so some files are definitely missing.

1272 01:38:23.140 ⇒ 01:38:29.320 Casie Aviles: I see, so it’s probably at an earlier branch than…

1273 01:38:30.500 ⇒ 01:38:35.409 Casie Aviles: Because I was getting… I took your files from…

1274 01:38:36.600 ⇒ 01:38:41.420 Casie Aviles: Where was it? The Migration Plan Branch, or Migration Progress Branch.

1275 01:38:42.400 ⇒ 01:38:44.520 Casie Aviles: Yeah, I was getting it from here.

1276 01:38:45.920 ⇒ 01:38:50.589 Mustafa Raja: Yeah, over here, there should just be more files.

1277 01:38:51.130 ⇒ 01:38:51.880 Mustafa Raja: What?

1278 01:38:53.010 ⇒ 01:38:58.039 Mustafa Raja: Nonetheless, we can always replicate this, right? No worries.

1279 01:39:00.650 ⇒ 01:39:01.160 Casie Aviles: Okay.

1280 01:39:01.260 ⇒ 01:39:04.879 Casie Aviles: Alright, so… Yeah, I’ll get back to that.

1281 01:39:07.980 ⇒ 01:39:17.299 Mustafa Raja: And I’ll also try to find the chat. And maybe we could also document this, because, this… this sort of stuff is a little unconventional, right?

1282 01:39:18.210 ⇒ 01:39:23.739 Casie Aviles: Yeah, since we did a monorepo thing for ABC, so…

1283 01:39:23.980 ⇒ 01:39:27.750 Mustafa Raja: Yeah, yeah. And now that we are EPs.

1284 01:39:28.170 ⇒ 01:39:32.890 Mustafa Raja: We should be thinking more with documentation, right?

1285 01:39:32.890 ⇒ 01:39:34.140 Casie Aviles: Yeah, I think so.

1286 01:39:35.930 ⇒ 01:39:38.589 Mustafa Raja: Yeah, I need to hold a meeting with my…

1287 01:39:39.020 ⇒ 01:39:41.379 Casie Aviles: with my SL and CSO…

1288 01:39:41.380 ⇒ 01:39:44.750 Mustafa Raja: And get some Gantt going.

1289 01:39:46.070 ⇒ 01:39:46.860 Mustafa Raja: it tomorrow.

1290 01:39:47.430 ⇒ 01:39:51.739 Casie Aviles: Yeah, me too, I’ll probably… I mean, I just got looped in earlier.

1291 01:39:51.850 ⇒ 01:39:59.590 Casie Aviles: to Eden Martech, Zoran talked to me, but… He just gave me, like… Yeah.

1292 01:39:59.590 ⇒ 01:40:19.009 Mustafa Raja: Yeah, yeah, I guess… I guess for my side, what I was thinking is, you know, to be able to come up with a Gantt chart, I need to be… I need to be up to the par with the goals that client has, you know, and the set of steps that we need to take to sort of get there, right?

1293 01:40:20.780 ⇒ 01:40:26.110 Mustafa Raja: So, yeah. So, so I think, yeah.

1294 01:40:26.680 ⇒ 01:40:32.329 Mustafa Raja: Need to meet with, with, as many stakeholders as we can.

1295 01:40:33.390 ⇒ 01:40:34.600 Casie Aviles: Yeah, that’s true.

1296 01:40:36.710 ⇒ 01:40:37.390 Mustafa Raja: Yeah.

1297 01:40:44.990 ⇒ 01:40:48.589 Mustafa Raja: Also, do we need to work on,

1298 01:40:48.940 ⇒ 01:40:54.639 Mustafa Raja: what’s it called? I know that we need to work on Notion documentations.

1299 01:40:55.320 ⇒ 01:41:02.080 Mustafa Raja: And decks every week. Does the… does… does it start from this week?

1300 01:41:04.380 ⇒ 01:41:06.679 Casie Aviles: Yeah, that’s actually unclear.

1301 01:41:08.200 ⇒ 01:41:09.610 Casie Aviles: I’m not sure, because…

1302 01:41:10.080 ⇒ 01:41:18.230 Casie Aviles: When I was asking Zora what I had to do, like, he was just asking me to help with tickets for now,

1303 01:41:18.420 ⇒ 01:41:24.160 Casie Aviles: But the thing is, they already created the Gantt chart, so I wasn’t sure what else I had to do.

1304 01:41:25.450 ⇒ 01:41:27.740 Mustafa Raja: Oh no, I have to do the Gantt.

1305 01:41:28.900 ⇒ 01:41:29.610 Casie Aviles: Yeah.

1306 01:41:29.920 ⇒ 01:41:35.840 Casie Aviles: So, I’m not sure if we should start doing… The decks as well.

1307 01:41:36.480 ⇒ 01:41:43.400 Casie Aviles: Maybe around Friday, we… it’s, like, more of, like, a week… weekly review, right? Maybe around Friday, we can…

1308 01:41:44.860 ⇒ 01:41:47.050 Casie Aviles: We can… or we can ask in the channel.

1309 01:41:47.050 ⇒ 01:41:49.889 Mustafa Raja: X are for the clients, or for internal.

1310 01:41:50.230 ⇒ 01:41:57.310 Casie Aviles: I think it might be internal, because the CSOs will be doing the decks for the clients, right?

1311 01:41:57.680 ⇒ 01:41:59.460 Casie Aviles: There’s that?

1312 01:41:59.730 ⇒ 01:42:00.230 Casie Aviles: Doesn’t.

1313 01:42:00.230 ⇒ 01:42:01.430 Mustafa Raja: for the clients.

1314 01:42:02.450 ⇒ 01:42:03.190 Casie Aviles: Really?

1315 01:42:03.530 ⇒ 01:42:04.260 Mustafa Raja: Yeah.

1316 01:42:05.230 ⇒ 01:42:12.560 Mustafa Raja: But I haven’t ever done… done a deck for default, ever. I don’t know how…

1317 01:42:14.560 ⇒ 01:42:17.929 Mustafa Raja: Demolade works, because he’s my CSO.

1318 01:42:18.560 ⇒ 01:42:19.220 Mustafa Raja: Cool.

1319 01:42:20.700 ⇒ 01:42:26.089 Casie Aviles: Oh, project review decks, yeah, that might be… might be for the Colorian.

1320 01:42:26.760 ⇒ 01:42:27.455 Mustafa Raja: Yeah.

1321 01:42:34.940 ⇒ 01:42:35.700 Casie Aviles: Okay.

1322 01:42:39.080 ⇒ 01:42:42.579 Mustafa Raja: Yeah, we’ll probably loop… loop in with them nowadays.

1323 01:42:43.060 ⇒ 01:42:47.110 Mustafa Raja: We’ll come up with that. Yeah, let’s just, let’s just assign this thing.

1324 01:42:47.110 ⇒ 01:42:48.080 Casie Aviles: Yeah, sorry.

1325 01:42:51.150 ⇒ 01:42:52.889 Casie Aviles: Where are we?

1326 01:42:53.740 ⇒ 01:42:56.750 Casie Aviles: GCP… Diploma.

1327 01:42:56.750 ⇒ 01:42:58.510 Mustafa Raja: And the musical.

1328 01:42:58.930 ⇒ 01:43:07.140 Mustafa Raja: I think… I think this is… this was something, above also, no?

1329 01:43:09.630 ⇒ 01:43:11.220 Casie Aviles: No, what do you mean, above?

1330 01:43:11.570 ⇒ 01:43:16.250 Mustafa Raja: Like, so 49 might be a duplicate if we scroll up.

1331 01:43:18.420 ⇒ 01:43:20.579 Casie Aviles: Duplicate of which one?

1332 01:43:21.050 ⇒ 01:43:27.979 Mustafa Raja: If you scroll up… And we would see… Implement Master Entry Point.

1333 01:43:32.020 ⇒ 01:43:34.049 Casie Aviles: An administrative one.

1334 01:43:36.230 ⇒ 01:43:37.490 Casie Aviles: Oh, this one.

1335 01:43:38.110 ⇒ 01:43:41.240 Mustafa Raja: Yeah, it’s essentially the same thing, right? Deploying the service.

1336 01:43:47.430 ⇒ 01:43:51.360 Casie Aviles: just deploying the master service client to GCP, this one.

1337 01:43:51.870 ⇒ 01:43:52.745 Mustafa Raja: Yeah.

1338 01:43:54.660 ⇒ 01:43:55.660 Casie Aviles: Hmm…

1339 01:43:56.040 ⇒ 01:43:59.089 Mustafa Raja: But I guess this one is, across all.

1340 01:43:59.270 ⇒ 01:44:02.680 Mustafa Raja: Across all, pipelines and that.

1341 01:44:03.160 ⇒ 01:44:05.049 Mustafa Raja: Does not have any… Yeah.

1342 01:44:05.050 ⇒ 01:44:05.540 Casie Aviles: Yeah.

1343 01:44:05.540 ⇒ 01:44:06.400 Mustafa Raja: if, yeah.

1344 01:44:06.400 ⇒ 01:44:09.829 Casie Aviles: Yeah, yeah, because… For the entry point.

1345 01:44:10.030 ⇒ 01:44:11.960 Casie Aviles: It’s… I think it’s just the…

1346 01:44:13.310 ⇒ 01:44:20.759 Casie Aviles: For the main request handling, I think that’s the one that we need to have on Cloud Run and, like, for the Google Chat API.

1347 01:44:21.240 ⇒ 01:44:25.559 Casie Aviles: Because it’s only one link, if you recall. I think that’s…

1348 01:44:25.560 ⇒ 01:44:26.190 Mustafa Raja: Yeah.

1349 01:44:26.190 ⇒ 01:44:28.109 Casie Aviles: For the entry endpoint.

1350 01:44:33.600 ⇒ 01:44:36.010 Casie Aviles: Well, a lot of these were…

1351 01:44:37.710 ⇒ 01:44:40.880 Casie Aviles: is deploy master service to client.

1352 01:44:41.580 ⇒ 01:44:43.510 Casie Aviles: Is there anything here that.

1353 01:44:43.510 ⇒ 01:44:48.020 Mustafa Raja: Yeah, you can assign that to me, 49… you can assign 49, 50…

1354 01:44:48.620 ⇒ 01:44:49.450 Casie Aviles: Okay.

1355 01:44:52.450 ⇒ 01:44:56.879 Mustafa Raja: Yeah, if I’m taking 50, then I should take 51 also.

1356 01:45:05.740 ⇒ 01:45:06.979 Mustafa Raja: What’s he doing?

1357 01:45:08.520 ⇒ 01:45:10.649 Casie Aviles: I think that’s end-to-end.

1358 01:45:10.840 ⇒ 01:45:22.370 Mustafa Raja: Oh… Oh, update was… oh, okay, okay, okay, I thought that… I thought that was in the… Update ticket.

1359 01:45:22.370 ⇒ 01:45:28.889 Casie Aviles: chat webhook to point to… yeah, it does sound like another duplicate. Update Google Chat webhook to point to…

1360 01:45:29.260 ⇒ 01:45:29.920 Casie Aviles: Hmm.

1361 01:45:31.160 ⇒ 01:45:33.530 Casie Aviles: Oh, okay, so this one is…

1362 01:45:33.950 ⇒ 01:45:39.310 Mustafa Raja: Something here is duplicate, but I guess we could take care of that as we go.

1363 01:45:40.410 ⇒ 01:45:43.300 Casie Aviles: Determine best marks, okay, yeah, D.

1364 01:45:43.520 ⇒ 01:45:49.860 Casie Aviles: Anything else that’s unassigned can… And can we filter?

1365 01:45:52.560 ⇒ 01:45:53.580 Casie Aviles: Hmm…

1366 01:45:54.690 ⇒ 01:45:56.939 Mustafa Raja: Yeah, this is, good then, right?

1367 01:45:58.560 ⇒ 01:46:03.540 Mustafa Raja: The rest of our… the rest of those are milestones.

1368 01:46:05.860 ⇒ 01:46:06.630 Casie Aviles: Yes.

1369 01:46:06.800 ⇒ 01:46:09.699 Casie Aviles: I think we’re… yeah, we’re good, we’re good. Okay.

1370 01:46:10.060 ⇒ 01:46:10.540 Mustafa Raja: Yes.

1371 01:46:10.540 ⇒ 01:46:14.400 Casie Aviles: So… I’ll just… I’ll just send… Kim…

1372 01:46:14.880 ⇒ 01:46:18.470 Casie Aviles: in the channel, I’ll let them know that we’ve assigned.

1373 01:46:19.440 ⇒ 01:46:21.679 Mustafa Raja: So, so in external talent.

1374 01:46:22.530 ⇒ 01:46:23.410 Casie Aviles: Oh, no, no, no.

1375 01:46:23.410 ⇒ 01:46:23.840 Mustafa Raja: Justin.

1376 01:46:23.890 ⇒ 01:46:25.040 Casie Aviles: internal.

1377 01:46:25.210 ⇒ 01:46:27.029 Casie Aviles: Oh, okay. I don’t think…

1378 01:46:27.030 ⇒ 01:46:29.750 Mustafa Raja: Oh, I… I thought you meant Tim.

1379 01:46:30.820 ⇒ 01:46:35.139 Casie Aviles: Yeah, no, no, I mean, no one’s active there, aside from Tim.

1380 01:46:35.260 ⇒ 01:46:38.969 Casie Aviles: It’s just him who’s replying there, I think.

1381 01:46:38.970 ⇒ 01:46:40.040 Mustafa Raja: Yeah, yeah, yeah.

1382 01:46:40.150 ⇒ 01:46:41.280 Mustafa Raja: Yeah. Okay.

1383 01:46:41.630 ⇒ 01:46:42.550 Mustafa Raja: Correct.

1384 01:46:43.450 ⇒ 01:46:44.190 Casie Aviles: on.

1385 01:46:44.840 ⇒ 01:46:49.699 Mustafa Raja: I’m just going to ping them again tomorrow for the service account.

1386 01:46:50.200 ⇒ 01:46:51.830 Casie Aviles: Sure, sure.

1387 01:46:52.370 ⇒ 01:46:55.059 Mustafa Raja: Yeah, he might get frustrated with me.

1388 01:46:55.490 ⇒ 01:46:59.580 Casie Aviles: Yeah, I mean… It does take some time.

1389 01:46:59.710 ⇒ 01:47:01.120 Mustafa Raja: Yeah, I bug a lot.

1390 01:47:04.320 ⇒ 01:47:05.110 Mustafa Raja: I used to do this.

1391 01:47:05.110 ⇒ 01:47:05.850 Casie Aviles: as well.

1392 01:47:05.850 ⇒ 01:47:06.940 Mustafa Raja: Yeah.

1393 01:47:07.100 ⇒ 01:47:13.529 Mustafa Raja: Yeah, I started… so my MacBook is able to run RDR2, you know?

1394 01:47:14.280 ⇒ 01:47:15.680 Casie Aviles: Oh, your MacBook?

1395 01:47:16.120 ⇒ 01:47:16.900 Mustafa Raja: Yeah.

1396 01:47:17.180 ⇒ 01:47:18.780 Casie Aviles: Oh, so now you can play it.

1397 01:47:19.400 ⇒ 01:47:20.420 Mustafa Raja: Yeah.

1398 01:47:20.420 ⇒ 01:47:21.949 Casie Aviles: Nice, yeah.

1399 01:47:22.360 ⇒ 01:47:25.989 Casie Aviles: I do hope you still find the time.

1400 01:47:26.260 ⇒ 01:47:31.319 Casie Aviles: Because that game is… I think it took me, like, almost a month to finish.

1401 01:47:31.660 ⇒ 01:47:39.449 Mustafa Raja: Oh, and yeah, I’m… I thought I could play it for, like, an hour today.

1402 01:47:39.780 ⇒ 01:47:40.630 Mustafa Raja: But…

1403 01:47:41.080 ⇒ 01:47:44.710 Mustafa Raja: I did not… I did not even find 5 minutes.

1404 01:47:44.710 ⇒ 01:47:50.220 Casie Aviles: Yeah, that’s… That’s how I feel, like, for the most part now.

1405 01:47:50.220 ⇒ 01:47:51.450 Mustafa Raja: Yeah.

1406 01:47:51.960 ⇒ 01:47:54.389 Mustafa Raja: Yeah, but it’s good, it’s good.

1407 01:47:54.970 ⇒ 01:47:56.390 Casie Aviles: Yeah, during the.

1408 01:47:56.390 ⇒ 01:48:01.059 Mustafa Raja: Yeah, even… yeah, yeah, yeah. We still have the weekends to us, so…

1409 01:48:04.810 ⇒ 01:48:08.809 Mustafa Raja: Yeah, I’ll… I’ll… I’ll play it then.

1410 01:48:12.560 ⇒ 01:48:13.500 Mustafa Raja: Hey, excellent.

1411 01:48:14.410 ⇒ 01:48:17.840 Casie Aviles: Yeah, yeah, I was just writing. Alright, wait, are you seeing my screen?

1412 01:48:19.190 ⇒ 01:48:22.449 Mustafa Raja: I’m only seeing the instant.

1413 01:48:23.600 ⇒ 01:48:27.230 Casie Aviles: Oh, weird, okay. Yeah, I was writing, I was typing.

1414 01:48:28.010 ⇒ 01:48:29.020 Casie Aviles: On Slack.

1415 01:48:29.020 ⇒ 01:48:29.570 Mustafa Raja: Okay.

1416 01:48:29.570 ⇒ 01:48:35.920 Casie Aviles: So… Yeah, okay, I’ll just, I’ll just swing you there, or… yeah, you’ll see it.

1417 01:48:36.230 ⇒ 01:48:37.820 Casie Aviles: I think that’s all.

1418 01:48:38.470 ⇒ 01:48:41.179 Mustafa Raja: Yeah, I’m going to be logging off now.

1419 01:48:41.610 ⇒ 01:48:43.560 Casie Aviles: Okay, yeah, thank you, thank you very much, Masal.

1420 01:48:43.650 ⇒ 01:48:47.820 Mustafa Raja: No, no, no, no worries, no worries. Also, let me know if, if the…

1421 01:48:48.150 ⇒ 01:48:54.290 Mustafa Raja: deployment works out for you. If not, I’ll be up in the morning, early, and I’ll set it up.

1422 01:48:55.000 ⇒ 01:49:01.829 Mustafa Raja: Just let me know the branch that you would deploy… you’d want to deploy. If you get it going, that’s good. If not, then…

1423 01:49:02.550 ⇒ 01:49:07.210 Mustafa Raja: I need to… I need to also deploy something, so I’ll…

1424 01:49:07.540 ⇒ 01:49:09.520 Casie Aviles: Okay, sure. I’ll be working on that.

1425 01:49:09.620 ⇒ 01:49:13.480 Mustafa Raja: Myself. If you do get that working, let me know.

1426 01:49:13.700 ⇒ 01:49:15.650 Mustafa Raja: Which branch it is?

1427 01:49:16.190 ⇒ 01:49:16.630 Casie Aviles: Trade off.

1428 01:49:16.630 ⇒ 01:49:17.690 Mustafa Raja: Let’s copy that.

1429 01:49:17.870 ⇒ 01:49:23.660 Casie Aviles: I’ll try it out myself first, and then if I really… Okay, thank you, thank you, bye-bye.

1430 01:49:23.870 ⇒ 01:49:24.900 Mustafa Raja: Yeah. Mike.