2026-04-15_abc_working_session_37cb3f9d

Meeting Title: ABC working session Date: 2026-04-15 Meeting participants: Casie Aviles, Samuel Roberts, Pranav Narahari, Mustafa Raja

---

WEBVTT

1 00:01:11.890 ⇒ 00:01:12.900 Samuel Roberts: Hey guys.

2 00:01:13.700 ⇒ 00:01:16.379 Samuel Roberts: Sorry about that, my granola didn’t pop up.

3 00:01:17.600 ⇒ 00:01:18.530 Pranav Narahari: Hey, guys.

4 00:01:22.850 ⇒ 00:01:24.359 Pranav Narahari: Okay.

5 00:01:24.360 ⇒ 00:01:26.659 Samuel Roberts: Talk to me about what’s happening here.

6 00:01:27.880 ⇒ 00:01:35.979 Pranav Narahari: Yeah, so Janiece messaged me while I was kind of on a… on a meeting, and then I told her to message into the external channel.

7 00:01:36.490 ⇒ 00:01:37.860 Pranav Narahari: So…

8 00:01:38.510 ⇒ 00:01:48.570 Pranav Narahari: Yeah, what it sounds like is that there’s, put very simply, that there’s certain data in the zip code DB that is just not being,

9 00:01:48.880 ⇒ 00:01:58.390 Pranav Narahari: that Andy is just not providing when asked. And then it looks like there’s another issue that she mentioned in the external channel that, there’s duplicate…

10 00:01:58.440 ⇒ 00:02:12.540 Pranav Narahari: answers being sent. So that’s all the context that I have. Like, yeah, but the DB is doubling answers. I believe that means it’s, like, sending the answer twice, potentially? .

11 00:02:12.830 ⇒ 00:02:19.469 Samuel Roberts: is… It’s sending the answer twice, or it’s including two? Like, how is it? It’s responding to messages? Okay.

12 00:02:19.950 ⇒ 00:02:22.030 Casie Aviles: Like, for example,

13 00:02:22.390 ⇒ 00:02:28.809 Casie Aviles: I, like, for example, me as an inspector, I have, like, two roles, or, like, two services, so…

14 00:02:29.050 ⇒ 00:02:32.809 Casie Aviles: I would show up on the list, like, twice, so one is for, like.

15 00:02:33.350 ⇒ 00:02:40.630 Casie Aviles: As an estimator, I would have pest, and then the other, I would have, lawn, something, yeah, something like that.

16 00:02:40.910 ⇒ 00:02:42.609 Casie Aviles: I’m just gonna share my screen.

17 00:02:44.460 ⇒ 00:02:45.120 Samuel Roberts: Okay.

18 00:02:50.310 ⇒ 00:02:52.480 Casie Aviles: Alright, do you guys see the screen now?

19 00:02:53.970 ⇒ 00:02:54.650 Samuel Roberts: Yes.

20 00:02:55.090 ⇒ 00:02:56.819 Casie Aviles: Oh, okay, so…

21 00:03:02.830 ⇒ 00:03:09.689 Casie Aviles: Let me just look through, like, the double ones, because for that, I think the fix there would just be…

22 00:03:10.650 ⇒ 00:03:13.729 Casie Aviles: In terms of, like, how Andy is…

23 00:03:14.170 ⇒ 00:03:22.210 Casie Aviles: generating the query, so I think we need… we need to make sure that the query generated, or, like, the output is not, like.

24 00:03:22.750 ⇒ 00:03:26.169 Casie Aviles: Just making sure that it’s just a single record, so…

25 00:03:33.490 ⇒ 00:03:49.299 Pranav Narahari: I see, so the process right now is that, the request is taken in. Using that request, it assesses whether this is a zip code question. Once that, has been established, then a…

26 00:03:49.820 ⇒ 00:03:55.669 Pranav Narahari: a SQL command is then generated to then extract the correct information from the data.

27 00:03:55.670 ⇒ 00:03:56.540 Samuel Roberts: Okay.

28 00:03:56.940 ⇒ 00:03:57.260 Casie Aviles: Yeah.

29 00:03:57.260 ⇒ 00:03:58.000 Pranav Narahari: Is that right?

30 00:03:58.670 ⇒ 00:04:00.780 Pranav Narahari: Okay.

31 00:04:00.780 ⇒ 00:04:01.410 Casie Aviles: So…

32 00:04:01.430 ⇒ 00:04:07.590 Pranav Narahari: What do you think about that technical… Sam, what do you think about that, like, technical approach?

33 00:04:07.870 ⇒ 00:04:15.090 Pranav Narahari: Does that sound like, what’s gonna give us the most accurate data back? .

34 00:04:15.280 ⇒ 00:04:17.100 Samuel Roberts: Hiddly, Hidory dog.

35 00:04:18.180 ⇒ 00:04:23.350 Samuel Roberts: Yeah, I mean, that’s why we had it that way. So, I just want to clarify, can you go back to that,

36 00:04:23.760 ⇒ 00:04:25.409 Samuel Roberts: That ticket that had the double.

37 00:04:25.710 ⇒ 00:04:26.270 Casie Aviles: Yes.

38 00:04:26.270 ⇒ 00:04:31.430 Samuel Roberts: So are you saying that there’s two different entries, and it’s not showing that they’re different here? Is that the issue?

39 00:04:31.430 ⇒ 00:04:44.810 Casie Aviles: Yes, yes, exactly. Okay, okay, listing them as, like, one item each. So if, like, this person is assigned to, like, multiple zips, it’s just gonna show them as, like, two separate, like, records.

40 00:04:44.810 ⇒ 00:04:46.080 Samuel Roberts: Got it, okay.

41 00:04:46.080 ⇒ 00:04:51.990 Casie Aviles: So, kind of like, if you could see it this way, right? So, each record, it’s like…

42 00:04:52.320 ⇒ 00:04:57.430 Casie Aviles: one person, but they have, like, multiple zip codes. So I guess, like…

43 00:04:57.710 ⇒ 00:04:59.950 Casie Aviles: Yeah, that’s how Andy’s returning them.

44 00:04:59.950 ⇒ 00:05:05.509 Samuel Roberts: Can we look at the lawn mowing crew for that zip? Is that, what is it, 78732?

45 00:05:06.300 ⇒ 00:05:08.560 Casie Aviles: 78732.

46 00:05:10.850 ⇒ 00:05:11.699 Samuel Roberts: That’s a reason.

47 00:05:12.830 ⇒ 00:05:20.350 Samuel Roberts: And then, lawn mowing crew… What day of the week?

48 00:05:20.350 ⇒ 00:05:21.739 Casie Aviles: Oh, I think it’s this one, they have an.

49 00:05:21.740 ⇒ 00:05:25.810 Samuel Roberts: Oh, it’s that one, okay, interesting. So that one… okay, so that’s an issue where it… it…

50 00:05:27.460 ⇒ 00:05:32.150 Samuel Roberts: Is that in the database? Is there a note for that one, or is that something that is…

51 00:05:33.060 ⇒ 00:05:35.440 Samuel Roberts: like, as it went to the DB query.

52 00:05:37.500 ⇒ 00:05:40.970 Samuel Roberts: But didn’t actually give anything about the time, the day of the week, or anything.

53 00:05:42.440 ⇒ 00:05:46.010 Casie Aviles: Yes, so… let’s see… So that first.

54 00:05:46.010 ⇒ 00:05:49.740 Samuel Roberts: So the doubling up is that… is that second query there? Okay.

55 00:05:50.790 ⇒ 00:05:56.610 Samuel Roberts: The first query looks… It wasn’t doubling up, okay.

56 00:05:58.120 ⇒ 00:05:59.560 Samuel Roberts: Sorry about that.

57 00:05:59.560 ⇒ 00:06:00.180 Casie Aviles: name.

58 00:06:03.460 ⇒ 00:06:05.839 Casie Aviles: So that must be a cruel.

59 00:06:06.940 ⇒ 00:06:10.789 Casie Aviles: Okay, it’s returning these, so that’s why there’s a double.

60 00:06:11.610 ⇒ 00:06:17.560 Samuel Roberts: Are they completely double… that looks like a double entry. Oh, no, assignment? Why are they two different assignments for the same zip code, though?

61 00:06:19.150 ⇒ 00:06:20.429 Casie Aviles: Oh, okay.

62 00:06:20.970 ⇒ 00:06:22.679 Casie Aviles: Yeah, that’s the same one.

63 00:06:23.510 ⇒ 00:06:24.170 Samuel Roberts: Yeah, I love it.

64 00:06:24.170 ⇒ 00:06:24.600 Casie Aviles: so weekly.

65 00:06:24.600 ⇒ 00:06:25.989 Samuel Roberts: Complete duplicate entry.

66 00:06:26.780 ⇒ 00:06:31.439 Casie Aviles: Okay, yeah, that’s a duplicate entry, so we need to deduplicate this.

67 00:06:32.390 ⇒ 00:06:33.919 Samuel Roberts: to, like, please don’t.

68 00:06:35.160 ⇒ 00:06:35.880 Casie Aviles: already.

69 00:06:37.660 ⇒ 00:06:43.220 Samuel Roberts: And then for that actual question, the day of the week, is that something that we have… in…

70 00:06:43.220 ⇒ 00:06:43.730 Casie Aviles: DB.

71 00:06:43.730 ⇒ 00:06:48.539 Samuel Roberts: at all. I know we added notes and stuff for some things, but is that in there for crews like that?

72 00:06:50.480 ⇒ 00:06:51.819 Casie Aviles: Yeah.

73 00:06:51.820 ⇒ 00:06:52.520 Samuel Roberts: Notes, okay.

74 00:06:52.520 ⇒ 00:06:54.889 Casie Aviles: Just under notes, Tuesday.

75 00:06:54.890 ⇒ 00:06:55.430 Samuel Roberts: Okay.

76 00:06:56.750 ⇒ 00:06:59.170 Samuel Roberts: So,

77 00:06:59.870 ⇒ 00:07:05.839 Samuel Roberts: My thought is that we either need to tweak the queer… well, we need to definitely deduplicate, obviously.

78 00:07:06.040 ⇒ 00:07:10.029 Samuel Roberts: We then either need to… is the…

79 00:07:10.780 ⇒ 00:07:22.590 Samuel Roberts: the note just saying Tuesday seems like it… even if it’s being returned, which I’m not sure if it is, and being passed into the LLM, we might want to make that note,

80 00:07:23.120 ⇒ 00:07:24.679 Samuel Roberts: more verbose.

81 00:07:26.670 ⇒ 00:07:27.300 Casie Aviles: Okay.

82 00:07:28.020 ⇒ 00:07:30.489 Samuel Roberts: Because I think if… if we do a full…

83 00:07:31.070 ⇒ 00:07:35.760 Samuel Roberts: query and get this entry, and the note just says Tuesday.

84 00:07:35.990 ⇒ 00:07:47.339 Samuel Roberts: Unless there’s something in the prompt that has explained the way lawn mowing crews work, my thought is that this note should be where the contacts live about they mow on Tuesdays in the zip code or something.

85 00:07:49.540 ⇒ 00:07:55.660 Casie Aviles: Okay, so essentially we’re going to specify in the instructions that for…

86 00:07:56.030 ⇒ 00:07:58.719 Casie Aviles: Days of the week, we would look at notes.

87 00:07:59.370 ⇒ 00:08:00.150 Casie Aviles: Is that, is that…

88 00:08:00.150 ⇒ 00:08:05.909 Samuel Roberts: Well, I’m thinking we just update the notes to say this crew

89 00:08:06.570 ⇒ 00:08:10.750 Samuel Roberts: only Moe’s on Tuesday, rather than just the word Tuesday.

90 00:08:10.750 ⇒ 00:08:12.490 Casie Aviles: I see, okay, so be, yeah.

91 00:08:12.490 ⇒ 00:08:13.850 Samuel Roberts: Wait, we don’t have to update?

92 00:08:13.970 ⇒ 00:08:14.660 Samuel Roberts: Go ahead, President.

93 00:08:14.660 ⇒ 00:08:15.760 Pranav Narahari: I…

94 00:08:16.100 ⇒ 00:08:23.279 Pranav Narahari: I’m not sure… are we sure that would work? Because it looks like it’s just not looking at the notes.

95 00:08:24.120 ⇒ 00:08:28.419 Samuel Roberts: Well, that’s what I’m wondering, it either… it might have looked at the note and not done anything with it.

96 00:08:28.730 ⇒ 00:08:31.929 Samuel Roberts: That’s why I want to see, let’s see what it actually returns and everything.

97 00:08:32.250 ⇒ 00:08:37.100 Samuel Roberts: So if the note is included and it didn’t,

98 00:08:42.809 ⇒ 00:08:46.709 Samuel Roberts: Like, is the word Tuesday anywhere in this page? Let’s do a controller for that.

99 00:08:48.360 ⇒ 00:08:49.380 Samuel Roberts: Okay, so it’s…

100 00:08:49.850 ⇒ 00:08:53.989 Samuel Roberts: Doesn’t look like it’s even returning the note, so definitely we need to make sure that that’s part of the query.

101 00:08:54.210 ⇒ 00:08:57.530 Samuel Roberts: Okay, it has notes there.

102 00:08:58.610 ⇒ 00:08:59.590 Casie Aviles: Yeah.

103 00:09:02.010 ⇒ 00:09:05.780 Samuel Roberts: Dev assignments, join dev crew…

104 00:09:06.340 ⇒ 00:09:09.690 Samuel Roberts: Did Dev Crew get the notes? Crew name, service name?

105 00:09:10.090 ⇒ 00:09:15.789 Samuel Roberts: A.notes from dev assignments A, join dev… okay, yeah.

106 00:09:16.220 ⇒ 00:09:21.039 Samuel Roberts: But it didn’t… or do we not see in this output the actual return of that?

107 00:09:21.040 ⇒ 00:09:24.849 Casie Aviles: Yeah, I don’t think we see it here. Just the generated SQL.

108 00:09:25.610 ⇒ 00:09:27.700 Samuel Roberts: Oh, we don’t see the return of the tool?

109 00:09:28.520 ⇒ 00:09:31.230 Casie Aviles: Yeah, for, for, like, the sequel.

110 00:09:31.680 ⇒ 00:09:35.670 Casie Aviles: Or, like, yeah, the return output of the SQL query.

111 00:09:35.670 ⇒ 00:09:36.250 Samuel Roberts: Okay.

112 00:09:36.250 ⇒ 00:09:38.230 Casie Aviles: just the generated SQL.

113 00:09:38.600 ⇒ 00:09:42.569 Samuel Roberts: Okay, so it should have at least returned the notes, and the notes might have just said Tuesday…

114 00:09:43.450 ⇒ 00:09:45.530 Samuel Roberts: And it didn’t… Do anything.

115 00:09:45.530 ⇒ 00:09:48.489 Pranav Narahari: Can we confirm, can we just, like, run that?

116 00:09:48.740 ⇒ 00:09:50.230 Samuel Roberts: Yeah, run that query, definitely.

117 00:09:50.230 ⇒ 00:09:50.860 Pranav Narahari: Got it.

118 00:09:52.020 ⇒ 00:09:52.630 Casie Aviles: Okay.

119 00:10:17.890 ⇒ 00:10:18.560 Casie Aviles: Done.

120 00:10:37.060 ⇒ 00:10:39.190 Samuel Roberts: Is your 1Password password in Bitwarden?

121 00:10:39.380 ⇒ 00:10:40.350 Casie Aviles: Yeah…

122 00:10:40.680 ⇒ 00:10:43.339 Samuel Roberts: Same, same. I understand that flow. Yeah.

123 00:10:43.870 ⇒ 00:10:46.130 Pranav Narahari: I can pull the password for you.

124 00:10:48.750 ⇒ 00:10:51.459 Samuel Roberts: No, I mean, once he gets into 1Password, he’ll be fine, he just needs to…

125 00:11:02.680 ⇒ 00:11:06.909 Casie Aviles: Yeah, because I wasn’t using 1Password when… when I first joined.

126 00:11:08.580 ⇒ 00:11:09.550 Pranav Narahari: Okay, cool.

127 00:11:24.340 ⇒ 00:11:25.010 Casie Aviles: See?

128 00:11:25.680 ⇒ 00:11:27.250 Casie Aviles: Not getting the code.

129 00:11:46.920 ⇒ 00:11:51.760 Casie Aviles: But, yeah, I’ve collected the other issues here, so…

130 00:11:52.240 ⇒ 00:11:52.910 Samuel Roberts: Okay.

131 00:12:01.890 ⇒ 00:12:02.590 Samuel Roberts: Cool.

132 00:12:04.060 ⇒ 00:12:06.260 Samuel Roberts: Oh, no, you want to put that in 1Password, right? Yeah.

133 00:12:12.630 ⇒ 00:12:13.530 Samuel Roberts: Yeah, cool.

134 00:12:14.210 ⇒ 00:12:14.880 Pranav Narahari: Nice.

135 00:12:16.980 ⇒ 00:12:21.470 Samuel Roberts: Yeah, I have the same flow every time I, like, reboot the computer or reopen 1Password or something, and it’s…

136 00:12:21.820 ⇒ 00:12:22.930 Pranav Narahari: Same for me.

137 00:12:23.290 ⇒ 00:12:23.930 Samuel Roberts: Yeah.

138 00:12:38.590 ⇒ 00:12:43.219 Casie Aviles: Let me just close some… Windows, I have a lot of windows.

139 00:12:43.220 ⇒ 00:12:43.990 Samuel Roberts: Okay.

140 00:12:45.870 ⇒ 00:12:47.680 Samuel Roberts: But yeah, that’s…

141 00:12:52.770 ⇒ 00:12:53.670 Casie Aviles: Okay.

142 00:13:03.290 ⇒ 00:13:04.829 Samuel Roberts: We go out right now.

143 00:13:05.760 ⇒ 00:13:08.149 Samuel Roberts: Oh, my goodness.

144 00:13:10.170 ⇒ 00:13:11.000 Samuel Roberts: Okay.

145 00:13:17.810 ⇒ 00:13:18.820 Samuel Roberts: Oh, yeah.

146 00:13:22.780 ⇒ 00:13:24.499 Samuel Roberts: Google does not make it easy.

147 00:13:26.310 ⇒ 00:13:27.200 Samuel Roberts: Love it.

148 00:13:53.520 ⇒ 00:13:54.190 Casie Aviles: Sure.

149 00:14:02.090 ⇒ 00:14:03.800 Samuel Roberts: Is it below it, or…

150 00:14:04.030 ⇒ 00:14:05.050 Samuel Roberts: Did it run?

151 00:14:08.460 ⇒ 00:14:11.079 Casie Aviles: Yeah, I have to clean up the… the new lines.

152 00:14:11.080 ⇒ 00:14:14.399 Samuel Roberts: Oh, I assumed it would have handled the new lines, that’s… okay.

153 00:14:16.120 ⇒ 00:14:21.979 Samuel Roberts: Sorry about that. I just assumed it would be… Okay. Cool, perfect.

154 00:14:22.250 ⇒ 00:14:23.829 Samuel Roberts: Yeah, the job. Okay.

155 00:14:25.020 ⇒ 00:14:28.709 Samuel Roberts: Yeah, so the notes aren’t getting processed, definitely, okay.

156 00:14:29.940 ⇒ 00:14:30.640 Pranav Narahari: Okay.

157 00:14:31.790 ⇒ 00:14:36.209 Samuel Roberts: Yeah, I think the… I think the cleanest thing is to make those notes more explicit.

158 00:14:37.490 ⇒ 00:14:40.880 Pranav Narahari: Yeah, well… May… so…

159 00:14:41.280 ⇒ 00:14:51.110 Pranav Narahari: Well, what is, like, the patch for right now? I think it’s going into the system prompt, and then just updating that to say, notes contain…

160 00:14:51.480 ⇒ 00:14:55.300 Pranav Narahari: day of the week assignments, right? Or, what is the right verbiage?

161 00:14:55.620 ⇒ 00:14:57.419 Samuel Roberts: Do notes contain anything else, though?

162 00:14:58.480 ⇒ 00:14:59.240 Pranav Narahari: Or you can say.

163 00:14:59.240 ⇒ 00:14:59.729 Samuel Roberts: For the vehicle.

164 00:15:00.410 ⇒ 00:15:00.990 Pranav Narahari: Like…

165 00:15:00.990 ⇒ 00:15:05.959 Samuel Roberts: So I just want to make sure we know, like, what… because I don’t know if these notes are being used for other things, other places, besides days of the Week.

166 00:15:06.730 ⇒ 00:15:07.170 Casie Aviles: Yeah, like…

167 00:15:07.170 ⇒ 00:15:13.059 Pranav Narahari: We can say, like, so it’s not going to break anything else, we can say notes can contain this information.

168 00:15:13.350 ⇒ 00:15:19.270 Pranav Narahari: So it’s not just… well, it’s not phrased as only contains.

169 00:15:21.210 ⇒ 00:15:24.539 Samuel Roberts: Yeah, I mean, I think that that is a way to do it. I don’t…

170 00:15:25.470 ⇒ 00:15:34.579 Samuel Roberts: I don’t necessarily see why that’s superior to just changing the notes so that it only adds, like, it adds that into the context when we get it, rather than every single time.

171 00:15:35.000 ⇒ 00:15:37.200 Pranav Narahari: Oh, are you saying that we do that right now?

172 00:15:38.520 ⇒ 00:15:47.849 Samuel Roberts: I was saying, like, just update that, like, note, maybe, and see if it… like, if we can change that to, this crew mows this zip code on…

173 00:15:48.850 ⇒ 00:15:52.999 Samuel Roberts: But you’re right, if you want to just do something quick, the system prompt is probably good, but I think a better…

174 00:15:53.380 ⇒ 00:15:54.220 Pranav Narahari: Well, I think the.

175 00:15:54.220 ⇒ 00:15:54.680 Samuel Roberts: Oh, right?

176 00:15:54.680 ⇒ 00:16:00.879 Pranav Narahari: The solution here is to not just have this in the notes anyways, like, we need to create a new field for day of the week assignments.

177 00:16:01.150 ⇒ 00:16:08.809 Pranav Narahari: Right, because like you said, notes can have other information, so since we know that…

178 00:16:10.860 ⇒ 00:16:15.030 Samuel Roberts: Oh, I see what you’re saying. Okay, so just ad hoc fields when we need them?

179 00:16:15.980 ⇒ 00:16:23.309 Samuel Roberts: But I feel like if the notes is something more verbose, then it can be a more useful field across multiple

180 00:16:24.520 ⇒ 00:16:27.040 Samuel Roberts: Crews or assignments, or anything like that.

181 00:16:27.820 ⇒ 00:16:37.379 Pranav Narahari: Yeah, but then it’s also susceptible to things like this. Like, they’re just gonna put lack of context, and then the AI’s not gonna be able to pull from it, clearly.

182 00:16:38.540 ⇒ 00:16:45.390 Samuel Roberts: Right, so then the issue becomes we need to make sure they know how to use the notes, rather than patching it every time something gets added to the notes that isn’t clear.

183 00:16:49.430 ⇒ 00:16:56.300 Pranav Narahari: I think the idea of having it structured this way is that we’re providing them the field so then they can easily insert it, right?

184 00:16:57.170 ⇒ 00:17:06.289 Pranav Narahari: Because, like, the reason we have, like, zip code 2 and service name is, like, so they don’t have to just, like, write it manually in the notes, because then that’s… they might put in the wrong.

185 00:17:06.290 ⇒ 00:17:09.899 Samuel Roberts: Right, but those are fields that match across. I feel like for the notes.

186 00:17:10.020 ⇒ 00:17:15.030 Samuel Roberts: Like, there might be other things that are important for these crews that will also go into a notes field.

187 00:17:16.280 ⇒ 00:17:18.330 Samuel Roberts: Or other, other entities.

188 00:17:18.339 ⇒ 00:17:22.999 Pranav Narahari: You probably keep a notes field still. I think we just create a new field that says…

189 00:17:23.589 ⇒ 00:17:24.579 Pranav Narahari: You know, days of the week.

190 00:17:24.579 ⇒ 00:17:26.179 Samuel Roberts: Right, what I’m worried about is…

191 00:17:26.369 ⇒ 00:17:30.409 Samuel Roberts: Constantly adding ad hoc fields to… various…

192 00:17:30.689 ⇒ 00:17:35.009 Samuel Roberts: Tables that are necessary for the crews, or for the departments, or for the inspect… like…

193 00:17:35.229 ⇒ 00:17:39.079 Samuel Roberts: My thought is the notes getting passed to the LLM should be enough for it to know

194 00:17:39.349 ⇒ 00:17:40.499 Samuel Roberts: what to do.

195 00:17:45.120 ⇒ 00:17:48.100 Pranav Narahari: Like… Yeah.

196 00:17:48.280 ⇒ 00:18:05.210 Samuel Roberts: I just feel like if it sets up, like, a… you know, zips make sense, because everyone has a zip code up there. Like, that all seems to, like, is the structure of the data. The notes, I think, is for more, oh, this specific one needs this piece of information that is not applicable across all

197 00:18:05.760 ⇒ 00:18:08.909 Samuel Roberts: All entries in this database.

198 00:18:09.580 ⇒ 00:18:16.259 Pranav Narahari: Okay. So, yeah, I see what you’re saying. What we can do then is… since we do have a…

199 00:18:16.970 ⇒ 00:18:21.860 Pranav Narahari: set number of fields, and I think, sam, you and Utami

200 00:18:22.290 ⇒ 00:18:25.109 Pranav Narahari: a little bit about, like, a long-term fix for this.

201 00:18:25.110 ⇒ 00:18:25.720 Samuel Roberts: That’s undesirable.

202 00:18:25.850 ⇒ 00:18:38.509 Pranav Narahari: make sense to support additional fields being created. I think that’s what’s gonna create the most deterministic outcome, having additional fields, but you and Utam can decide on that. What we should do for… The issue with that is…

203 00:18:38.510 ⇒ 00:18:39.000 Samuel Roberts: Yep.

204 00:18:39.000 ⇒ 00:18:49.690 Pranav Narahari: What we should… like, we need to patch this, like, right now, right? So, I think the patch for right now is probably update to the system prompt. Like, that seems like the quickest fix. So we can just message Denise saying, hey, this is fixed now.

205 00:18:49.690 ⇒ 00:18:50.300 Samuel Roberts: Amazing.

206 00:18:50.420 ⇒ 00:18:50.910 Samuel Roberts: Branch?

207 00:18:50.910 ⇒ 00:18:51.690 Pranav Narahari: Do you agree?

208 00:18:53.070 ⇒ 00:18:59.849 Samuel Roberts: If it works, yeah, as long as it handles that well. I think we have to test that and make sure, definitely, but.

209 00:19:00.860 ⇒ 00:19:03.680 Pranav Narahari: Or… okay, then another option…

210 00:19:03.680 ⇒ 00:19:04.360 Samuel Roberts: You could also just…

211 00:19:05.010 ⇒ 00:19:15.200 Pranav Narahari: We can have a script go out to all of the records in the database right now, check the notes to see if there is a day of the week, and then update the…

212 00:19:15.440 ⇒ 00:19:23.770 Pranav Narahari: Then the notes in there to, you know, to update that string from being just a day of the week to being…

213 00:19:24.110 ⇒ 00:19:25.950 Pranav Narahari: You know, whatever string that you think is.

214 00:19:25.950 ⇒ 00:19:28.860 Samuel Roberts: better. That’s what I was gonna suggest, yes. I think, I think…

215 00:19:29.010 ⇒ 00:19:32.660 Samuel Roberts: Is it… so, are lawn crews…

216 00:19:33.800 ⇒ 00:19:37.300 Samuel Roberts: the only ones that have this day of the week limitation? Do we know that?

217 00:19:39.560 ⇒ 00:19:42.000 Casie Aviles: For, like, day of the week.

218 00:19:42.000 ⇒ 00:19:47.430 Samuel Roberts: like, the… I just know we have… since we have notes, and it’s… it’s… I want to make sure…

219 00:19:47.670 ⇒ 00:19:50.769 Samuel Roberts: We’re updating it across everything that needs it, because, like…

220 00:19:51.580 ⇒ 00:19:58.739 Samuel Roberts: Are there other… like, I’m wondering if there are any other notes fields that would not be this, but I think… I think for now, if we just update the ones that say just day of the week, too.

221 00:19:59.010 ⇒ 00:20:01.109 Samuel Roberts: Something that would be the best fix.

222 00:20:01.110 ⇒ 00:20:12.239 Casie Aviles: Yeah, there are… I’ve inputted, like, other free text in the notes, so it’s not limited to just days of the week. Like, it’s kind of like a catch-all for additional context that…

223 00:20:12.240 ⇒ 00:20:12.690 Samuel Roberts: Right.

224 00:20:12.690 ⇒ 00:20:15.650 Casie Aviles: assignment may have… that’s how it’s working right now.

225 00:20:16.210 ⇒ 00:20:25.619 Samuel Roberts: Yeah, and I think that’s fine. I just… yeah, so when we… if we write a script to fetch all those, check if it matches against a list of days of the week, and then update it.

226 00:20:26.610 ⇒ 00:20:27.720 Pranav Narahari: I think that’s the…

227 00:20:27.720 ⇒ 00:20:29.379 Samuel Roberts: The cleanest, like, most…

228 00:20:29.380 ⇒ 00:20:35.379 Pranav Narahari: Also, just do, like, a proof of concept of that. Casey, can you manually update this notes field to…

229 00:20:35.380 ⇒ 00:20:38.279 Samuel Roberts: That’s exactly what I was gonna say, yeah, test this against that.

230 00:20:38.930 ⇒ 00:20:43.650 Casie Aviles: Okay, so, what would be the change here?

231 00:20:45.720 ⇒ 00:20:49.099 Samuel Roberts: So the question was exactly, how did they phrase it?

232 00:20:49.830 ⇒ 00:20:51.840 Casie Aviles: Oh, C.

233 00:20:53.260 ⇒ 00:20:55.839 Casie Aviles: What day of the week does the crew go?

234 00:20:56.590 ⇒ 00:21:04.769 Samuel Roberts: Okay, so this is specifically a lawnmower, so I would say the day of the week that this crew, yeah, or something like that, day of the week this crew knows the zip code is, something like that.

235 00:21:09.130 ⇒ 00:21:15.790 Pranav Narahari: Or instead of… to make it more global, this service is provided? Or do you want to say this crew mows?

236 00:21:16.130 ⇒ 00:21:19.019 Samuel Roberts: I mean, this is specifically a crew role and a crew person.

237 00:21:19.020 ⇒ 00:21:25.640 Pranav Narahari: Well, if… what are we gonna do, like, say if this goes beyond just a lawn mowing service, are you gonna put a unique…

238 00:21:25.980 ⇒ 00:21:27.050 Pranav Narahari: Note.

239 00:21:27.170 ⇒ 00:21:28.329 Pranav Narahari: for each one.

240 00:21:30.340 ⇒ 00:21:33.179 Samuel Roberts: Yeah, I think that’s the only way that we can make sure the data’s clean.

241 00:21:34.640 ⇒ 00:21:37.440 Samuel Roberts: I mean, we can make it more generic if you want, but I think…

242 00:21:38.770 ⇒ 00:21:39.150 Pranav Narahari: Excuse me.

243 00:21:39.150 ⇒ 00:21:40.029 Samuel Roberts: Yeah, I mean, we could…

244 00:21:40.030 ⇒ 00:21:44.830 Pranav Narahari: this proof of concept should just be based on, like, what patch are we gonna push right after, right? Yeah.

245 00:21:44.830 ⇒ 00:21:50.330 Samuel Roberts: If I run this and see if that fixes it, then yeah. Probably just delete this other one, right?

246 00:21:51.120 ⇒ 00:21:53.080 Casie Aviles: Oh, right, yeah. Oh.

247 00:21:53.080 ⇒ 00:21:53.909 Samuel Roberts: If it’s our kids.

248 00:21:53.910 ⇒ 00:21:54.320 Casie Aviles: identical.

249 00:21:54.320 ⇒ 00:21:55.990 Samuel Roberts: It’s just a duplicate assignment.

250 00:21:55.990 ⇒ 00:21:57.419 Pranav Narahari: Oh, but they have different days.

251 00:21:57.420 ⇒ 00:22:02.660 Casie Aviles: Yeah, they had different days, so one is Tuesday, and the other is Thursday. So that’s.

252 00:22:02.870 ⇒ 00:22:11.170 Samuel Roberts: Okay, so it’s showing up in this, in this row. Got it. Okay, in that case, definitely, yes, we just want to change this one to say the same thing, and then test that out.

253 00:22:11.570 ⇒ 00:22:18.350 Samuel Roberts: Okay, so the notes also need to be displayed on the admin UI, potentially? Or a way to view them in the admin UI?

254 00:22:19.060 ⇒ 00:22:20.759 Casie Aviles: Yeah, not… right now, there’s…

255 00:22:20.760 ⇒ 00:22:29.789 Pranav Narahari: probably… shouldn’t we… like, the… what this should actually look like in the DB is just one record with two days of the week in the know, right?

256 00:22:32.990 ⇒ 00:22:33.630 Samuel Roberts: Yeah, I think…

257 00:22:34.000 ⇒ 00:22:38.299 Samuel Roberts: I don’t know how this… what got added here, but I think that’s probably a better solution, too.

258 00:22:38.570 ⇒ 00:22:43.339 Samuel Roberts: Okay, sure. Unless we want to add the day of the week field, in which case that’s… Different, yeah.

259 00:22:43.660 ⇒ 00:22:45.789 Casie Aviles: Okay, so I’ll just delete one of them.

260 00:22:46.520 ⇒ 00:22:49.409 Samuel Roberts: And then update the other note to, say, Tuesday, Thursday, or something.

261 00:22:52.420 ⇒ 00:22:57.839 Pranav Narahari: Yeah, so there’s likely not duplicates then, it’s just records that need to be consolidated.

262 00:22:58.240 ⇒ 00:22:58.960 Samuel Roberts: Yes.

263 00:23:03.370 ⇒ 00:23:04.060 Casie Aviles: Okay.

264 00:23:06.360 ⇒ 00:23:11.590 Samuel Roberts: And when was he assigned? Or what was the… is there a creation date that we’re seeing here, or is that just in the…

265 00:23:12.120 ⇒ 00:23:13.130 Samuel Roberts: the DD.

266 00:23:14.130 ⇒ 00:23:16.700 Casie Aviles: This is the assigned date, so…

267 00:23:17.020 ⇒ 00:23:18.240 Samuel Roberts: That’s for them, though.

268 00:23:18.700 ⇒ 00:23:20.310 Samuel Roberts: Is that when it got created?

269 00:23:21.090 ⇒ 00:23:22.579 Casie Aviles: Yeah, yeah, it’s when it got free.

270 00:23:22.580 ⇒ 00:23:23.180 Samuel Roberts: Okay.

271 00:23:23.620 ⇒ 00:23:25.869 Samuel Roberts: Okay, they were both created at the same document. Just making sure, so…

272 00:23:25.970 ⇒ 00:23:29.960 Samuel Roberts: Whenever those got put in, they got put in two separate ones with different notes. Okay.

273 00:23:32.260 ⇒ 00:23:33.719 Casie Aviles: Okay, so I’ll test, like…

274 00:23:45.810 ⇒ 00:23:49.460 Samuel Roberts: Now, we didn’t update the other one, so it’ll be interesting to see how this handles that, but…

275 00:23:58.920 ⇒ 00:23:59.989 Samuel Roberts: Did that even…

276 00:24:04.490 ⇒ 00:24:08.689 Samuel Roberts: What is… No ongoing maintenance on Wednesdays?

277 00:24:10.010 ⇒ 00:24:12.399 Samuel Roberts: Where… is that in the central dock or something?

278 00:24:14.020 ⇒ 00:24:16.890 Casie Aviles: Yeah, that could be another thing, the routing.

279 00:24:19.130 ⇒ 00:24:25.729 Casie Aviles: So it could… it may look… it may be looking at, like, just central dock, and then the other is…

280 00:24:26.700 ⇒ 00:24:32.439 Casie Aviles: It will call, like, the vector agent, so I’m going to check the trace here.

281 00:24:32.440 ⇒ 00:24:33.520 Samuel Roberts: Yeah, definitely.

282 00:24:37.950 ⇒ 00:24:38.720 Casie Aviles: Hmm.

283 00:24:39.180 ⇒ 00:24:41.890 Casie Aviles: It did execute with the DB agent.

284 00:24:42.610 ⇒ 00:24:43.400 Samuel Roberts: Okay.

285 00:24:43.920 ⇒ 00:24:45.979 Samuel Roberts: And now run that query again.

286 00:24:54.370 ⇒ 00:24:55.190 Samuel Roberts: Oh, yeah.

287 00:25:00.180 ⇒ 00:25:03.380 Samuel Roberts: Did it generate the exact same query, actually? That’s a good question, too.

288 00:25:13.300 ⇒ 00:25:16.039 Samuel Roberts: Did it select notes there? It did not select notes.

289 00:25:17.820 ⇒ 00:25:19.479 Pranav Narahari: Looks like it did, right?

290 00:25:19.480 ⇒ 00:25:23.789 Samuel Roberts: Oh, C notes, sorry, I was looking for the NA notes, yeah, I generated it slightly differently, my apologies.

291 00:25:25.690 ⇒ 00:25:27.629 Casie Aviles: Yeah, it was slightly different.

292 00:25:28.140 ⇒ 00:25:32.910 Samuel Roberts: Yeah, that’s fine, as long as it got it. I just wanna… I was reading it the wrong way, I was going from the bottom up, looking for the notes that it matched.

293 00:25:33.990 ⇒ 00:25:36.169 Casie Aviles: Oh, so it’s reading these now.

294 00:25:39.560 ⇒ 00:25:44.890 Samuel Roberts: So where’s this other… note… Coming from?

295 00:25:56.220 ⇒ 00:25:58.120 Casie Aviles: Chicken.

296 00:25:58.260 ⇒ 00:25:59.030 Casie Aviles: Let’s see…

297 00:26:00.880 ⇒ 00:26:03.179 Samuel Roberts: Because that is not what we saw with the other query, was it?

298 00:26:03.530 ⇒ 00:26:06.000 Casie Aviles: Yeah, we got Regan.

299 00:26:07.550 ⇒ 00:26:08.360 Samuel Roberts: And those do look like.

300 00:26:08.360 ⇒ 00:26:08.860 Casie Aviles: here at the.

301 00:26:08.860 ⇒ 00:26:11.409 Samuel Roberts: Duplicates, because the notes are even the same.

302 00:26:20.270 ⇒ 00:26:21.889 Casie Aviles: Where’s that coming from?

303 00:26:22.870 ⇒ 00:26:28.070 Casie Aviles: no initial on… Going maintenance on Wednesdays.

304 00:26:42.340 ⇒ 00:26:45.270 Pranav Narahari: Can you, search just on the crew name again?

305 00:26:46.360 ⇒ 00:26:48.499 Pranav Narahari: And I can actually do this on my end, too.

306 00:26:48.860 ⇒ 00:26:52.470 Pranav Narahari: Can you send me this DB link again?

307 00:26:53.600 ⇒ 00:26:54.620 Casie Aviles: I’m sure.

308 00:26:57.110 ⇒ 00:27:02.820 Casie Aviles: I’m sending it here… Yeah, and it’s in the chat.

309 00:27:03.410 ⇒ 00:27:04.320 Pranav Narahari: Perfect, thank you.

310 00:27:13.690 ⇒ 00:27:16.489 Pranav Narahari: And it’s admin, and then password is what?

311 00:27:16.980 ⇒ 00:27:22.069 Casie Aviles: Abc Home and Commercial 2026.

312 00:27:36.250 ⇒ 00:27:37.880 Samuel Roberts: Which was the zip we’re looking at again?

313 00:27:38.900 ⇒ 00:27:39.850 Pranav Narahari: 802.

314 00:27:40.800 ⇒ 00:27:41.950 Pranav Narahari: I believe, right?

315 00:27:41.950 ⇒ 00:27:43.710 Casie Aviles: 77802.

316 00:27:44.450 ⇒ 00:27:46.600 Samuel Roberts: Okay, that’s the one we updated.

317 00:27:47.230 ⇒ 00:27:51.270 Samuel Roberts: All the other ones just have days of the week for that guy.

318 00:27:54.320 ⇒ 00:27:55.529 Casie Aviles: Oh, I think it’s the.

319 00:27:55.530 ⇒ 00:27:57.179 Samuel Roberts: Oh, the crew hasn’t moved.

320 00:27:57.510 ⇒ 00:27:59.919 Samuel Roberts: The… yep, I see it. Okay.

321 00:28:00.530 ⇒ 00:28:01.110 Casie Aviles: Because it’s…

322 00:28:01.110 ⇒ 00:28:03.600 Pranav Narahari: Oh, right, that’s on C.Notes, yeah, okay.

323 00:28:03.600 ⇒ 00:28:06.370 Samuel Roberts: Yeah, so we… yeah, that’s why I was… it was different.

324 00:28:06.810 ⇒ 00:28:10.010 Samuel Roberts: So we want to make sure to select notes from the crew and from the…

325 00:28:10.320 ⇒ 00:28:14.470 Samuel Roberts: Assignment, or the, yeah.

326 00:28:18.170 ⇒ 00:28:22.260 Casie Aviles: Yeah, it’s this one that Andy’s reading from.

327 00:28:23.110 ⇒ 00:28:28.689 Samuel Roberts: So, the notes on the crew and the notes on the assignment.

328 00:28:28.890 ⇒ 00:28:31.739 Samuel Roberts: So the crew applies to all the zips, right?

329 00:28:32.700 ⇒ 00:28:33.520 Casie Aviles: Yeah.

330 00:28:34.540 ⇒ 00:28:40.970 Samuel Roberts: Okay, and then the other ones are specific to Zips, so we still want both, we just need to make sure the prompt knows to get notes from both crew and assignments.

331 00:28:42.170 ⇒ 00:28:42.650 Casie Aviles: No question.

332 00:28:42.650 ⇒ 00:28:43.270 Samuel Roberts: Okay.

333 00:28:43.390 ⇒ 00:28:45.189 Samuel Roberts: That’s a prompt update there.

334 00:29:04.010 ⇒ 00:29:10.519 Casie Aviles: Okay, let’s see… yeah, same thing here…

335 00:29:20.690 ⇒ 00:29:22.270 Casie Aviles: Let me just clean this up.

336 00:30:06.460 ⇒ 00:30:10.830 Casie Aviles: Should we go through, like, each of the tickets still?

337 00:30:12.490 ⇒ 00:30:17.699 Samuel Roberts: If they’re… are these similar doubling up? What is the relation of all the tickets?

338 00:30:19.870 ⇒ 00:30:25.080 Casie Aviles: Yeah, there are some doubling here, let me just get them.

339 00:30:25.680 ⇒ 00:30:27.739 Samuel Roberts: So it’s probably a similar thing, where it’s,

340 00:30:28.130 ⇒ 00:30:30.420 Samuel Roberts: Not recognizing the difference in the notes.

341 00:30:30.420 ⇒ 00:30:33.350 Casie Aviles: Okay, so this one is doubling as well, so…

342 00:30:33.350 ⇒ 00:30:34.000 Samuel Roberts: Okay.

343 00:30:38.230 ⇒ 00:30:40.160 Casie Aviles: Who said that I steal you?

344 00:31:09.880 ⇒ 00:31:11.830 Casie Aviles: Boy, this is a central dock.

345 00:31:12.400 ⇒ 00:31:13.000 Samuel Roberts: Yeah.

346 00:31:13.640 ⇒ 00:31:15.090 Casie Aviles: Central dot issue.

347 00:31:22.860 ⇒ 00:31:27.620 Samuel Roberts: Okay, so I’m gonna try to get a script together to update those notes if they contain just a day of the week.

348 00:31:35.420 ⇒ 00:31:37.420 Samuel Roberts: let’s see…

349 00:31:45.070 ⇒ 00:31:48.919 Pranav Narahari: Yes, and I would also just be careful with, like, updating

350 00:31:49.850 ⇒ 00:31:55.319 Pranav Narahari: Like, I think that script should also just check to make sure that the day of the week is the only…

351 00:31:56.750 ⇒ 00:32:02.139 Pranav Narahari: only information in there, because like we just saw, the client, or with.

352 00:32:02.400 ⇒ 00:32:04.639 Samuel Roberts: Yeah, it would have to match exactly to… yeah.

353 00:32:04.640 ⇒ 00:32:05.420 Pranav Narahari: Yeah.

354 00:32:06.600 ⇒ 00:32:11.669 Pranav Narahari: And I think it’s still an open question with, what do we want to add there?

355 00:32:12.240 ⇒ 00:32:14.780 Pranav Narahari: How are you going to, with the script.

356 00:32:15.080 ⇒ 00:32:18.909 Pranav Narahari: Pull in the, like, the specific service that they’re a part of.

357 00:32:20.920 ⇒ 00:32:29.559 Samuel Roberts: Alright, snack time! I mean, the crew, the service, the role, the department, are all part of those entries.

358 00:32:30.120 ⇒ 00:32:33.329 Pranav Narahari: Okay, so you’ll pull that, and then create a dynamic…

359 00:32:33.460 ⇒ 00:32:35.830 Pranav Narahari: String to append to the notes section.

360 00:32:37.030 ⇒ 00:32:43.260 Samuel Roberts: Yeah, I mean, I’m gonna do a quick pass and see which ones have these dates and stuff, and if it’s all just the same, it might be easy to just…

361 00:32:44.630 ⇒ 00:32:46.470 Pranav Narahari: Sure, yeah, if it’s all just lawn, then.

362 00:32:46.470 ⇒ 00:32:47.020 Samuel Roberts: Yeah.

363 00:32:47.020 ⇒ 00:32:48.580 Pranav Narahari: You can just hardcode it, but…

364 00:32:48.580 ⇒ 00:32:50.660 Samuel Roberts: Yeah, I’m gonna take a look and see what I can…

365 00:33:03.330 ⇒ 00:33:04.880 Pranav Narahari: Okay,

366 00:33:18.180 ⇒ 00:33:25.049 Pranav Narahari: Another question I got from Janiece was, template for cancellations?

367 00:33:26.780 ⇒ 00:33:29.200 Pranav Narahari: Is that in the central dock?

368 00:33:30.960 ⇒ 00:33:33.859 Casie Aviles: I believe the template for cancellations is…

369 00:33:34.180 ⇒ 00:33:38.940 Casie Aviles: From the verbatim agent, so that should be, like, pulling… the exact.

370 00:33:39.400 ⇒ 00:33:41.109 Casie Aviles: Templates that we have.

371 00:33:42.510 ⇒ 00:33:51.629 Casie Aviles: So, it’s in a different location outside of the central dock, and I think maybe Mustafa knows more about, like, how that was set up, but…

372 00:33:52.750 ⇒ 00:33:58.779 Casie Aviles: Yeah, if I remember correctly, that’s how we did it, since we wanted to have, like, the exact template.

373 00:33:59.880 ⇒ 00:34:00.610 Pranav Narahari: Gotcha.

374 00:34:03.410 ⇒ 00:34:07.039 Pranav Narahari: Mustafa, do you know where that is? Because it looks… it sounds like.

375 00:34:07.040 ⇒ 00:34:08.360 Mustafa Raja: What are we talking about.

376 00:34:08.360 ⇒ 00:34:09.090 Pranav Narahari: Yeah.

377 00:34:09.199 ⇒ 00:34:19.659 Pranav Narahari: So, separate issue that Janiece messaged me about was, what’s… what’s going on with the templates for cancellations. I think there’s some issue regarding that.

378 00:34:20.389 ⇒ 00:34:22.270 Pranav Narahari: Is that something we can test right now?

379 00:34:23.510 ⇒ 00:34:25.389 Mustafa Raja: Yeah. What’s the issue?

380 00:34:26.570 ⇒ 00:34:38.269 Pranav Narahari: Yeah, I mean, she just kind of messaged me, like, a flea of different messages, a flood of different messages, I mean. So, the DB stuff that we’re looking into, but then also, she just said, not sure what happened to the template for cancellations.

381 00:34:39.090 ⇒ 00:34:41.419 Samuel Roberts: Like, it stopped working or something, or…

382 00:34:41.570 ⇒ 00:34:45.900 Pranav Narahari: Yeah, that’s my guess, but we can just test that right now. If we see that it’s working, then I can ask her to, like…

383 00:34:46.440 ⇒ 00:34:47.809 Pranav Narahari: Give more context.

384 00:35:16.590 ⇒ 00:35:17.770 Samuel Roberts: Excuse me.

385 00:35:18.130 ⇒ 00:35:22.159 Pranav Narahari: Okay, so Mustafa, does this… Give you more information?

386 00:35:24.240 ⇒ 00:35:26.390 Samuel Roberts: Oh, yeah, good.

387 00:35:29.430 ⇒ 00:35:32.660 Mustafa Raja: I’ll have to take a look at the execution to see what’s happening here.

388 00:35:33.340 ⇒ 00:35:33.819 Samuel Roberts: Oh my god.

389 00:35:33.820 ⇒ 00:35:36.810 Pranav Narahari: Okay, maybe in parallel you can be doing that?

390 00:35:36.810 ⇒ 00:35:37.450 Mustafa Raja: Yeah.

391 00:35:38.070 ⇒ 00:35:38.780 Pranav Narahari: Okay, cool.

392 00:35:44.470 ⇒ 00:35:46.529 Casie Aviles: You can get the trace here.

393 00:35:49.630 ⇒ 00:35:51.460 Mustafa Raja: How are you accessing that?

394 00:35:51.910 ⇒ 00:35:58.590 Casie Aviles: Let me send it to you. But it should also be in… In the,

395 00:35:58.860 ⇒ 00:36:00.779 Casie Aviles: ABC logs, yeah.

396 00:36:01.770 ⇒ 00:36:03.640 Casie Aviles: Yeah, it’s here in the Zoom chat.

397 00:36:03.640 ⇒ 00:36:04.689 Mustafa Raja: Oh, okay, fine.

398 00:36:29.430 ⇒ 00:36:39.749 Mustafa Raja: If I recall correctly, what we did is, they sent us a new document, or a new central doc, that had all of the

399 00:36:39.960 ⇒ 00:36:41.269 Mustafa Raja: templates, right?

400 00:36:42.880 ⇒ 00:36:44.760 Mustafa Raja: And this is regarding that, right?

401 00:36:45.540 ⇒ 00:36:46.400 Mustafa Raja: Fair enough.

402 00:36:47.240 ⇒ 00:36:50.090 Pranav Narahari: Yeah, so…

403 00:36:50.550 ⇒ 00:36:56.300 Pranav Narahari: the… there was a bunch of other templates that I sent over, is it… so, the cancellation templates were in there?

404 00:36:57.620 ⇒ 00:37:02.429 Mustafa Raja: Yeah, what I’m asking is, this should be coming from that document, right?

405 00:37:04.370 ⇒ 00:37:13.939 Pranav Narahari: I think so? Should we try to confirm that right now? Like, were these cancellation templates there before? Because I think there was some global information before, too, right?

406 00:37:14.130 ⇒ 00:37:17.600 Mustafa Raja: Yes, yes, let me double-check that, actually.

407 00:37:18.380 ⇒ 00:37:19.050 Pranav Narahari: Okay.

408 00:37:21.870 ⇒ 00:37:28.010 Pranav Narahari: Because I think now, at this point, everything is consolidated, right? So the old information, new information is all in just one place.

409 00:37:28.380 ⇒ 00:37:29.220 Pranav Narahari: Right?

410 00:37:30.120 ⇒ 00:37:31.550 Mustafa Raja: Yes.

411 00:37:31.550 ⇒ 00:37:32.180 Pranav Narahari: Okay.

412 00:37:33.690 ⇒ 00:37:37.000 Pranav Narahari: So whatever dock that that should be pulling from?

413 00:37:37.510 ⇒ 00:37:40.559 Pranav Narahari: It looks like there’s some permission issue, or…

414 00:37:40.560 ⇒ 00:37:46.979 Mustafa Raja: No, no, no, so, this is being… this is… this is calling a, an agent that we developed.

415 00:37:46.980 ⇒ 00:37:47.630 Samuel Roberts: Whoa!

416 00:37:48.050 ⇒ 00:37:49.210 Mustafa Raja: A lot earlier.

417 00:37:50.060 ⇒ 00:37:55.160 Mustafa Raja: a verbatim agent that we have disregarded now. It’s still in the code.

418 00:37:55.470 ⇒ 00:37:57.420 Mustafa Raja: So, prompt calling that…

419 00:37:58.240 ⇒ 00:37:59.000 Casie Aviles: Okay.

420 00:38:01.040 ⇒ 00:38:01.770 Casie Aviles: Okay.

421 00:38:02.320 ⇒ 00:38:04.360 Casie Aviles: So it should be looking at the central dock now.

422 00:38:05.410 ⇒ 00:38:07.820 Mustafa Raja: Yeah, it should be rather looking into…

423 00:38:08.480 ⇒ 00:38:19.270 Mustafa Raja: what’s it called, into the, vector agent, rather than the verbatim agent. So we need to delete that agent also, and any context regarding that in the

424 00:38:19.900 ⇒ 00:38:21.530 Mustafa Raja: System prompt also.

425 00:38:22.410 ⇒ 00:38:23.030 Casie Aviles: Okay.

426 00:38:28.300 ⇒ 00:38:30.129 Mustafa Raja: But I’m going to double-check it.

427 00:38:30.580 ⇒ 00:38:31.510 Mustafa Raja: with the…

428 00:40:36.020 ⇒ 00:40:36.840 Samuel Roberts: Oh, God.

429 00:40:37.220 ⇒ 00:40:41.690 Samuel Roberts: The thing is so… I’m straight into that.

430 00:40:44.270 ⇒ 00:40:45.450 Samuel Roberts: giggling.

431 00:40:46.400 ⇒ 00:40:47.780 Samuel Roberts: Oops!

432 00:40:50.130 ⇒ 00:40:54.099 Samuel Roberts: Okay, Glenn! Role ID…

433 00:41:02.680 ⇒ 00:41:05.509 Pranav Narahari: I’m also looking at some of these, Casey, to help you out.

434 00:41:05.950 ⇒ 00:41:06.520 Casie Aviles: Okay.

435 00:41:09.040 ⇒ 00:41:15.719 Pranav Narahari: So, the Patricia Velta, or Vela one, that’s another template.

436 00:41:15.960 ⇒ 00:41:21.019 Pranav Narahari: Issue, it’s looking like. So, where are we organizing that in the investigative field?

437 00:41:21.140 ⇒ 00:41:22.430 Pranav Narahari: section, I should say.

438 00:41:25.780 ⇒ 00:41:30.920 Casie Aviles: Let’s see… Yeah, we could just create, like, another… yeah.

439 00:41:31.050 ⇒ 00:41:32.000 Casie Aviles: Template.

440 00:41:35.300 ⇒ 00:41:42.280 Mustafa Raja: Yeah, so, let me share my screen. Can I share my screen for a moment?

441 00:41:42.660 ⇒ 00:41:43.949 Casie Aviles: Sure, sure. Sure.

442 00:41:48.340 ⇒ 00:41:55.880 Mustafa Raja: Okay, yeah, so there’s no cancellation template over here.

443 00:41:56.640 ⇒ 00:42:06.190 Mustafa Raja: I’m wondering… I mean… So this is initial pest control service. Yeah. So, so,

444 00:42:06.780 ⇒ 00:42:10.699 Mustafa Raja: ABC4 told you that cancellation templates weren’t working right.

445 00:42:11.150 ⇒ 00:42:15.550 Mustafa Raja: So, can you ask where they live, where those templates live?

446 00:42:17.120 ⇒ 00:42:19.840 Pranav Narahari: Don’t we have, like, a global dock?

447 00:42:21.710 ⇒ 00:42:23.540 Mustafa Raja: Over.

448 00:42:25.760 ⇒ 00:42:27.030 Mustafa Raja: Global Doc?

449 00:42:28.420 ⇒ 00:42:30.819 Pranav Narahari: Yeah, in our central dog, we don’t have a global.

450 00:42:31.470 ⇒ 00:42:32.480 Pranav Narahari: Restructured.

451 00:42:32.480 ⇒ 00:42:32.840 Mustafa Raja: virtual dog.

452 00:42:34.200 ⇒ 00:42:36.240 Mustafa Raja: We have an all-coverage dog, but…

453 00:42:40.500 ⇒ 00:42:47.389 Mustafa Raja: We don’t have, yeah, I don’t think so we have, we have something like…

454 00:42:47.390 ⇒ 00:42:47.970 Samuel Roberts: Here.

455 00:42:48.720 ⇒ 00:42:53.730 Pranav Narahari: So I shared, like, that, like, that universal… folder, right?

456 00:42:54.310 ⇒ 00:42:55.910 Mustafa Raja: Cancellation follow-up…

457 00:42:56.310 ⇒ 00:42:57.050 Pranav Narahari: Yeah.

458 00:42:57.410 ⇒ 00:43:03.029 Mustafa Raja: Okay, so the thing that you… the folder that you, sent is this, right?

459 00:43:03.030 ⇒ 00:43:03.690 Pranav Narahari: Yeah.

460 00:43:10.090 ⇒ 00:43:12.089 Mustafa Raja: This is… is this a temporary work?

461 00:43:13.410 ⇒ 00:43:15.899 Pranav Narahari: It looks like, a template.

462 00:43:17.780 ⇒ 00:43:18.410 Mustafa Raja: Okay.

463 00:43:30.620 ⇒ 00:43:33.480 Pranav Narahari: So are these currently vectorized, or were these just not…

464 00:43:33.480 ⇒ 00:43:35.910 Mustafa Raja: Yeah, these are… these are in the vector store.

465 00:43:36.520 ⇒ 00:43:37.150 Pranav Narahari: Okay.

466 00:43:51.790 ⇒ 00:43:53.599 Mustafa Raja: Yeah, I guess,

467 00:43:54.270 ⇒ 00:43:59.399 Mustafa Raja: If this is what they’re talking about, what we need to do is just get rid of, verbatim agent.

468 00:44:00.890 ⇒ 00:44:05.699 Pranav Narahari: Okay, good. Yeah, let’s get rid of the verbatim agent.

469 00:44:08.610 ⇒ 00:44:10.700 Pranav Narahari: I’ll add that to the comment, to the ticket.

470 00:44:16.710 ⇒ 00:44:17.360 Samuel Roberts: And…

471 00:44:19.470 ⇒ 00:44:22.079 Pranav Narahari: So that should fix the template issues that we’re seeing.

472 00:44:26.570 ⇒ 00:44:29.529 Mustafa Raja: So the verbatim agent is,

473 00:44:34.000 ⇒ 00:44:41.170 Mustafa Raja: Is what we developed to throw some of the verbatim stuff, but we are not using it anymore.

474 00:44:42.260 ⇒ 00:44:43.119 Mustafa Raja: We’ll say this?

475 00:44:43.120 ⇒ 00:44:48.859 Pranav Narahari: Did we confirm that, it is calling the verbatim agent when we ask that question?

476 00:44:50.010 ⇒ 00:44:51.699 Mustafa Raja: I did confirm that.

477 00:44:51.700 ⇒ 00:44:56.059 Pranav Narahari: Okay, perfect. Yeah, if we confirm that, then that is… that is good.

478 00:44:56.170 ⇒ 00:44:57.030 Pranav Narahari: Okay.

479 00:44:57.290 ⇒ 00:44:57.910 Mustafa Raja: Yeah.

480 00:45:01.370 ⇒ 00:45:02.170 Mustafa Raja: Here.

481 00:45:03.050 ⇒ 00:45:05.080 Mustafa Raja: Executive verbatim agent.

482 00:45:05.790 ⇒ 00:45:06.770 Pranav Narahari: Perfect, perfect, perfect.

483 00:45:06.770 ⇒ 00:45:11.360 Mustafa Raja: I’ll actually rather just… For now, update the workflow.

484 00:45:11.940 ⇒ 00:45:13.229 Mustafa Raja: Not this one.

485 00:45:14.780 ⇒ 00:45:16.129 Mustafa Raja: Not this one.

486 00:45:19.460 ⇒ 00:45:21.560 Mustafa Raja: I have a lot of windows open.

487 00:45:22.810 ⇒ 00:45:27.120 Samuel Roberts: Okay. Yeah, they’re all lawn mowing crews that have just days of the week.

488 00:45:29.050 ⇒ 00:45:30.519 Pranav Narahari: Awesome, that makes it simple.

489 00:45:31.170 ⇒ 00:45:33.690 Samuel Roberts: Yeah, there’s just 65 of them,

490 00:45:33.990 ⇒ 00:45:42.799 Samuel Roberts: And I think I’m just gonna update it to say whatever is in there will be tagged onto the end of this crew mows lawns in this zip code on blank.

491 00:45:44.780 ⇒ 00:45:49.950 Pranav Narahari: Cool. And that’s gonna go specifically to which notes again? Not the crew notes, the…

492 00:45:50.590 ⇒ 00:45:54.230 Samuel Roberts: The assignment notes, because the crew notes have different notes.

493 00:45:54.640 ⇒ 00:45:55.800 Pranav Narahari: Perfect. Okay.

494 00:45:57.720 ⇒ 00:46:00.489 Pranav Narahari: And then, will we still need to update the…

495 00:46:00.800 ⇒ 00:46:04.180 Pranav Narahari: The system prompt to look at both assignment and crew notes.

496 00:46:04.280 ⇒ 00:46:07.549 Pranav Narahari: Okay. Correct. Yep. KCR, you wrote that down, perfect.

497 00:46:07.680 ⇒ 00:46:08.330 Samuel Roberts: Yep, he did.

498 00:46:12.720 ⇒ 00:46:13.390 Pranav Narahari: Oh.

499 00:46:20.430 ⇒ 00:46:21.439 Pranav Narahari: Yeah, good question.

500 00:47:02.590 ⇒ 00:47:06.430 Samuel Roberts: And we edited Reagan Peterman.

501 00:47:06.980 ⇒ 00:47:08.779 Samuel Roberts: Just wanna see, make sure I’m matching that.

502 00:47:10.140 ⇒ 00:47:15.579 Samuel Roberts: The day of the week this crew mows is… oh, and right, there’s several, we want to join them too, don’t we?

503 00:47:17.860 ⇒ 00:47:20.639 Pranav Narahari: Yeah, we need to consolidate these.

504 00:47:20.640 ⇒ 00:47:21.310 Samuel Roberts: Yep, yep.

505 00:47:21.310 ⇒ 00:47:28.959 Pranav Narahari: But I also don’t… Yeah, so since it’s showing double, we need to be consolidating, right? Yeah, okay.

506 00:47:29.730 ⇒ 00:47:33.640 Pranav Narahari: Okay, so maybe that can be a separate script that we run, but…

507 00:47:33.640 ⇒ 00:47:37.670 Samuel Roberts: That’s what I’m just looking to see if anything… lines up here…

508 00:47:40.940 ⇒ 00:47:42.220 Samuel Roberts: Sort…

509 00:47:46.890 ⇒ 00:47:51.799 Pranav Narahari: I mean, that can also be something on the system prompt as well, like consolidate…

510 00:47:52.830 ⇒ 00:47:53.799 Samuel Roberts: Yeah, that’s true.

511 00:47:54.650 ⇒ 00:48:03.179 Pranav Narahari: consolidate, the same… like, if… if… yeah, consolidate the same, what is it? Person.

512 00:48:12.350 ⇒ 00:48:14.699 Pranav Narahari: Okay, for these other 4, Casey?

513 00:48:20.930 ⇒ 00:48:27.220 Pranav Narahari: Update needed for… Yeah, for the CSR feedback for Deborah.

514 00:48:27.570 ⇒ 00:48:32.469 Pranav Narahari: From Debra, I mean. Update needed for Trelena Termite Services.

515 00:48:33.550 ⇒ 00:48:35.750 Pranav Narahari: This is just missing, right?

516 00:48:37.320 ⇒ 00:48:39.520 Casie Aviles: For which one again? Sorry.

517 00:48:41.700 ⇒ 00:48:45.990 Pranav Narahari: Trelena… or it’s, Deborah Huggins-Jones.

518 00:48:46.880 ⇒ 00:48:48.780 Casie Aviles: Oh, this one, the first one, okay.

519 00:48:48.780 ⇒ 00:48:49.330 Pranav Narahari: Yep.

520 00:48:50.630 ⇒ 00:48:56.030 Casie Aviles: Okay, yeah, this… this seems to be an update, because there’s… there’s an output returned.

521 00:48:57.080 ⇒ 00:49:00.860 Pranav Narahari: Yeah, I’m just going to ask… Andy again.

522 00:49:29.060 ⇒ 00:49:32.060 Pranav Narahari: Oh, instead of main… instead of maintenance, they wrote paint?

523 00:49:33.150 ⇒ 00:49:35.650 Pranav Narahari: Okay, so I was able to figure that out.

524 00:49:38.200 ⇒ 00:49:39.969 Pranav Narahari: I got an answer.

525 00:49:44.560 ⇒ 00:49:45.960 Casie Aviles: Yeah.

526 00:49:46.340 ⇒ 00:49:49.590 Casie Aviles: But not every termite technician can perform.

527 00:49:54.100 ⇒ 00:49:54.880 Casie Aviles: Okay.

528 00:49:56.540 ⇒ 00:49:59.279 Pranav Narahari: Here, this is the answer I got. Just gonna leave it as a comment here.

529 00:50:18.050 ⇒ 00:50:21.389 Pranav Narahari: Oh, I guess that is the output that they got, but they’re saying that’s incorrect.

530 00:50:26.610 ⇒ 00:50:34.390 Pranav Narahari: Oh, okay, their issue isn’t with that, it’s with… they want to do the update with… they were trying to do the update within Andy.

531 00:50:38.520 ⇒ 00:50:39.319 Pranav Narahari: Did you see that?

532 00:50:39.320 ⇒ 00:50:41.360 Samuel Roberts: change the… oh, interesting.

533 00:50:45.240 ⇒ 00:50:49.330 Casie Aviles: For, for the technician, Travis? Is that it?

534 00:50:49.880 ⇒ 00:50:52.240 Pranav Narahari: Yeah, yeah, I’m seeing where you’re highlighting, yeah.

535 00:50:53.780 ⇒ 00:50:55.309 Pranav Narahari: Right here.

536 00:50:55.310 ⇒ 00:50:55.900 Casie Aviles: opener.

537 00:50:58.170 ⇒ 00:50:59.080 Casie Aviles: Okay.

538 00:51:00.040 ⇒ 00:51:01.890 Casie Aviles: So that… Yes.

539 00:51:01.890 ⇒ 00:51:04.310 Pranav Narahari: Incorrect usability. Yeah.

540 00:51:05.600 ⇒ 00:51:06.320 Pranav Narahari: 10.

541 00:51:06.490 ⇒ 00:51:09.610 Pranav Narahari: Cool. So, let me delete my comment, actually, it’s a little confusing.

542 00:51:17.640 ⇒ 00:51:21.830 Pranav Narahari: And so, how do you now define that that’s a DB update?

543 00:51:24.010 ⇒ 00:51:27.530 Casie Aviles: If it’s… Not in the Debian.

544 00:51:28.430 ⇒ 00:51:31.129 Casie Aviles: Andy did not return anything.

545 00:51:32.780 ⇒ 00:51:34.780 Casie Aviles: So, we need to, like, add…

546 00:51:34.990 ⇒ 00:51:41.249 Casie Aviles: Or if it’s, like, an outdated record, then we have to change it or update it. That’s how I define it. Right.

547 00:51:41.840 ⇒ 00:51:50.059 Pranav Narahari: I guess not, that’s, like, how we’re diagnosing it, right? Like, how do we now define it in the ticket, so then it gets updated by Janice?

548 00:51:52.020 ⇒ 00:51:54.879 Pranav Narahari: Or are you the one that does that? I think it’s Janiece, right?

549 00:51:55.830 ⇒ 00:51:59.500 Casie Aviles: Yeah, we… I just assigned it back to her when… when it’s…

550 00:52:00.320 ⇒ 00:52:03.149 Casie Aviles: You know, when it’s about, like, updating something, or…

551 00:52:03.970 ⇒ 00:52:05.780 Pranav Narahari: Oh, okay, and then you put a comment?

552 00:52:06.550 ⇒ 00:52:07.130 Casie Aviles: Yep.

553 00:52:08.420 ⇒ 00:52:09.100 Pranav Narahari: Okay.

554 00:52:09.900 ⇒ 00:52:18.629 Mustafa Raja: Okay, so I, I got rid of the verbatim agent. Now, when I rerun it, what it does is it calls a tool.

555 00:52:19.090 ⇒ 00:52:23.989 Mustafa Raja: Google… Google Sheets template tool, because initially.

556 00:52:24.780 ⇒ 00:52:30.450 Mustafa Raja: the templates were stored in Google Sheets. I don’t know if this is going to be maintained anymore or not.

557 00:52:30.720 ⇒ 00:52:33.439 Mustafa Raja: So let me know if I should get rid of this also.

558 00:52:34.080 ⇒ 00:52:40.060 Mustafa Raja: and only… Look at the… Central dots.

559 00:52:45.870 ⇒ 00:52:52.330 Pranav Narahari: Yeah, I don’t think the Google Sheets is being maintained. Can you send me the Google Sheet link? I don’t even think I’ve seen that.

560 00:52:54.510 ⇒ 00:52:55.510 Mustafa Raja: Let me find…

561 00:53:10.890 ⇒ 00:53:14.739 Pranav Narahari: Yeah, it’s not in the restructured Central Dogs.

562 00:53:19.510 ⇒ 00:53:22.499 Mustafa Raja: Yeah, it’s not in there, let me… let me get you that.

563 00:54:02.250 ⇒ 00:54:03.929 Mustafa Raja: I’ve sent it in Zoom chat.

564 00:54:10.300 ⇒ 00:54:11.400 Pranav Narahari: Yeah, let me take a look.

565 00:54:21.570 ⇒ 00:54:23.430 Pranav Narahari: Oh,

566 00:54:28.010 ⇒ 00:54:29.100 Pranav Narahari: Interesting.

567 00:54:30.280 ⇒ 00:54:32.349 Pranav Narahari: So this isn’t currently being…

568 00:54:35.510 ⇒ 00:54:39.339 Pranav Narahari: Oh, so it pulled the cancellation template from this? Is that what you’re saying?

569 00:54:40.390 ⇒ 00:54:44.670 Mustafa Raja: Re… So,

570 00:54:45.010 ⇒ 00:54:56.490 Mustafa Raja: initially, the output we saw when KC ran it, a verbatim agent ran, and it didn’t have access. Now that I have removed verbatim agent.

571 00:54:56.600 ⇒ 00:54:59.449 Mustafa Raja: The vector agent is calling this tool.

572 00:54:59.700 ⇒ 00:55:02.710 Mustafa Raja: This Google Sheet to find the template.

573 00:55:03.360 ⇒ 00:55:06.530 Pranav Narahari: And so, was the response correct? Did it give the.

574 00:55:06.530 ⇒ 00:55:08.639 Mustafa Raja: I mean, according to this one.

575 00:55:08.830 ⇒ 00:55:13.939 Mustafa Raja: According to this one, this is correct, but I don’t know if this is going to be maintained.

576 00:55:14.190 ⇒ 00:55:17.989 Pranav Narahari: Yeah, just send me a screenshot of that, and I’m just gonna say…

577 00:55:18.390 ⇒ 00:55:19.800 Pranav Narahari: Hey, this is what we’re seeing.

578 00:55:19.950 ⇒ 00:55:22.670 Pranav Narahari: Is this.

579 00:55:22.670 ⇒ 00:55:25.080 Mustafa Raja: Yeah, but I haven’t deployed it yet.

580 00:55:27.260 ⇒ 00:55:28.670 Pranav Narahari: Oh, I see, okay.

581 00:55:28.900 ⇒ 00:55:34.849 Mustafa Raja: Yeah, because if you say that you’re seeing this, then they won’t be able to see this, because I haven’t deployed.

582 00:55:36.100 ⇒ 00:55:39.940 Mustafa Raja: So, I just need a decision from team here.

583 00:55:39.940 ⇒ 00:55:43.969 Pranav Narahari: Yeah, I think, let’s just continue to…

584 00:55:44.310 ⇒ 00:55:57.980 Pranav Narahari: Yeah, maintain this as well, for right now, and then I will… I’ll sync with, Janiece to just… because there’s only 13 templates here, like, there’s no reason that it should exist here. We can just put this into a Google Doc as well.

585 00:56:00.090 ⇒ 00:56:12.469 Pranav Narahari: But for right now, like, for an immediate patch, right, you can just push that. Because all we did basically right now is just remove the verbatim agent. And now we can talk about removing this central doc template agent.

586 00:56:13.280 ⇒ 00:56:17.889 Mustafa Raja: This isn’t an agent, this is just a tool with the vector agent.

587 00:56:20.600 ⇒ 00:56:24.189 Pranav Narahari: Oh, I see, so the vector agent also checks central dock?

588 00:56:25.040 ⇒ 00:56:27.799 Pranav Narahari: But then it also looks at this Google Sheet, is that what you’re saying?

589 00:56:27.800 ⇒ 00:56:29.189 Mustafa Raja: Yes, yes.

590 00:56:29.190 ⇒ 00:56:29.630 Pranav Narahari: Okay.

591 00:56:29.630 ⇒ 00:56:34.500 Mustafa Raja: Initially, templates, templates were being stored in Google Sheets.

592 00:56:34.660 ⇒ 00:56:41.160 Mustafa Raja: when we moved to Maestra, So, we, we all, we had the solution.

593 00:56:41.260 ⇒ 00:56:44.409 Mustafa Raja: Already in place, that is why this exists.

594 00:56:46.880 ⇒ 00:56:54.800 Pranav Narahari: Okay, gotcha. So, I mean, then in the future, if we just remove all this information from this template, then it shouldn’t even require a redeploy, right?

595 00:56:55.230 ⇒ 00:56:57.060 Mustafa Raja: Yes. We can just…

596 00:56:57.060 ⇒ 00:56:57.860 Pranav Narahari: That’s really cool.

597 00:56:58.250 ⇒ 00:57:00.940 Mustafa Raja: Maybe update the system prompt.

598 00:57:01.120 ⇒ 00:57:02.329 Mustafa Raja: put our,

599 00:57:02.810 ⇒ 00:57:12.029 Mustafa Raja: vector agent, and then it should… it should be only looking into the Google Docs, all of the embeddings in the Google Docs.

600 00:57:13.280 ⇒ 00:57:21.079 Pranav Narahari: Okay, perfect. Yeah, so then I would just deploy what we have right now, because all you did right now, basically, is just remove the verbatim agent.

601 00:57:21.820 ⇒ 00:57:23.599 Mustafa Raja: Yes, let me do that.

602 00:57:24.280 ⇒ 00:57:25.330 Pranav Narahari: Okay, perfect.

603 00:57:28.240 ⇒ 00:57:35.469 Samuel Roberts: Alright, so the ones are merged and updated. I’m now updating the ones that are not merged, that are just single days.

604 00:57:39.200 ⇒ 00:57:39.890 Casie Aviles: Okay.

605 00:58:02.060 ⇒ 00:58:06.379 Pranav Narahari: Casey, did you want to assign this to Janiece, or should I do it right now?

606 00:58:07.360 ⇒ 00:58:08.130 Pranav Narahari: The devil.

607 00:58:08.130 ⇒ 00:58:09.040 Casie Aviles: Hi, Deborah.

608 00:58:09.710 ⇒ 00:58:10.360 Pranav Narahari: Yeah.

609 00:58:11.150 ⇒ 00:58:18.510 Casie Aviles: Okay, yeah, let me… I’m just doing another check of the results, so… over a shore with…

610 00:58:19.250 ⇒ 00:58:19.910 Pranav Narahari: No problem.

611 00:58:23.150 ⇒ 00:58:24.470 Casie Aviles: I can share…

612 00:59:05.700 ⇒ 00:59:08.889 Mustafa Raja: I’ve pushed the code, it’ll take, some time to deploy.

613 00:59:10.110 ⇒ 00:59:13.420 Mustafa Raja: And then maybe you can… Update them regarding this.

614 00:59:14.270 ⇒ 00:59:16.570 Pranav Narahari: Cool, yeah, Mustava, once that just,

615 00:59:17.830 ⇒ 00:59:27.949 Pranav Narahari: Well, once that’s redeployed, we’ll just do, or once that finishes redeploying, we’re just gonna test all of these things that we just investigated, because some of them we already made solutions for as well, so…

616 00:59:28.710 ⇒ 00:59:31.400 Pranav Narahari: Okay. Then I’ll… we’ll do a final update to the client.

617 00:59:40.860 ⇒ 00:59:43.610 Samuel Roberts: Okay, looks like we’re good, I’m just doing one last check.

618 01:00:36.730 ⇒ 01:00:37.340 Samuel Roberts: Okay.

619 01:00:37.760 ⇒ 01:00:40.650 Samuel Roberts: Looks like it’s good.

620 01:00:48.020 ⇒ 01:00:51.399 Samuel Roberts: Let me go back, though, make sure I don’t have the duplicates in there anymore.

621 01:01:01.580 ⇒ 01:01:02.420 Samuel Roberts: Nope, okay.

622 01:01:42.630 ⇒ 01:01:47.519 Pranav Narahari: Sorry, Sam, I was writing a message. You said that you just made a… you just executed both of the scripts?

623 01:01:48.210 ⇒ 01:01:52.979 Samuel Roberts: Yeah, I’m… there’s a couple little cleanup things, I just… I was double-checking to make sure things…

624 01:01:53.350 ⇒ 01:01:54.280 Samuel Roberts: are good, and I think…

625 01:01:54.280 ⇒ 01:01:54.899 Pranav Narahari: I’ll feed you through.

626 01:01:54.900 ⇒ 01:01:56.349 Samuel Roberts: move now? Yep.

627 01:01:56.860 ⇒ 01:02:00.379 Pranav Narahari: I just, created this QA complete section, so,

628 01:02:01.070 ⇒ 01:02:07.079 Pranav Narahari: for the ones that were affected by the DB stuff, let’s just do some QA, just so we can,

629 01:02:07.570 ⇒ 01:02:08.849 Pranav Narahari: Test that it’s working.

630 01:02:10.130 ⇒ 01:02:11.839 Pranav Narahari: Let us know when we can do that.

631 01:02:13.410 ⇒ 01:02:14.010 Samuel Roberts: Okay.

632 01:02:20.910 ⇒ 01:02:23.700 Samuel Roberts: Yeah, so right now, we’re… we’re pretty close. We’re,

633 01:02:24.090 ⇒ 01:02:30.490 Samuel Roberts: the notes are all updated, except it… it didn’t clean out the duplicate ones completely, so I’m just clearing those out now.

634 01:02:32.740 ⇒ 01:02:38.229 Pranav Narahari: Okay. I also just wanna be sure that they’re duplicates, so… I just wanna… Yeah.

635 01:02:38.510 ⇒ 01:02:39.270 Pranav Narahari: Okay.

636 01:02:40.660 ⇒ 01:02:56.690 Samuel Roberts: Yeah, so basically what happened is, like, it… for one, like, the Reagan-Peterman one that had the same zip code, all the other stuff’s the same, except it was Monday, and one was Wednesday, it combined that until a Monday-Wednesday one, but it didn’t delete the Wednesday one yet, so I’m just clearing those ones out now.

637 01:02:57.120 ⇒ 01:02:58.850 Pranav Narahari: Okay, that makes sense, yeah.

638 01:03:06.350 ⇒ 01:03:08.410 Pranav Narahari: Casey, what’s this for a review section?

639 01:03:09.240 ⇒ 01:03:11.640 Casie Aviles: Oh, this is the one that I sent to…

640 01:03:11.890 ⇒ 01:03:16.059 Casie Aviles: Or I assigned to Jenny’s, like, I’m just asking her to confirm if…

641 01:03:16.460 ⇒ 01:03:19.290 Casie Aviles: The response is correct now. Since this is…

642 01:03:19.850 ⇒ 01:03:22.770 Casie Aviles: Reading from the central dock now.

643 01:03:23.770 ⇒ 01:03:26.269 Pranav Narahari: Well, if you scroll up, right?

644 01:03:26.740 ⇒ 01:03:30.749 Pranav Narahari: That’s, like, the output that they already displayed.

645 01:03:33.890 ⇒ 01:03:36.309 Pranav Narahari: Right, no, up one more, the output.

646 01:03:39.250 ⇒ 01:03:44.450 Casie Aviles: Okay, and then it asked, like, approved… it, yeah, it returned these.

647 01:03:46.980 ⇒ 01:03:48.990 Pranav Narahari: Oh, oh, that’s part of the output.

648 01:03:51.020 ⇒ 01:03:51.840 Pranav Narahari: Okay.

649 01:03:52.050 ⇒ 01:03:57.120 Pranav Narahari: I thought this… this was… Created based on that second section.

650 01:03:57.240 ⇒ 01:04:01.189 Pranav Narahari: I mean, that second part, you see where it says input, please update, tech info?

651 01:04:04.160 ⇒ 01:04:07.250 Pranav Narahari: Let me… I’ll highlight it.

652 01:04:08.230 ⇒ 01:04:09.539 Pranav Narahari: This part right here.

653 01:04:11.660 ⇒ 01:04:13.110 Casie Aviles: Oh, okay.

654 01:04:13.730 ⇒ 01:04:18.079 Pranav Narahari: Yeah. So this screenshot isn’t really giving any additional information.

655 01:04:18.590 ⇒ 01:04:21.020 Pranav Narahari: Right? Am I… am I understanding this correctly?

656 01:04:23.050 ⇒ 01:04:29.050 Casie Aviles: Yeah, you’re right, this is… it’s… it’s the same, like, output. I think… I guess it was just confusing, like.

657 01:04:29.990 ⇒ 01:04:34.559 Casie Aviles: They should have been in the, in the feedback field, so… but I guess they, they sent, like.

658 01:04:35.850 ⇒ 01:04:41.309 Casie Aviles: an input instead. They sent it as a message to Andy. That’s why I was confused.

659 01:04:41.900 ⇒ 01:04:42.690 Pranav Narahari: Yeah.

660 01:04:43.320 ⇒ 01:04:48.359 Pranav Narahari: Yeah, I think they tried to… well, the feedback that they gave was just, like.

661 01:04:49.060 ⇒ 01:04:49.530 Casie Aviles: Oh, wow.

662 01:04:49.530 ⇒ 01:04:54.329 Pranav Narahari: to do the update, but then they also tried to ask Andy to do the update, which didn’t work.

663 01:04:55.780 ⇒ 01:04:57.480 Casie Aviles: Okay, yeah, yeah, that’s right.

664 01:04:58.740 ⇒ 01:04:59.300 Pranav Narahari: Okay.

665 01:04:59.930 ⇒ 01:05:02.299 Pranav Narahari: So, I would probably just delete your comment.

666 01:05:02.910 ⇒ 01:05:03.730 Pranav Narahari: Yeah.

667 01:05:09.380 ⇒ 01:05:18.100 Pranav Narahari: And then, do you usually, like, when you assign it to her, give her, like, some context of, like, why we assigned it to her? Or does she usually just kind of understand?

668 01:05:18.550 ⇒ 01:05:31.439 Casie Aviles: When it’s, like, well, I add, like, a comment that this should be fixed now, or this should… this feedback should be resolved, that’s… at minimum, that’s the context that I give her, and then I give her, like, a screenshot.

669 01:05:32.120 ⇒ 01:05:33.570 Casie Aviles: So, and then it gets assigned.

670 01:05:33.570 ⇒ 01:05:38.670 Pranav Narahari: Here, since it’s assigned to… oh, so, did Janiece assign it to you?

671 01:05:39.510 ⇒ 01:05:41.280 Casie Aviles: Yeah, she assigned it to me.

672 01:05:41.610 ⇒ 01:05:46.099 Pranav Narahari: So she probably accidentally assigned it to you. She didn’t realize it was a DB update.

673 01:05:49.810 ⇒ 01:05:50.540 Casie Aviles: Yeah.

674 01:05:50.980 ⇒ 01:05:54.289 Pranav Narahari: Right, so I would just put in the comments, like,

675 01:05:54.670 ⇒ 01:05:57.019 Pranav Narahari: Looks like… looks to be a DB update.

676 01:05:59.950 ⇒ 01:06:01.840 Pranav Narahari: Let me know if you need anything from me.

677 01:06:15.610 ⇒ 01:06:16.360 Pranav Narahari: Okay.

678 01:06:19.070 ⇒ 01:06:24.140 Pranav Narahari: Cool. So yeah, this one is off our plate.

679 01:06:26.220 ⇒ 01:06:28.810 Pranav Narahari: Mustafa, how are we doing with the deploy?

680 01:06:29.130 ⇒ 01:06:31.110 Pranav Narahari: Should be ready now, right?

681 01:06:31.610 ⇒ 01:06:34.350 Mustafa Raja: Let me… Check.

682 01:06:34.910 ⇒ 01:06:35.530 Pranav Narahari: Okay.

683 01:06:47.690 ⇒ 01:06:48.440 Mustafa Raja: Yes.

684 01:06:52.260 ⇒ 01:06:52.980 Pranav Narahari: Okay.

685 01:06:53.610 ⇒ 01:06:56.299 Pranav Narahari: Okay, I’m gonna test the Patricia Vela one.

686 01:06:58.620 ⇒ 01:07:00.970 Mustafa Raja: Can you share the screen? Oh, yeah, you’re sharing, sorry.

687 01:07:01.850 ⇒ 01:07:03.260 Casie Aviles: No, that was on my screen.

688 01:07:03.930 ⇒ 01:07:04.690 Pranav Narahari: I can share.

689 01:07:15.070 ⇒ 01:07:17.330 Pranav Narahari: You can see my screen? Oh, wait one sec.

690 01:07:18.090 ⇒ 01:07:20.380 Pranav Narahari: I think now you should be able to see my screen.

691 01:07:21.960 ⇒ 01:07:22.610 Casie Aviles: Yes.

692 01:07:23.280 ⇒ 01:07:23.950 Pranav Narahari: Huh.

693 01:07:43.370 ⇒ 01:07:45.029 Pranav Narahari: Compliment form.

694 01:07:58.880 ⇒ 01:08:02.199 Pranav Narahari: Feed us a… feed us a template for compliments.

695 01:08:06.420 ⇒ 01:08:08.910 Pranav Narahari: Compliments? What does compliments even mean?

696 01:08:23.939 ⇒ 01:08:27.300 Pranav Narahari: I don’t see anything about compliments on the Universal Docs.

697 01:08:35.620 ⇒ 01:08:39.470 Pranav Narahari: Also, nothing in… Essential.templates.

698 01:08:43.569 ⇒ 01:08:48.550 Pranav Narahari: Mustafa, so is this just basically that template doesn’t exist, and should get assigned back to Janiece?

699 01:08:50.750 ⇒ 01:08:54.650 Mustafa Raja: Can you go back to the, the Gmail?

700 01:08:58.660 ⇒ 01:09:03.609 Mustafa Raja: Yeah, comp… yeah, the compliment from… yeah. If it’s not there, then, yeah.

701 01:09:07.710 ⇒ 01:09:08.370 Pranav Narahari: Okay.

702 01:09:08.470 ⇒ 01:09:10.919 Pranav Narahari: So then, for this ticket.

703 01:09:12.970 ⇒ 01:09:15.720 Mustafa Raja: I mean, yeah, I guess, I guess what we…

704 01:09:16.729 ⇒ 01:09:20.219 Pranav Narahari: Oh, so she already gave this information, actually, that’s why she assigned it back to us.

705 01:09:20.419 ⇒ 01:09:24.309 Pranav Narahari: Send all compliments to a compliment, include the following information.

706 01:09:25.000 ⇒ 01:09:37.449 Mustafa Raja: Yeah, I guess we should, we should give, give them a… I don’t know if we want to, maintain this Google Sheet or not, or what, what do we want, what we want, what we want.

707 01:09:37.990 ⇒ 01:09:40.650 Mustafa Raja: The source of truth for templates.

708 01:09:41.029 ⇒ 01:09:44.429 Mustafa Raja: But yeah, I don’t think that we should be taken care of.

709 01:09:44.609 ⇒ 01:09:46.630 Mustafa Raja: Inputting this information, right?

710 01:09:49.750 ⇒ 01:09:56.619 Pranav Narahari: No, I think that is on us, right? We’re going to be the ones to update the central dock.

711 01:09:56.840 ⇒ 01:09:57.900 Pranav Narahari: Going forward.

712 01:09:57.900 ⇒ 01:10:02.270 Mustafa Raja: Yeah, okay, yeah, that’s right, yeah, that had a misconception then. Yeah.

713 01:10:02.510 ⇒ 01:10:05.160 Mustafa Raja: We can just add this in the Google Sheet, then.

714 01:10:05.720 ⇒ 01:10:06.450 Mustafa Raja: Okay.

715 01:10:06.990 ⇒ 01:10:16.909 Pranav Narahari: Okay, do you want to add this into the… Can we add this to the central doc? I think the Google Sheet, let’s stop adding things to.

716 01:10:17.940 ⇒ 01:10:18.520 Pranav Narahari: Unless you.

717 01:10:18.520 ⇒ 01:10:19.440 Mustafa Raja: That’s not good.

718 01:10:19.440 ⇒ 01:10:20.000 Pranav Narahari: Life.

719 01:10:20.600 ⇒ 01:10:23.820 Mustafa Raja: Yeah, we can add that, but right now…

720 01:10:24.480 ⇒ 01:10:29.560 Mustafa Raja: the agent is just going to take a look at the Google Sheet and not at the Google Doc.

721 01:10:30.400 ⇒ 01:10:31.549 Mustafa Raja: For the template.

722 01:10:32.980 ⇒ 01:10:35.700 Pranav Narahari: Oh, all templates are in the Google Sheets, okay.

723 01:10:36.800 ⇒ 01:10:41.229 Pranav Narahari: I thought there was some in the Central… the Universal Central dock as well.

724 01:10:41.670 ⇒ 01:10:46.020 Mustafa Raja: Yeah, that’s what I’m saying, the issue is, you know?

725 01:10:48.200 ⇒ 01:10:49.470 Pranav Narahari: Hmm, I see what you’re saying.

726 01:10:49.470 ⇒ 01:10:52.790 Mustafa Raja: We have two sources, we need to decide on one.

727 01:10:53.360 ⇒ 01:10:57.880 Mustafa Raja: I guess we can communicate with them and then come up with the… what we want.

728 01:11:02.500 ⇒ 01:11:04.590 Pranav Narahari: For universal usage across all centers.

729 01:11:06.450 ⇒ 01:11:07.800 Pranav Narahari: Yeah, okay.

730 01:11:07.960 ⇒ 01:11:14.630 Pranav Narahari: That makes sense. Yeah, I guess what we need to do then for right now is,

731 01:11:17.550 ⇒ 01:11:19.129 Pranav Narahari: That’s what I’ve been thinking.

732 01:11:19.290 ⇒ 01:11:37.059 Mustafa Raja: Also, I believe, there should be one Google Doc, if we want to move, to, storing these templates, and in a Google Doc, there should be one Google Doc regarding templates, where all templates should live, right? I don’t think we have that.

733 01:11:40.450 ⇒ 01:11:41.410 Pranav Narahari: Gotcha.

734 01:11:41.990 ⇒ 01:11:43.560 Samuel Roberts: Okay, that is good now.

735 01:11:43.980 ⇒ 01:11:47.329 Samuel Roberts: Sorry, I’m not trying to interrupt, but .

736 01:11:47.330 ⇒ 01:11:56.140 Pranav Narahari: No, you’re good. Okay, let’s just… yeah, we’ll test that right after. Or, Sam, if you want to do the test and then just, show us the output that works while we work on this.

737 01:11:56.450 ⇒ 01:12:02.479 Pranav Narahari: Okay, so I’m looking at Universal Docs right here, let’s see…

738 01:12:04.000 ⇒ 01:12:09.420 Pranav Narahari: These are all templates, essentially, right? Cancellation follow-up is a template.

739 01:12:10.210 ⇒ 01:12:14.979 Pranav Narahari: Collection… creating an estimate is likely a template as well.

740 01:12:15.560 ⇒ 01:12:17.449 Pranav Narahari: Or are these more like…

741 01:12:20.480 ⇒ 01:12:22.430 Pranav Narahari: Yeah, processes…

742 01:12:23.430 ⇒ 01:12:26.800 Samuel Roberts: Casey, which ticket had that, duplicate,

743 01:12:27.930 ⇒ 01:12:30.450 Samuel Roberts: To test. I just want to see how they asked it again.

744 01:12:31.300 ⇒ 01:12:33.980 Casie Aviles: Oh, it’s… Meaning.

745 01:12:37.600 ⇒ 01:12:41.230 Pranav Narahari: Okay, Mustafa, on this one, let’s…

746 01:12:43.090 ⇒ 01:12:52.390 Pranav Narahari: I have to talk to Janiece, then, to talk about, hey, we’re gonna create a central dog. I need to understand a little bit about, like, what’s in the universal docs.

747 01:12:52.390 ⇒ 01:12:56.989 Mustafa Raja: Yes, that is… that’s actually how we should cater the list, yeah.

748 01:12:57.270 ⇒ 01:12:58.200 Pranav Narahari: Yeah, actually…

749 01:12:58.200 ⇒ 01:12:59.070 Mustafa Raja: We need to be, we need to be…

750 01:13:00.350 ⇒ 01:13:05.740 Pranav Narahari: Right now, can you, in the external channel, just create a thread about this?

751 01:13:05.890 ⇒ 01:13:19.450 Pranav Narahari: And then, you know, you can link this ticket, you can also link the Universal Docs, and then also that Google Sheet, and just say, hey, we would like to have one… create one…

752 01:13:19.650 ⇒ 01:13:29.330 Pranav Narahari: Google Doc that has all of the templates, can you let us know which one of these central… universal central docs is a template?

753 01:13:29.450 ⇒ 01:13:32.310 Pranav Narahari: Or you would define as a template, and then we can then…

754 01:13:32.560 ⇒ 01:13:34.490 Pranav Narahari: Move those all into one dock.

755 01:13:35.340 ⇒ 01:13:42.250 Mustafa Raja: Yes, and I should also mention that we would be moving away from this Google Sheet, right? If I’m going to mention that.

756 01:13:42.890 ⇒ 01:13:44.810 Pranav Narahari: Yes, exactly. Thanks.

757 01:13:50.730 ⇒ 01:13:53.280 Mustafa Raja: And I should, tag Janice? Or just…

758 01:13:53.280 ⇒ 01:13:54.000 Pranav Narahari: Yeah, tagging.

759 01:13:54.850 ⇒ 01:13:55.340 Mustafa Raja: Ugh.

760 01:13:55.340 ⇒ 01:13:55.730 Pranav Narahari: Tagged in.

761 01:13:57.800 ⇒ 01:14:05.420 Pranav Narahari: Let me maybe, you know what, just to make it super clear, I’m going to create the thread, and then you can…

762 01:14:06.110 ⇒ 01:14:09.659 Pranav Narahari: you can respond in that thread, because I’m just gonna forward Janiece’s message.

763 01:14:54.350 ⇒ 01:14:55.250 Pranav Narahari: Okay.

764 01:14:55.550 ⇒ 01:14:56.780 Pranav Narahari: Just created it.

765 01:15:08.090 ⇒ 01:15:10.210 Pranav Narahari: Yeah, and then just make sure to tag Janice.

766 01:15:13.230 ⇒ 01:15:15.100 Pranav Narahari: Okay, Sam.

767 01:15:15.480 ⇒ 01:15:20.619 Samuel Roberts: Hold on, up front, I was trying to screenshot the, Andy output, it looks like it worked.

768 01:15:21.020 ⇒ 01:15:21.590 Samuel Roberts: Oh.

769 01:15:21.590 ⇒ 01:15:22.130 Pranav Narahari: Okay.

770 01:15:38.100 ⇒ 01:15:40.069 Samuel Roberts: What the hell in here? Why are you so…

771 01:15:59.110 ⇒ 01:16:00.160 Samuel Roberts: And a few new shirts.

772 01:16:00.160 ⇒ 01:16:00.480 Pranav Narahari: For sure.

773 01:16:00.480 ⇒ 01:16:04.140 Samuel Roberts: That’s what Janice, about the double entries, and we’ll try to,

774 01:16:05.280 ⇒ 01:16:08.489 Samuel Roberts: Make sure the notes are clear going forward, maybe?

775 01:16:10.050 ⇒ 01:16:10.960 Pranav Narahari: Can you repeat that?

776 01:16:11.930 ⇒ 01:16:16.319 Samuel Roberts: For Denise, if she’s making updates to the DB, rather than make two entries.

777 01:16:16.820 ⇒ 01:16:21.480 Samuel Roberts: For a given zip and, you know, crew or…

778 01:16:21.950 ⇒ 01:16:30.129 Samuel Roberts: that the notes can contain the different days, rather than two different entries. I don’t know if that was her, if that was us, but just going forward, if she’s making updates, she should know that.

779 01:16:33.060 ⇒ 01:16:38.139 Pranav Narahari: Yeah, yeah. Once I give her that message of, we fixed things, I will…

780 01:16:38.140 ⇒ 01:16:45.359 Samuel Roberts: Yeah, yeah, I just want to make sure that that gets communicated, because now that… there might even be other ones, this was just what was very clear.

781 01:16:45.520 ⇒ 01:16:52.769 Samuel Roberts: from what we had done here. I didn’t investigate beyond that, so I think it’s probably something we need to do, because there might be other times where

782 01:16:52.890 ⇒ 01:16:56.860 Samuel Roberts: This is how it’s specified in the database right now.

783 01:16:58.170 ⇒ 01:16:59.300 Samuel Roberts: If that makes sense.

784 01:17:17.030 ⇒ 01:17:19.589 Samuel Roberts: I’m adding a comment to that linear ticket.

785 01:17:30.750 ⇒ 01:17:33.100 Samuel Roberts: Oh… It’s really small, I don’.

786 01:18:06.720 ⇒ 01:18:10.409 Pranav Narahari: So, did all of these tickets just come in?

787 01:18:13.110 ⇒ 01:18:15.240 Pranav Narahari: I’m wondering why, like…

788 01:18:15.550 ⇒ 01:18:22.150 Pranav Narahari: like, is this just kind of how, like, Janiece’s usually works? Like, she just sees all of these problems and then kind of, like, gives them to you at the same time?

789 01:18:22.780 ⇒ 01:18:31.819 Casie Aviles: Yeah, that’s kind of what happens. So she doesn’t always, like, assign this to me, like, each day. So I think there are days where she would just check, like.

790 01:18:32.500 ⇒ 01:18:35.510 Casie Aviles: The triages that have come in, and then…

791 01:18:36.050 ⇒ 01:18:40.840 Casie Aviles: She, based on, like, whether it’s on us, she assigns it to me.

792 01:18:48.170 ⇒ 01:18:51.610 Mustafa Raja: You can just reply to the thread, Pranav, if you want to take a look.

793 01:18:59.280 ⇒ 01:19:00.060 Samuel Roberts: Okay.

794 01:19:10.130 ⇒ 01:19:17.270 Samuel Roberts: I’m gonna do another quick pass and see if… how many other entries are identical except for notes for everything except for Lawn Cruz now.

795 01:19:21.410 ⇒ 01:19:22.650 Samuel Roberts: Good planning.

796 01:19:33.100 ⇒ 01:19:34.499 Samuel Roberts: There we go.

797 01:19:42.250 ⇒ 01:19:43.890 Samuel Roberts: For the end real quick.

798 01:19:51.070 ⇒ 01:19:51.980 Samuel Roberts: I’m familiar.

799 01:19:53.400 ⇒ 01:19:55.590 Pranav Narahari: Are we sure that Janiece made those duplicate entries?

800 01:19:56.300 ⇒ 01:20:02.540 Samuel Roberts: No, I mean, they’re from back… way back when. I wanted to try to check that, but the audit log in the.

801 01:20:03.750 ⇒ 01:20:09.710 Casie Aviles: Yeah, that could also be, like, with how it was loaded when we were… Loading, like…

802 01:20:10.640 ⇒ 01:20:13.070 Casie Aviles: rows from the zips sheet.

803 01:20:14.650 ⇒ 01:20:14.965 Samuel Roberts: So…

804 01:20:15.670 ⇒ 01:20:20.090 Casie Aviles: Because the way it was structured in the spreadsheets was that… Yeah.

805 01:20:20.370 ⇒ 01:20:24.680 Casie Aviles: Yeah, it wasn’t, like, when… Yeah.

806 01:20:25.450 ⇒ 01:20:28.920 Casie Aviles: That’s… that’s kind of… that’s also, like, another thing.

807 01:20:29.410 ⇒ 01:20:31.840 Casie Aviles: Where it could have been how we ingested it.

808 01:21:40.300 ⇒ 01:21:41.480 Samuel Roberts: There’s the change?

809 01:21:44.860 ⇒ 01:21:47.420 Samuel Roberts: He’s like, I want to mention it.

810 01:21:47.850 ⇒ 01:21:52.480 Samuel Roberts: Alright, let’s see, let’s see… alright, so we have this assignment ID…

811 01:23:54.870 ⇒ 01:24:00.250 Pranav Narahari: I feel like we’re pretty close on this, right? Now we’re just kind of in the review stage for most of this,

812 01:24:02.840 ⇒ 01:24:06.439 Pranav Narahari: Sam, what’s, still kind of holding on,

813 01:24:06.570 ⇒ 01:24:09.219 Pranav Narahari: This do- this, duplication part.

814 01:24:09.970 ⇒ 01:24:20.509 Samuel Roberts: I’m just looking to see if there are other duplicates, but that might be something we can tackle after. I think the lawn is good, so that one is resolved, but there might be others, I’m trying to find if it’s…

815 01:24:20.510 ⇒ 01:24:21.800 Pranav Narahari: Just in the chat.

816 01:24:22.200 ⇒ 01:24:35.009 Pranav Narahari: what I’ll do, let’s first just move all of these things into, like, QA complete, and then we can kind of debrief about what are, like, some of the next steps we need to do, and then more, like, long-term, robust

817 01:24:35.290 ⇒ 01:24:37.290 Pranav Narahari: Kind of, features we need to add.

818 01:24:38.540 ⇒ 01:24:40.989 Casie Aviles: We, we also have these…

819 01:24:41.670 ⇒ 01:24:44.680 Casie Aviles: incorrect SQL query generated, so these are…

820 01:24:45.130 ⇒ 01:24:49.329 Casie Aviles: These require, like, prompt updates, so… which I’m updating now.

821 01:24:50.190 ⇒ 01:24:59.479 Casie Aviles: Because, yeah, basically the query that was generated was looking at the wrong… using the wrong keys, or, you know, looking at the wrong columns.

822 01:25:02.240 ⇒ 01:25:05.150 Samuel Roberts: And make sure… is that also when we update that, we’ll pull the right notes?

823 01:25:05.470 ⇒ 01:25:06.790 Samuel Roberts: Is that what you’re talking about?

824 01:25:07.420 ⇒ 01:25:10.530 Casie Aviles: It’s not just related to notes.

825 01:25:10.530 ⇒ 01:25:11.830 Samuel Roberts: Come with other things, too. Okay.

826 01:25:11.830 ⇒ 01:25:12.600 Casie Aviles: Yeah, yeah.

827 01:26:16.640 ⇒ 01:26:18.820 Samuel Roberts: Yeah… what the hell?

828 01:26:19.320 ⇒ 01:26:20.030 Samuel Roberts: Hmm.

829 01:26:42.950 ⇒ 01:26:46.170 Samuel Roberts: Yeah, I think there’s other duplicates, potentially.

830 01:26:46.790 ⇒ 01:26:48.040 Samuel Roberts: In here…

831 01:27:07.270 ⇒ 01:27:11.359 Pranav Narahari: Okay, Casey, we’re gonna have to do another redeploy after you update the system prompt, right?

832 01:27:12.310 ⇒ 01:27:13.060 Casie Aviles: Yeah.

833 01:27:13.550 ⇒ 01:27:14.190 Pranav Narahari: Okay.

834 01:27:15.080 ⇒ 01:27:18.740 Pranav Narahari: Can you maybe just share your screen so we can see what you’re changing it to?

835 01:27:20.460 ⇒ 01:27:23.470 Casie Aviles: Sure, I’m just having cursor that go through.

836 01:27:25.540 ⇒ 01:27:26.390 Casie Aviles: the prompt.

837 01:27:34.730 ⇒ 01:27:36.530 Casie Aviles: Yeah, I’m sharing my screen right.

838 01:27:49.190 ⇒ 01:27:51.170 Samuel Roberts: Did we run an update on the second?

839 01:27:51.350 ⇒ 01:27:52.370 Samuel Roberts: Of the month.

840 01:27:57.250 ⇒ 01:27:58.300 Samuel Roberts: Would have been…

841 01:28:03.250 ⇒ 01:28:07.659 Samuel Roberts: Not last week, but the week before, on the 2nd.

842 01:28:17.850 ⇒ 01:28:21.120 Pranav Narahari: Ted, whether you get your money back or not, if I…

843 01:28:42.290 ⇒ 01:28:46.419 Samuel Roberts: Yeah, Casey, do you know if we ran a big update for the DB on the 2nd, or…

844 01:28:48.360 ⇒ 01:28:50.400 Casie Aviles: Which? April?

845 01:28:50.940 ⇒ 01:28:51.500 Samuel Roberts: Yeah.

846 01:28:52.150 ⇒ 01:28:53.509 Samuel Roberts: Remember, like, 2 weeks ago?

847 01:28:55.760 ⇒ 01:29:02.439 Casie Aviles: I mean, I didn’t do, like, we didn’t do, like, any batch updates ourselves around that.

848 01:29:03.260 ⇒ 01:29:03.990 Samuel Roberts: Okay.

849 01:29:05.280 ⇒ 01:29:08.060 Samuel Roberts: They can do batch updates right now.

850 01:29:15.820 ⇒ 01:29:17.080 Samuel Roberts: Okay, let’s see…

851 01:31:00.040 ⇒ 01:31:02.990 Pranav Narahari: Sorry, I was just typing up a few messages.

852 01:31:03.720 ⇒ 01:31:07.799 Pranav Narahari: Casey, did you say you were gonna share your screen? To show what’s going on?

853 01:31:08.330 ⇒ 01:31:11.730 Casie Aviles: Yeah, it’s currently shared. Is it already, or there’s nothing…

854 01:31:12.950 ⇒ 01:31:14.120 Pranav Narahari: Oh, I’m not seeing it.

855 01:31:14.680 ⇒ 01:31:15.840 Pranav Narahari: Yeah, that’s on me.

856 01:31:16.330 ⇒ 01:31:18.020 Casie Aviles: Let me reshare them.

857 01:31:20.350 ⇒ 01:31:22.020 Samuel Roberts: No, Panav, you’re still sharing it.

858 01:31:23.220 ⇒ 01:31:24.659 Pranav Narahari: Oh, I’m sharing that’s why.

859 01:31:25.120 ⇒ 01:31:27.999 Samuel Roberts: Yeah, they had… I think you were both sharing, so you have to flip it to them.

860 01:31:28.000 ⇒ 01:31:29.900 Pranav Narahari: Okay, okay.

861 01:31:31.610 ⇒ 01:31:32.610 Pranav Narahari: Gotcha.

862 01:31:36.250 ⇒ 01:31:37.279 Pranav Narahari: Okay, cool.

863 01:31:50.220 ⇒ 01:31:52.810 Pranav Narahari: Okay, Sam, you’re just kind of waiting on this right now, right?

864 01:31:55.060 ⇒ 01:31:57.589 Samuel Roberts: Sorry, one cycle’s running on the query. Waiting on what?

865 01:31:58.890 ⇒ 01:32:03.849 Pranav Narahari: You’re just waiting on the system prompt update to be made?

866 01:32:05.880 ⇒ 01:32:16.050 Samuel Roberts: Yeah, well, it worked for the… the, the…

867 01:32:16.500 ⇒ 01:32:19.829 Samuel Roberts: lawn duplicates, but I think I’m finding more duplicates.

868 01:32:22.900 ⇒ 01:32:29.489 Pranav Narahari: Okay. I think, let’s just address… I mean, we’ll definitely have to address those too, Sam. Like, thanks for finding.

869 01:32:29.490 ⇒ 01:32:32.089 Samuel Roberts: Yeah, I just wanted to make sure that I wasn’t missing something while we were doing it.

870 01:32:32.090 ⇒ 01:32:32.630 Pranav Narahari: Yeah.

871 01:32:32.630 ⇒ 01:32:34.460 Samuel Roberts: I thought I was doing this, yeah.

872 01:32:34.460 ⇒ 01:32:43.650 Pranav Narahari: What, what Utam wants before, like, end of day, because I don’t know when you’re logging off today, is just kind of, like, a assessment of, like, what you found.

873 01:32:45.820 ⇒ 01:32:54.100 Pranav Narahari: Yeah, so just, like, a quick write-up of, like, the issue and, like, what you did to solve it. So, like, yeah. Casey, Sam, Mustafa, like…

874 01:32:54.240 ⇒ 01:32:59.990 Pranav Narahari: All three of y’all, if you could do that. Casey, keep doing what you’re doing, because, like, we wanna… we wanna fix the issue first, but…

875 01:33:00.350 ⇒ 01:33:04.079 Pranav Narahari: Sam, since we’re just kind of waiting on this, you could probably start writing yours.

876 01:33:05.350 ⇒ 01:33:08.480 Samuel Roberts: Yeah, okay, we’re really… I can.

877 01:33:08.480 ⇒ 01:33:13.749 Pranav Narahari: And you can add it just to this ticket, if you want. Just put it in the comments.

878 01:33:18.110 ⇒ 01:33:19.510 Samuel Roberts: The comments on the…

879 01:33:19.850 ⇒ 01:33:20.560 Pranav Narahari: The linear ticket?

880 01:33:27.460 ⇒ 01:33:32.089 Samuel Roberts: Wait, what… wait, which ticket are you looking at? Sorry, I was in the other one with the duplicate.

881 01:33:32.670 ⇒ 01:33:39.319 Pranav Narahari: Actually, you know what? Instead of the linear ticket, you can just put it in the Slack. That’s what I was thinking.

882 01:33:51.690 ⇒ 01:33:53.509 Samuel Roberts: No, I’m catching up on the slack here, yeah.

883 01:39:50.200 ⇒ 01:39:52.530 Pranav Narahari: Oh, did Janice just add a bunch more?

884 01:39:53.460 ⇒ 01:39:55.780 Casie Aviles: Yeah… -Oh.

885 01:39:57.010 ⇒ 01:39:58.300 Casie Aviles: I’m checking right now.

886 01:40:04.050 ⇒ 01:40:05.960 Casie Aviles: Okay, this one we already did.

887 01:40:05.960 ⇒ 01:40:10.039 Pranav Narahari: Part is also super automatable, like, just the assignment of things.

888 01:40:12.330 ⇒ 01:40:13.080 Casie Aviles: Hmm.

889 01:40:13.460 ⇒ 01:40:17.219 Pranav Narahari: Right? Like, because I think at this point, I already know, like, how the assignments work.

890 01:40:21.660 ⇒ 01:40:25.239 Pranav Narahari: Okay, that’s… A topic for another time.

891 01:40:29.750 ⇒ 01:40:32.810 Pranav Narahari: So is there more that you want to add to that ticket?

892 01:40:34.750 ⇒ 01:40:39.469 Casie Aviles: Yeah, there’s… I’m gonna… I think there’s 3 more that… that she added.

893 01:40:43.450 ⇒ 01:40:47.480 Pranav Narahari: Power offers for April… okay, interesting.

894 01:40:48.410 ⇒ 01:40:55.340 Pranav Narahari: Yeah, if you want to add those into the ticket right now, I can take a look while… UQA.

895 01:46:27.530 ⇒ 01:46:33.660 Pranav Narahari: So, Casey, on the Scott Glandin one, Which is…

896 01:46:33.880 ⇒ 01:46:38.639 Pranav Narahari: The handyman service one, which is basically… Two different answers.

897 01:46:40.050 ⇒ 01:46:48.849 Pranav Narahari: that they’re getting based on, you know, the intention… the intention, I think, semantically, both these questions are asking the same exact thing. So…

898 01:46:49.150 ⇒ 01:46:53.450 Pranav Narahari: It’s just an area where it’s being non-deterministic in the worst way.

899 01:46:53.550 ⇒ 01:46:54.800 Pranav Narahari: Yeah.

900 01:46:55.800 ⇒ 01:47:01.479 Pranav Narahari: why is that? Like, it’s hard for me to believe that the AI is just…

901 01:47:02.030 ⇒ 01:47:11.069 Pranav Narahari: not understanding… like, it looks like to me that there’s some conflicting information in the DB, where it’s saying handyman is not…

902 01:47:11.360 ⇒ 01:47:12.520 Pranav Narahari: a service.

903 01:47:13.030 ⇒ 01:47:14.310 Pranav Narahari: And that DP.

904 01:47:14.450 ⇒ 01:47:16.490 Pranav Narahari: In that zip code.

905 01:47:16.690 ⇒ 01:47:17.850 Pranav Narahari: So…

906 01:47:18.510 ⇒ 01:47:28.570 Pranav Narahari: I’m guessing the right way to interpret this, like, from what’s in the DB, is that there is a handyman that services that area, and there’s probably another handyman that doesn’t service that area.

907 01:47:30.670 ⇒ 01:47:33.250 Casie Aviles: Yeah, that, that’s…

908 01:47:33.550 ⇒ 01:47:38.760 Casie Aviles: Let’s see, like, this is, like, the source that they have, and this is what we copied, basically.

909 01:47:39.960 ⇒ 01:47:42.700 Pranav Narahari: Okay, let’s check our DB, though, right? Because that’s where…

910 01:47:42.700 ⇒ 01:47:43.499 Casie Aviles: Yeah, that’s true.

911 01:47:45.160 ⇒ 01:47:46.150 Casie Aviles: There’s…

912 01:48:24.430 ⇒ 01:48:25.150 Casie Aviles: Oh.

913 01:48:25.920 ⇒ 01:48:28.270 Casie Aviles: To decorate for this.

914 01:49:19.290 ⇒ 01:49:20.470 Pranav Narahari: Yeah, so I think…

915 01:49:22.510 ⇒ 01:49:29.709 Pranav Narahari: for… for this one, like, I’m interested to see, like, what is the SQL command that’s being generated?

916 01:49:30.960 ⇒ 01:49:34.650 Pranav Narahari: I think it’s doing… it’s just doing, like, a poor job of generating a SQL command.

917 01:49:36.120 ⇒ 01:49:42.160 Pranav Narahari: Because I can clearly see, like, that there’s two people that satisfy the handyman service and that zip code.

918 01:50:05.660 ⇒ 01:50:06.490 Samuel Roberts: The zip code?

919 01:50:08.430 ⇒ 01:50:11.560 Casie Aviles: It’s 78640.

920 01:50:20.040 ⇒ 01:50:22.800 Pranav Narahari: Is serviced is… no.

921 01:50:23.270 ⇒ 01:50:24.550 Pranav Narahari: Okay…

922 01:50:31.170 ⇒ 01:50:35.240 Pranav Narahari: So, when I look at… I’m just gonna share my screen real quick.

923 01:50:47.140 ⇒ 01:50:48.860 Pranav Narahari: I see these two people.

924 01:50:49.180 ⇒ 01:50:50.939 Pranav Narahari: For the zip code, this handyman?

925 01:50:56.670 ⇒ 01:50:58.750 Pranav Narahari: Where do I see the is serviced field?

926 01:51:00.850 ⇒ 01:51:08.950 Casie Aviles: Oh, it’s a different table that’s based on the sheet that I’m looking at right now. That’s not currently exposed in this

927 01:51:09.390 ⇒ 01:51:12.479 Casie Aviles: In the admin UI, since…

928 01:51:12.980 ⇒ 01:51:21.759 Casie Aviles: It’s mainly, like, we’re mainly, like, making updates to the assignments, primarily, but not, like, the… so that one isn’t exposed.

929 01:51:22.280 ⇒ 01:51:28.359 Casie Aviles: But I can share right now, like, this is the sheet that… We basically got it from…

930 01:51:32.110 ⇒ 01:51:41.420 Casie Aviles: So this is, like, the source, and we created a… Another table, dev service areas.

931 01:51:42.040 ⇒ 01:51:42.690 Samuel Roberts: Hmm.

932 01:51:43.610 ⇒ 01:51:44.840 Casie Aviles: So it looks like this.

933 01:51:47.670 ⇒ 01:51:53.790 Casie Aviles: Yeah, I was just asking if we service handyman in… 78640.

934 01:51:53.970 ⇒ 01:51:55.310 Casie Aviles: Or does ABC do handle?

935 01:51:55.310 ⇒ 01:51:57.549 Samuel Roberts: Well, with the generating query, I mean.

936 01:52:01.100 ⇒ 01:52:02.609 Samuel Roberts: From the log, yeah.

937 01:52:02.930 ⇒ 01:52:04.570 Casie Aviles: Yeah, it’s this one.

938 01:52:05.850 ⇒ 01:52:10.039 Samuel Roberts: And that return… so it’s pulling from… so it’s not looking at assignments, then?

939 01:52:10.490 ⇒ 01:52:13.069 Casie Aviles: Yes, it’s in the dev service areas.

940 01:52:13.190 ⇒ 01:52:13.760 Casie Aviles: table.

941 01:52:15.960 ⇒ 01:52:19.469 Pranav Narahari: Yeah, I think the issue here is that it’s not showing a user’s name.

942 01:52:19.980 ⇒ 01:52:32.990 Pranav Narahari: or, sorry, a person’s name. And so, the only context that the AI is given is that there’s conflicting information about the service, and if it’s… if that area is serviced.

943 01:52:36.410 ⇒ 01:52:42.249 Pranav Narahari: Does that make sense? Because I think if the AI got the person’s name, it would understand that each record is just basically…

944 01:52:42.940 ⇒ 01:52:46.030 Pranav Narahari: The… the assign… like, an assignment.

945 01:52:46.370 ⇒ 01:52:55.830 Pranav Narahari: And so, if somebody’s asking, is there a handyman in this area, it would see one person that says no, one person that says yes, and it would answer, yes, it is serviced in this area.

946 01:52:55.830 ⇒ 01:53:01.930 Samuel Roberts: People would say yes and no, the people would just be assigned, right? But why do we have this entry here that we’re seeing for the same zip? It’s yes and no.

947 01:53:02.980 ⇒ 01:53:04.279 Samuel Roberts: This is an issue.

948 01:53:06.610 ⇒ 01:53:10.009 Casie Aviles: Yeah, that could be the… could be a duplicate.

949 01:53:10.010 ⇒ 01:53:14.780 Samuel Roberts: Right, if it was gonna look, but it should specifically know that, like, this is not, or this is…

950 01:53:15.140 ⇒ 01:53:18.200 Pranav Narahari: Oh, I see, because it’s pulling from this, right? Yeah, yeah.

951 01:53:18.310 ⇒ 01:53:19.180 Pranav Narahari: Okay.

952 01:53:19.310 ⇒ 01:53:25.060 Samuel Roberts: If there is a yes, then theoretically there should be someone, and if there’s a no, then there shouldn’t be someone, but…

953 01:53:25.060 ⇒ 01:53:26.960 Pranav Narahari: Right, right, no, you’re… you’re absolutely.

954 01:53:26.960 ⇒ 01:53:29.849 Samuel Roberts: It has both, which is definitely an issue.

955 01:53:29.850 ⇒ 01:53:30.890 Pranav Narahari: Okay…

956 01:53:36.180 ⇒ 01:53:36.720 Casie Aviles: Yeah, there is.

957 01:53:36.720 ⇒ 01:53:40.009 Pranav Narahari: Casey, you want to throw this in the chat, too, so I can look with you?

958 01:53:42.350 ⇒ 01:53:43.549 Casie Aviles: Sh- yeah.

959 01:53:44.150 ⇒ 01:53:45.809 Casie Aviles: This is the link.

960 01:53:47.330 ⇒ 01:53:48.190 Pranav Narahari: Thanks.

961 01:53:52.180 ⇒ 01:53:56.320 Casie Aviles: Yeah, it should only be yes, 78640.

962 01:54:07.370 ⇒ 01:54:08.240 Casie Aviles: Let me see…

963 01:54:16.240 ⇒ 01:54:17.129 Casie Aviles: Demand.

964 01:54:19.590 ⇒ 01:54:22.750 Casie Aviles: 78640… yes.

965 01:54:45.820 ⇒ 01:54:52.220 Pranav Narahari: Okay, so it’s from dev services areas, joining dev services and dev locations.

966 01:54:52.420 ⇒ 01:54:54.959 Pranav Narahari: Are those two different sheets?

967 01:54:57.650 ⇒ 01:55:01.330 Casie Aviles: Yeah, these are diff- different tables, so… yeah.

968 01:55:02.850 ⇒ 01:55:05.029 Pranav Narahari: Different sheets, like, in the… okay.

969 01:55:07.500 ⇒ 01:55:08.740 Pranav Narahari: Okay, gotcha.

970 01:56:24.330 ⇒ 01:56:27.930 Pranav Narahari: Casey, When you run that query.

971 01:56:28.580 ⇒ 01:56:36.609 Pranav Narahari: Let’s get a little bit more information than just the… The service name?

972 01:56:39.830 ⇒ 01:56:41.629 Casie Aviles: Okay.

973 01:56:45.420 ⇒ 01:56:49.720 Pranav Narahari: Yeah, so here we’re getting service name, zip is service, can we also just get…

974 01:56:50.590 ⇒ 01:56:54.500 Pranav Narahari: You can just do select star, honestly, if you want, just so we can see everything.

975 01:56:55.420 ⇒ 01:56:58.229 Pranav Narahari: Cause, like, we need to know why there’s two different things.

976 01:56:59.930 ⇒ 01:57:00.830 Pranav Narahari: Okay.

977 01:57:05.030 ⇒ 01:57:06.990 Pranav Narahari: Does this give any more context?

978 01:57:10.930 ⇒ 01:57:12.930 Pranav Narahari: Branch Chem Free.

979 01:57:15.040 ⇒ 01:57:20.150 Casie Aviles: who… I see… okay, I see now. It’s because there are two different…

980 01:57:20.930 ⇒ 01:57:24.640 Casie Aviles: locations. One is CAM-free, and one is not CAM-free.

981 01:57:26.090 ⇒ 01:57:29.519 Casie Aviles: So, if we look here in the source sheet.

982 01:57:30.220 ⇒ 01:57:36.520 Casie Aviles: And then handyman… so it’s gonna say yes for 78640, yes, Handyman.

983 01:57:36.890 ⇒ 01:57:43.980 Casie Aviles: But if we go to… Chem-free, this Chem-free sheet here, it’s gonna say handyman, no.

984 01:57:46.070 ⇒ 01:57:52.389 Casie Aviles: So technically, it’s correct, but it’s confusing, like, why are there… why is one yes, why is one no?

985 01:57:55.990 ⇒ 01:57:57.040 Pranav Narahari: Okay…

986 01:57:57.580 ⇒ 01:58:07.430 Pranav Narahari: So, what’s the difference between these two? Oh, this is for chem-free, not for ABC, so why are we even… why do we even bring in the chem-free service areas, then?

987 01:58:10.900 ⇒ 01:58:18.299 Casie Aviles: I think it’s just trying to give, like, everything in the records, but it’s not doing a good job at…

988 01:58:18.490 ⇒ 01:58:21.539 Casie Aviles: Clarifying which one is chem-free and which one isn’t.

989 01:58:23.280 ⇒ 01:58:24.669 Pranav Narahari: I see, okay.

990 01:58:25.820 ⇒ 01:58:31.819 Casie Aviles: But yeah, like, here it does say ABC, so… I think we could…

991 01:58:32.330 ⇒ 01:58:35.440 Casie Aviles: We need to fix the querying again, such that

992 01:58:36.070 ⇒ 01:58:39.880 Casie Aviles: It takes a look at just the… Non-chem-free branch.

993 01:58:42.000 ⇒ 01:58:44.330 Casie Aviles: Because it also took chemistry.

994 01:58:44.620 ⇒ 01:58:48.390 Casie Aviles: So yeah, that’s… that’s the part that we… that I think needs to be fixed.

995 01:58:52.610 ⇒ 01:58:53.260 Pranav Narahari: Okay.

996 01:58:58.160 ⇒ 01:59:00.999 Pranav Narahari: Yeah, that makes sense. So where do we fix that?

997 01:59:02.130 ⇒ 01:59:03.510 Pranav Narahari: In the system prompt?

998 01:59:04.930 ⇒ 01:59:07.239 Casie Aviles: Yeah, I would… yeah, that’s where I would go.

999 01:59:12.820 ⇒ 01:59:13.360 Samuel Roberts: Yes.

1000 01:59:29.170 ⇒ 01:59:31.490 Pranav Narahari: Yeah, man, I’m just gonna add a little bit more context here.

1001 01:59:35.690 ⇒ 01:59:42.210 Pranav Narahari: So, ABC versus… Temphrey is just the… the company?

1002 01:59:43.840 ⇒ 01:59:50.439 Casie Aviles: So when they specify ABC only, then… We should not… query for chemistry at all.

1003 01:59:52.220 ⇒ 01:59:53.040 Pranav Narahari: Yep.

1004 02:00:01.070 ⇒ 02:00:04.459 Pranav Narahari: I also feel like a lot of these…

1005 02:00:05.370 ⇒ 02:00:09.710 Pranav Narahari: problems could have just been solved by the AI if instead of…

1006 02:00:10.170 ⇒ 02:00:14.109 Pranav Narahari: Just outputting certain fields if it just did a select star on the command.

1007 02:00:14.400 ⇒ 02:00:17.320 Pranav Narahari: And that’s not gonna add too much context, I feel.

1008 02:00:21.560 ⇒ 02:00:26.100 Casie Aviles: Oh, like, so we’re gonna get everything, and then we’ll let the AI kind of do the filtering?

1009 02:00:26.670 ⇒ 02:00:27.640 Pranav Narahari: Yeah.

1010 02:00:27.780 ⇒ 02:00:30.230 Pranav Narahari: Because the AI’s still doing the filtering.

1011 02:00:30.490 ⇒ 02:00:33.179 Pranav Narahari: Right? Like, but it’s doing the filtering.

1012 02:00:33.600 ⇒ 02:00:38.579 Pranav Narahari: It’s doing a filtering on the SQL command

1013 02:00:38.920 ⇒ 02:00:47.400 Pranav Narahari: part. If we just prevent it, hey, don’t do any filtering on the SQL command, just do select star on whatever tables you want to join.

1014 02:00:47.950 ⇒ 02:00:49.840 Pranav Narahari: Or whatever.

1015 02:00:49.840 ⇒ 02:00:50.440 Casie Aviles: Okay.

1016 02:00:51.780 ⇒ 02:00:54.500 Pranav Narahari: I don’t know, maybe that’s gonna cause other…

1017 02:00:54.500 ⇒ 02:00:59.140 Casie Aviles: Yeah, but I do think we still need to tune that, if that’s… if…

1018 02:00:59.270 ⇒ 02:01:03.930 Casie Aviles: That would be how our approach would be, because it might just return, like, so much noise.

1019 02:01:06.860 ⇒ 02:01:11.240 Casie Aviles: I mean, like, they’re in the response, so we’ll have to probably tune that as well.

1020 02:01:12.040 ⇒ 02:01:13.989 Pranav Narahari: Yeah, that’s a good point. Yeah.

1021 02:01:15.710 ⇒ 02:01:19.800 Pranav Narahari: It just needs to do a better job, I guess, of generating these SQL queries.

1022 02:01:21.460 ⇒ 02:01:27.540 Pranav Narahari: Maybe we need to be… In the system prompt, are we giving it a description of

1023 02:01:27.730 ⇒ 02:01:29.149 Pranav Narahari: Each of the fields.

1024 02:01:31.490 ⇒ 02:01:34.769 Casie Aviles: Each of the fields, or, like, beginning columns.

1025 02:01:34.800 ⇒ 02:01:37.030 Pranav Narahari: Yeah, each of the table columns.

1026 02:01:37.610 ⇒ 02:01:43.770 Pranav Narahari: Like, does it have the… DB… the DB schema with, like.

1027 02:01:43.770 ⇒ 02:01:45.180 Casie Aviles: Yeah, just the schema.

1028 02:01:45.700 ⇒ 02:01:50.579 Pranav Narahari: Okay, for the schema, I think for each of the columns, we need to give it a short description.

1029 02:01:51.900 ⇒ 02:01:53.289 Casie Aviles: Okay, okay.

1030 02:01:54.710 ⇒ 02:01:55.370 Pranav Narahari: Yeah.

1031 02:02:01.740 ⇒ 02:02:02.910 Pranav Narahari: Sam, what do you think?

1032 02:02:04.220 ⇒ 02:02:06.530 Samuel Roberts: Sorry, I was listening, I was typing up this thing, what was that?

1033 02:02:08.060 ⇒ 02:02:16.169 Pranav Narahari: So right now, for… the… how we’re giving the AI context is just giving it the raw DB schema.

1034 02:02:16.530 ⇒ 02:02:26.639 Pranav Narahari: what we need to do is give it the DB schema, but then for each of the columns, for each table, we need to give a description of what that column is.

1035 02:02:31.690 ⇒ 02:02:35.650 Samuel Roberts: Yeah, because it’s not… Selecting the right ones at this point, is that the issue?

1036 02:02:35.650 ⇒ 02:02:36.300 Pranav Narahari: Yeah.

1037 02:02:36.300 ⇒ 02:02:37.560 Samuel Roberts: Terrific. Yeah. Yeah.

1038 02:02:38.180 ⇒ 02:02:40.940 Samuel Roberts: Yeah, description would definitely be helpful.

1039 02:02:41.270 ⇒ 02:02:48.010 Samuel Roberts: that even could be something we add to the DB, and then, like, export the schema kind of thing, or just do it in the prompt, either way.

1040 02:04:24.060 ⇒ 02:04:28.449 Samuel Roberts: Casey, I’m making a quick ticket that has all the context of the duplicate stuff I was just searching.

1041 02:04:29.060 ⇒ 02:04:30.420 Samuel Roberts: After the lock.

1042 02:04:31.340 ⇒ 02:04:36.240 Samuel Roberts: It’ll have the SQL that I used to generate the CSV that it exported.

1043 02:04:36.390 ⇒ 02:04:40.399 Samuel Roberts: And, I think we… basically, I just want to do a more…

1044 02:04:41.170 ⇒ 02:04:53.079 Samuel Roberts: closer pass on all the entries, and just make sure that it is, in fact, what I think it is, before I just go in and blow them away. Because the lawn care ones were, like, 20, this is, like, over 1,000, so it would be, like, 500 different ones to remove. So I don’t think it’s…

1045 02:04:53.080 ⇒ 02:04:53.520 Casie Aviles: Okay.

1046 02:04:53.520 ⇒ 02:04:57.040 Samuel Roberts: I just want to make sure we do it before we just remove them.

1047 02:04:57.260 ⇒ 02:04:58.889 Samuel Roberts: And that I’m not missing something else.

1048 02:04:59.500 ⇒ 02:05:00.520 Casie Aviles: Okay, okay.

1049 02:05:00.960 ⇒ 02:05:04.449 Samuel Roberts: I’m just making a quick ticket, it should have everything you need there.

1050 02:06:02.310 ⇒ 02:06:05.969 Samuel Roberts: Alright, so I’m typing up a note about the doubled-up responses for the lawn care.

1051 02:06:14.090 ⇒ 02:06:19.799 Samuel Roberts: Is there anything else I should talk about here? Or, Casey, you can type up one for…

1052 02:06:20.270 ⇒ 02:06:25.679 Casie Aviles: Yeah, I can… I can handle, like, the prompt changes for the query generation.

1053 02:06:25.880 ⇒ 02:06:26.500 Samuel Roberts: Okay.

1054 02:06:26.830 ⇒ 02:06:29.089 Samuel Roberts: Oh, what was the other Dublin one? Okay, yeah.

1055 02:06:31.380 ⇒ 02:06:38.400 Casie Aviles: I’m just… what’s… like, I’m not sure, like, if similar, like.

1056 02:06:38.900 ⇒ 02:06:47.510 Casie Aviles: Issues will happen again, like, in the future, like, in terms of, like, querying columns or, like, specific keywords, right?

1057 02:06:47.740 ⇒ 02:06:54.510 Casie Aviles: So… I’m wondering if we should, like… there should be another ticket for just…

1058 02:06:56.170 ⇒ 02:07:02.060 Casie Aviles: thinking about that, or… I’m not entirely sure yet, like, what the best… I’m fixing.

1059 02:07:02.060 ⇒ 02:07:07.580 Pranav Narahari: Yeah, I think what happens next now is we just need to…

1060 02:07:07.870 ⇒ 02:07:17.869 Pranav Narahari: look into this full system with, like… and we’re kind of doing it now, but just a little bit after the fact. Now, once we’ve just fixed… we’ve adjusted all these comments, let’s…

1061 02:07:17.970 ⇒ 02:07:24.009 Pranav Narahari: let’s look at each stage of the retrieval process. When we are…

1062 02:07:24.130 ⇒ 02:07:36.570 Pranav Narahari: delegating to different agents, for what a reason. What’s in the system prompt? How can that be improved? I think I’m already writing a bunch of, like, comments here on the… in the activity.

1063 02:07:36.750 ⇒ 02:07:41.799 Pranav Narahari: Sam, I think you and Uten will probably talk a little bit about, like, okay, how do we modify

1064 02:07:42.030 ⇒ 02:07:47.480 Pranav Narahari: this system… So then, there’s less of these, like, issues with.

1065 02:07:47.480 ⇒ 02:07:52.220 Samuel Roberts: What, yeah. Casey, can you real quick look at, ABC 2531?

1066 02:07:53.530 ⇒ 02:07:54.190 Casie Aviles: Okay.

1067 02:07:54.910 ⇒ 02:07:59.099 Samuel Roberts: I’m just trying to make sure, because I looked at the lawn care one, I want to make sure I’m seeing…

1068 02:07:59.480 ⇒ 02:08:00.469 Samuel Roberts: This one right.

1069 02:08:00.470 ⇒ 02:08:02.090 Casie Aviles: 13. Cool.

1070 02:08:05.640 ⇒ 02:08:07.219 Casie Aviles: So this is another discipline.

1071 02:08:07.650 ⇒ 02:08:08.200 Samuel Roberts: Yeah.

1072 02:08:09.000 ⇒ 02:08:13.900 Samuel Roberts: So, this one… It doubled it up, if you scroll down.

1073 02:08:16.330 ⇒ 02:08:24.029 Samuel Roberts: But it looks like… like, Ed Matthews here is different services, so is it…

1074 02:08:25.760 ⇒ 02:08:28.929 Casie Aviles: Oh, okay. Yeah, that’s, like, that’s what I was…

1075 02:08:29.290 ⇒ 02:08:32.730 Casie Aviles: Talking about at first, before we found out the…

1076 02:08:33.170 ⇒ 02:08:39.769 Casie Aviles: to duplicate records, but this one, technically, he… Ed Matthews just has, like, different services.

1077 02:08:40.010 ⇒ 02:08:40.750 Casie Aviles: Right, one.

1078 02:08:40.750 ⇒ 02:08:47.119 Samuel Roberts: So, is this a bug, or is this just they didn’t like the way it came back because of the way they have their stuff organized?

1079 02:08:47.910 ⇒ 02:08:51.410 Samuel Roberts: Map that differently, but I don’t…

1080 02:08:51.890 ⇒ 02:09:00.640 Casie Aviles: I think it’s just a matter of formatting this such that the services would just say… it would just be, like, one bullet for Ed Matthews, and then…

1081 02:09:00.940 ⇒ 02:09:02.950 Casie Aviles: Zero dent and sea termite.

1082 02:09:03.660 ⇒ 02:09:04.740 Samuel Roberts: Yeah. Okay.

1083 02:09:05.820 ⇒ 02:09:06.639 Samuel Roberts: Because the data here.

1084 02:09:06.640 ⇒ 02:09:12.740 Pranav Narahari: That’s not a central dock issue, then, right? I think we put it here in central dock, but it’s actually, in doubling.

1085 02:09:13.910 ⇒ 02:09:16.800 Samuel Roberts: Well, it’s not… Really doubling, though, is my point.

1086 02:09:17.470 ⇒ 02:09:19.599 Samuel Roberts: It’s returning the data that is in there.

1087 02:09:20.830 ⇒ 02:09:26.510 Pranav Narahari: Okay, so I’m just gonna create a new… Section just called, like… response.

1088 02:09:26.810 ⇒ 02:09:28.170 Pranav Narahari: formatting.

1089 02:09:29.490 ⇒ 02:09:33.199 Samuel Roberts: Yeah, that’s more of a… yeah, if they don’t like the way that’s coming out, that we can update that.

1090 02:09:37.970 ⇒ 02:09:47.039 Samuel Roberts: Yeah, the lawn care ones, there were the double ones with the dates that were… or the days of the week that were being ignored by the LLM. That seems to be resolved. This one is just a formatting thing.

1091 02:09:53.990 ⇒ 02:09:54.530 Samuel Roberts: Okay.

1092 02:09:54.530 ⇒ 02:09:57.139 Pranav Narahari: Casey, did you look at both of these, Marty Cruz ones?

1093 02:10:00.280 ⇒ 02:10:01.040 Casie Aviles: See?

1094 02:10:04.360 ⇒ 02:10:08.100 Pranav Narahari: Mowing crew for 7878378?

1095 02:10:11.910 ⇒ 02:10:14.579 Casie Aviles: This one, on their list of tickets, right?

1096 02:10:14.870 ⇒ 02:10:16.430 Casie Aviles: 078…

1097 02:10:16.430 ⇒ 02:10:21.779 Pranav Narahari: Yeah. This is just for Janiece to do, I think, right? She… even in the feedback, it’s just giving updates.

1098 02:10:24.740 ⇒ 02:10:33.840 Casie Aviles: Yeah, it’s just missing… Now, I haven’t checked this, because I was QA-ing the other tickets.

1099 02:10:33.840 ⇒ 02:10:39.230 Pranav Narahari: Okay, that’s fine. I was just asking, because I can move these over. I think,

1100 02:10:39.610 ⇒ 02:10:46.750 Pranav Narahari: Okay, that one, I’ll just… If… I’ll just message Janice and say…

1101 02:10:47.080 ⇒ 02:10:48.889 Pranav Narahari: I’ll put in the comments, Janice.

1102 02:10:49.360 ⇒ 02:10:52.380 Pranav Narahari: Looks like a DB update.

1103 02:10:52.720 ⇒ 02:10:53.520 Pranav Narahari: on.

1104 02:10:54.880 ⇒ 02:10:55.679 Samuel Roberts: For which one?

1105 02:10:58.800 ⇒ 02:11:07.479 Pranav Narahari: It’s the first one for Mari Cruz in that, parent ticket.

1106 02:11:08.600 ⇒ 02:11:10.050 Pranav Narahari: Do you want a ticket number?

1107 02:11:11.070 ⇒ 02:11:13.600 Samuel Roberts: Yeah, if you just… I just want to take a look, I’m just trying to understand what that one is.

1108 02:11:13.600 ⇒ 02:11:16.019 Pranav Narahari: I’ll put it in the chat. Here’s a link.

1109 02:11:16.020 ⇒ 02:11:16.590 Samuel Roberts: Okay.

1110 02:11:17.030 ⇒ 02:11:17.790 Samuel Roberts: Thank you.

1111 02:11:21.710 ⇒ 02:11:23.310 Samuel Roberts: C256…

1112 02:11:37.650 ⇒ 02:11:38.890 Samuel Roberts: Oh, I see, okay.

1113 02:11:41.350 ⇒ 02:11:44.740 Pranav Narahari: Wait… Casey, is this something that we need to do on our end?

1114 02:11:45.920 ⇒ 02:11:49.759 Casie Aviles: I’m checking, like, if it really is our end or not.

1115 02:11:52.090 ⇒ 02:11:54.190 Casie Aviles: Okay. I just updated, I just linked it here.

1116 02:11:54.690 ⇒ 02:11:57.910 Pranav Narahari: Yeah, I won’t tag Janice. I’m just gonna put my comment in.

1117 02:12:06.190 ⇒ 02:12:08.439 Pranav Narahari: Yeah, I guess we should first check.

1118 02:12:09.930 ⇒ 02:12:11.550 Pranav Narahari: Yeah, good point.

1119 02:12:14.880 ⇒ 02:12:16.010 Samuel Roberts: And then who do they think?

1120 02:12:23.080 ⇒ 02:12:24.330 Casie Aviles: enrollment group.

1121 02:12:27.670 ⇒ 02:12:29.260 Pranav Narahari: Yeah, there’s no assignments found.

1122 02:12:29.470 ⇒ 02:12:30.839 Pranav Narahari: For that zip code.

1123 02:12:32.630 ⇒ 02:12:33.530 Pranav Narahari: I’m missing a foot.

1124 02:12:33.710 ⇒ 02:12:35.160 Casie Aviles: Yeah, it’s missing.

1125 02:12:35.840 ⇒ 02:12:36.430 Pranav Narahari: Yeah.

1126 02:12:38.420 ⇒ 02:12:40.850 Pranav Narahari: Okay, so we can probably just tag Janice and say that, right?

1127 02:12:44.610 ⇒ 02:12:45.390 Casie Aviles: Here.

1128 02:12:48.230 ⇒ 02:12:51.079 Pranav Narahari: Here, I can just show you real quick, so you can just fact check it.

1129 02:12:52.510 ⇒ 02:12:53.930 Pranav Narahari: I’m just gonna share my screen.

1130 02:12:55.170 ⇒ 02:12:57.270 Pranav Narahari: Yeah, I just put zip code here.

1131 02:12:57.380 ⇒ 02:13:01.740 Pranav Narahari: And then… Service, lawn mowing, nothing pops up.

1132 02:13:03.830 ⇒ 02:13:04.980 Casie Aviles: Okay, yeah.

1133 02:13:08.650 ⇒ 02:13:09.230 Pranav Narahari: Okay.

1134 02:13:10.230 ⇒ 02:13:14.709 Pranav Narahari: Alright, so that one’s good. I will… you wanna leave a comment for that?

1135 02:13:17.690 ⇒ 02:13:18.540 Casie Aviles: Sure, sure.

1136 02:13:18.960 ⇒ 02:13:19.640 Pranav Narahari: Go.

1137 02:13:25.650 ⇒ 02:13:28.529 Pranav Narahari: So, for review just means for Janice’s review, right?

1138 02:13:29.890 ⇒ 02:13:32.139 Pranav Narahari: Okay, so I’m just gonna update that…

1139 02:14:12.190 ⇒ 02:14:16.359 Pranav Narahari: Is Tree Crew? Is that the Tree Airspace service?

1140 02:14:18.280 ⇒ 02:14:23.289 Casie Aviles: Hmm, this looks… the triage page seems to be a new service that Janice added.

1141 02:14:23.840 ⇒ 02:14:25.050 Casie Aviles: So…

1142 02:14:26.680 ⇒ 02:14:34.469 Pranav Narahari: But that’s in the… yeah, okay, it’s in the lawn department, that makes sense. When they say Austin, and they don’t provide a zip code, what do we do?

1143 02:14:35.130 ⇒ 02:14:36.729 Casie Aviles: We look at the branch.

1144 02:14:36.930 ⇒ 02:14:38.360 Casie Aviles: So we have.

1145 02:14:38.360 ⇒ 02:14:44.150 Pranav Narahari: Branch Austin. Okay, gotcha. Okay, so there’s nothing in Branch Austin for tree air spade.

1146 02:14:45.190 ⇒ 02:14:45.940 Casie Aviles: Okay.

1147 02:14:56.080 ⇒ 02:15:00.730 Pranav Narahari: Oh, you already updated this. Okay, cool. So, looks like you already did your QA on this.

1148 02:15:01.390 ⇒ 02:15:04.140 Pranav Narahari: Okay, I’m gonna move this for Janiece’s review, then.

1149 02:15:07.450 ⇒ 02:15:08.340 Pranav Narahari: Right?

1150 02:15:09.830 ⇒ 02:15:13.229 Casie Aviles: Oh, which, which one? Sorry, I was looking at a different ticket.

1151 02:15:13.660 ⇒ 02:15:14.849 Casie Aviles: For those…

1152 02:15:16.140 ⇒ 02:15:17.460 Pranav Narahari: 2571.

1153 02:15:17.460 ⇒ 02:15:18.750 Casie Aviles: 2571.

1154 02:15:19.160 ⇒ 02:15:19.710 Pranav Narahari: Yo.

1155 02:15:22.580 ⇒ 02:15:25.690 Casie Aviles: 2471… Okay.

1156 02:15:27.160 ⇒ 02:15:30.350 Casie Aviles: offers, and it’s not providing any crew info.

1157 02:15:33.170 ⇒ 02:15:36.479 Casie Aviles: Oh, okay, yeah, it’s missing the cruise.

1158 02:15:41.550 ⇒ 02:15:44.889 Casie Aviles: So, she would need… she would need to update.

1159 02:15:46.780 ⇒ 02:15:49.589 Pranav Narahari: Yeah, so I put that in the… for Genesis review.

1160 02:15:51.140 ⇒ 02:15:52.130 Pranav Narahari: section.

1161 02:15:52.920 ⇒ 02:15:53.570 Casie Aviles: Okay.

1162 02:15:54.480 ⇒ 02:15:55.840 Pranav Narahari: Yup, okay, cool.

1163 02:15:56.150 ⇒ 02:16:01.139 Pranav Narahari: Alright, so… Now let’s.

1164 02:16:02.250 ⇒ 02:16:03.849 Samuel Roberts: Yeah, where are we?

1165 02:16:03.850 ⇒ 02:16:06.220 Pranav Narahari: Go through the rest of it and investigate it.

1166 02:16:07.630 ⇒ 02:16:09.089 Samuel Roberts: Yeah, what else do we go through?

1167 02:16:10.100 ⇒ 02:16:13.400 Pranav Narahari: So, incorrect SQL generated,

1168 02:16:15.870 ⇒ 02:16:21.549 Pranav Narahari: for Scott Glandon, I think for this one, we know that what we need to do is…

1169 02:16:24.270 ⇒ 02:16:32.819 Pranav Narahari: update the… the system prompt for the section where we’re just adding the db schema raw, and give it a…

1170 02:16:33.070 ⇒ 02:16:35.880 Pranav Narahari: Description per field and per table.

1171 02:16:37.070 ⇒ 02:16:38.969 Samuel Roberts: Yeah, that’s probably the best idea.

1172 02:16:39.700 ⇒ 02:16:40.320 Pranav Narahari: Okay.

1173 02:16:41.280 ⇒ 02:16:44.050 Samuel Roberts: And hopefully that… I mean, try… tested that, actually.

1174 02:16:44.219 ⇒ 02:16:46.090 Samuel Roberts: You know, gives enough context for it.

1175 02:16:46.540 ⇒ 02:16:47.860 Pranav Narahari: Yeah, yeah.

1176 02:16:49.959 ⇒ 02:16:54.280 Pranav Narahari: Which ones can we QA now, because we feel confident that they’re fixed?

1177 02:16:54.870 ⇒ 02:17:00.580 Samuel Roberts: I was gonna say, 2533 is the one I have a screenshot in there of me asking the same question.

1178 02:17:01.090 ⇒ 02:17:01.939 Samuel Roberts: And it…

1179 02:17:01.940 ⇒ 02:17:03.400 Pranav Narahari: Okay, and it worked.

1180 02:17:06.340 ⇒ 02:17:06.840 Pranav Narahari: Right.

1181 02:17:07.830 ⇒ 02:17:10.120 Pranav Narahari: Okay, I’m gonna move that to QA complete.

1182 02:17:16.309 ⇒ 02:17:17.799 Samuel Roberts: Should we move the status to?

1183 02:17:21.070 ⇒ 02:17:23.880 Samuel Roberts: Are we putting them into the review, is that what we’re doing?

1184 02:17:25.309 ⇒ 02:17:26.639 Pranav Narahari: Yeah…

1185 02:17:26.879 ⇒ 02:17:34.699 Pranav Narahari: I… I think so, yeah. Casey, what’s usually your process? You, once you do your QA, you put it into review?

1186 02:17:35.400 ⇒ 02:17:39.850 Casie Aviles: Yeah, I put it into client review, and then I reassigned it to Janiece.

1187 02:17:40.820 ⇒ 02:17:43.510 Pranav Narahari: Okay. Why are some of these, say, complete?

1188 02:17:44.350 ⇒ 02:17:45.020 Pranav Narahari: Or dumb.

1189 02:17:45.020 ⇒ 02:17:51.080 Casie Aviles: I think Janice said that to Don. She was looking over at some of these already.

1190 02:17:51.440 ⇒ 02:17:52.609 Casie Aviles: Yeah, that’s her.

1191 02:17:54.580 ⇒ 02:17:56.859 Casie Aviles: Nice. You can check the activity as well.

1192 02:17:57.760 ⇒ 02:18:00.639 Pranav Narahari: Okay. Yeah, Sam, I’ll let you update those statuses.

1193 02:18:01.420 ⇒ 02:18:02.969 Samuel Roberts: Okay, so update and reassign.

1194 02:18:11.350 ⇒ 02:18:16.209 Pranav Narahari: For the Scott Glandon one, that one is still… we need to do stuff on our end.

1195 02:18:17.270 ⇒ 02:18:18.880 Pranav Narahari: We’ll put that in progress.

1196 02:18:28.870 ⇒ 02:18:31.790 Pranav Narahari: Here’s our feedback from Patricia Vela.

1197 02:18:41.139 ⇒ 02:18:45.610 Pranav Narahari: What is the minimum for landscape? This would be in the central dock, right?

1198 02:18:47.629 ⇒ 02:18:51.739 Casie Aviles: Hmm, yeah, I believe that should be a central talk query.

1199 02:18:52.889 ⇒ 02:18:58.529 Samuel Roberts: And they’re saying the landscape estimates is 1,000, so it looks like the central dock is not accurate, then?

1200 02:19:00.199 ⇒ 02:19:02.119 Samuel Roberts: Or missing that information, maybe?

1201 02:19:07.459 ⇒ 02:19:11.179 Pranav Narahari: Yeah, Casey, so what’s usually your process for these? You just go into the central doc and update it?

1202 02:19:13.180 ⇒ 02:19:17.309 Casie Aviles: I don’t do, like, direct updates to the central dock.

1203 02:19:20.080 ⇒ 02:19:29.830 Casie Aviles: Yeah, I don’t really, like, do that for, yeah, for those kinds of questions, and sometimes it’s just them that addresses that.

1204 02:19:30.520 ⇒ 02:19:33.529 Casie Aviles: So they, they would make changes to the central dock.

1205 02:19:35.250 ⇒ 02:19:37.079 Pranav Narahari: Why did you assign it to you?

1206 02:19:37.450 ⇒ 02:19:39.030 Casie Aviles: Which, which ticket?

1207 02:19:39.920 ⇒ 02:19:44.179 Pranav Narahari: Patricia Velo, it is 2551.

1208 02:19:46.080 ⇒ 02:19:48.170 Casie Aviles: 25551.

1209 02:19:53.110 ⇒ 02:19:58.570 Casie Aviles: Okay… 02451.

1210 02:19:58.740 ⇒ 02:20:00.230 Casie Aviles: Looking at the wrong one.

1211 02:20:04.020 ⇒ 02:20:06.720 Casie Aviles: What is the minimum for a landscape?

1212 02:20:09.730 ⇒ 02:20:17.770 Casie Aviles: Hmm, I guess that’s something that… that I need to clarify, like, is this something I have to edit since…

1213 02:20:18.590 ⇒ 02:20:24.629 Casie Aviles: Because what… from what I understand, like, they would… they would flag that something is…

1214 02:20:25.730 ⇒ 02:20:28.120 Casie Aviles: They would flag that,

1215 02:20:28.960 ⇒ 02:20:32.690 Casie Aviles: Response is wrong, so they would go ahead and

1216 02:20:33.400 ⇒ 02:20:40.890 Casie Aviles: So Janiece would assign that to a trainer who handles, like, the specific central dock, right?

1217 02:20:41.240 ⇒ 02:20:43.839 Casie Aviles: And then they go ahead and make the change.

1218 02:20:44.370 ⇒ 02:20:46.690 Casie Aviles: And then they asked me to review if…

1219 02:20:46.880 ⇒ 02:20:50.550 Casie Aviles: The change looks good, or if it’s returning the right answer.

1220 02:20:52.840 ⇒ 02:20:56.689 Samuel Roberts: So this is the flow that will get, kind of, supplanted by the…

1221 02:20:57.000 ⇒ 02:20:58.940 Casie Aviles: Yeah, yeah.

1222 02:21:01.030 ⇒ 02:21:03.320 Casie Aviles: So this is… yeah, I’m not too sure about…

1223 02:21:03.320 ⇒ 02:21:03.689 Pranav Narahari: if this was…

1224 02:21:03.690 ⇒ 02:21:04.779 Casie Aviles: assigned to me.

1225 02:21:05.250 ⇒ 02:21:08.290 Pranav Narahari: You should put a comment there, probably just, like, asking Janiece, and…

1226 02:21:08.660 ⇒ 02:21:10.009 Samuel Roberts: I think so. Okay.

1227 02:21:10.010 ⇒ 02:21:10.830 Pranav Narahari: relaying that.

1228 02:21:15.020 ⇒ 02:21:16.779 Samuel Roberts: What else is there under?

1229 02:21:16.780 ⇒ 02:21:23.380 Pranav Narahari: For the response format issue, I think that is still ongoing, right? You need to make an update on that.

1230 02:21:25.080 ⇒ 02:21:28.599 Samuel Roberts: Sorry, hold on, which one was that? Let me go back.

1231 02:21:28.800 ⇒ 02:21:30.040 Pranav Narahari: 2531.

1232 02:21:30.810 ⇒ 02:21:33.319 Samuel Roberts: Yeah, hold on, it’s… ugh, I hate the way.

1233 02:21:33.790 ⇒ 02:21:37.280 Samuel Roberts: There does this back button, there we go. Okay,

1234 02:21:42.170 ⇒ 02:21:46.509 Samuel Roberts: Oh, this one, yeah, yeah, yeah, so it’s not really a double, it’s a formatting thing.

1235 02:21:47.650 ⇒ 02:21:49.280 Samuel Roberts: Right, okay.

1236 02:21:57.940 ⇒ 02:21:58.960 Samuel Roberts: So we want to make a…

1237 02:21:59.200 ⇒ 02:22:02.780 Samuel Roberts: Prompt update there, or… do we want to do…

1238 02:22:04.930 ⇒ 02:22:11.619 Samuel Roberts: I mean, we’re just taking the LLM return, so it’s not like we want to do a formatting on the outside there, so I think we just need to update the prompt to…

1239 02:22:11.740 ⇒ 02:22:27.359 Samuel Roberts: If… what, multiple… If multiple… Records return with… The same… person?

1240 02:22:27.520 ⇒ 02:22:31.570 Samuel Roberts: display them… Hmm.

1241 02:22:37.450 ⇒ 02:22:40.050 Samuel Roberts: I wonder if we even just wanted to do, like, a table or something here.

1242 02:22:44.350 ⇒ 02:22:50.539 Samuel Roberts: Oh, time is… okay, I… I don’t… I gotta run. I hear him crying. The babysitter’s gone.

1243 02:22:51.840 ⇒ 02:22:57.590 Pranav Narahari: Okay, gotcha. Do you have time to give a quick update in the Slack channel at some point today?

1244 02:22:57.590 ⇒ 02:23:01.810 Samuel Roberts: I posted in the thread with the doubling one, and then I assigned…

1245 02:23:01.810 ⇒ 02:23:03.910 Pranav Narahari: Not for ABC, for Eden.

1246 02:23:06.290 ⇒ 02:23:09.629 Samuel Roberts: Oh, what was that? Sorry, I missed that one then while I was doing the other stuff. What was that?

1247 02:23:09.930 ⇒ 02:23:13.899 Pranav Narahari: So, like, just, like, on that, thread that…

1248 02:23:14.650 ⇒ 02:23:17.579 Pranav Narahari: Utam asked for, like, to keep in the loop on.

1249 02:23:18.030 ⇒ 02:23:20.429 Pranav Narahari: I’ll, I can just tag you in there.

1250 02:23:22.530 ⇒ 02:23:28.100 Pranav Narahari: I created a thread, like, early this morning, and then Utam just said, let me know where this lands today.

1251 02:23:28.690 ⇒ 02:23:31.340 Samuel Roberts: Oh, oh, oh, Irving this morning, yeah, yeah, I’ll jump in there, okay.

1252 02:23:32.280 ⇒ 02:23:33.999 Pranav Narahari: I’m just gonna bump it.

1253 02:23:34.700 ⇒ 02:23:36.910 Samuel Roberts: Yeah, no, I see it, I see it there, you’re good, you’re good.

1254 02:23:36.910 ⇒ 02:23:38.050 Pranav Narahari: Okay, okay.

1255 02:23:38.370 ⇒ 02:23:39.160 Samuel Roberts: Don’t.

1256 02:23:40.400 ⇒ 02:23:45.570 Pranav Narahari: Just so it’s visible for everybody, okay, cool.

1257 02:23:45.920 ⇒ 02:23:48.669 Pranav Narahari: Yeah, if you can do that at some point, that’d be great.

1258 02:23:48.670 ⇒ 02:23:50.140 Samuel Roberts: Yeah, I’ll do the renewal.

1259 02:23:50.660 ⇒ 02:23:55.700 Pranav Narahari: This one, I guess, we’ll just have to leave open until tomorrow. Okay, that’s fine.

1260 02:24:00.260 ⇒ 02:24:01.839 Samuel Roberts: Which one is open for tomorrow?

1261 02:24:04.050 ⇒ 02:24:06.310 Pranav Narahari: The response format issue?

1262 02:24:12.100 ⇒ 02:24:22.039 Samuel Roberts: Oh, okay. Yeah, I can… yeah, I can take care of that tomorrow, or if, you know, we want to take care of it sooner, someone else can, and I can look at it later, but I’m happy to dive into it tomorrow, but…

1263 02:24:22.480 ⇒ 02:24:23.789 Casie Aviles: Okay, sounds good.

1264 02:24:24.010 ⇒ 02:24:26.249 Casie Aviles: I can probably just change the prompt.

1265 02:24:26.550 ⇒ 02:24:28.870 Casie Aviles: But I’ll have to test it first.

1266 02:24:29.420 ⇒ 02:24:30.050 Samuel Roberts: Yeah.

1267 02:24:32.400 ⇒ 02:24:34.550 Pranav Narahari: Yeah, we’ll close that out for sure tomorrow, then.

1268 02:24:35.140 ⇒ 02:24:35.710 Samuel Roberts: Okay.

1269 02:24:36.420 ⇒ 02:24:40.030 Pranav Narahari: Alright. And then, Casey, for…

1270 02:24:40.920 ⇒ 02:24:46.949 Pranav Narahari: the central dock ones. So, it looks like on one of these, you gave a bunch of activity.

1271 02:24:49.290 ⇒ 02:24:51.390 Pranav Narahari: So, does this need to be…

1272 02:24:51.990 ⇒ 02:24:59.359 Pranav Narahari: This is not… this is showing Travis is not picked for residential, Gilbert is also… yeah, so Janiece gave a response here, actually.

1273 02:25:02.170 ⇒ 02:25:02.990 Casie Aviles: Okay.

1274 02:25:06.030 ⇒ 02:25:06.860 Casie Aviles: C.

1275 02:25:11.330 ⇒ 02:25:18.080 Casie Aviles: Which Gilbert… it’s a Gilbert Juarez, so that’s fine. What I showed was a different Gilbert.

1276 02:25:22.800 ⇒ 02:25:25.609 Pranav Narahari: Yeah, I guess they didn’t say which Gilbert.

1277 02:25:26.210 ⇒ 02:25:30.460 Pranav Narahari: Oh, well, they did ask about Gilbert Juarez.

1278 02:25:30.670 ⇒ 02:25:36.080 Pranav Narahari: So I think you… You were looking at a different Gilbert when they were referring to Gilbert Juarez.

1279 02:25:36.210 ⇒ 02:25:40.150 Casie Aviles: Yes, yes, so… yeah, it’s a different Gilbert.

1280 02:25:41.520 ⇒ 02:25:44.080 Casie Aviles: They deleted the Juarez one.

1281 02:25:44.580 ⇒ 02:25:47.340 Casie Aviles: And that’s correct, it’s not in the DB anymore.

1282 02:25:49.300 ⇒ 02:25:52.210 Pranav Narahari: Okay, so is this closed, then? I’m confused.

1283 02:25:53.200 ⇒ 02:25:57.300 Casie Aviles: This is showing drivers, it’s not big for residential.

1284 02:25:59.550 ⇒ 02:26:01.050 Pranav Narahari: Let me ask this question again.

1285 02:26:01.160 ⇒ 02:26:02.509 Pranav Narahari: to Andy.

1286 02:26:18.770 ⇒ 02:26:21.520 Casie Aviles: Yeah, these issues… Confusing…

1287 02:26:40.960 ⇒ 02:26:42.980 Pranav Narahari: So I think the output should be Ryan.

1288 02:26:53.800 ⇒ 02:26:56.560 Casie Aviles: Brian, then, Helen, okay.

1289 02:27:02.550 ⇒ 02:27:05.269 Pranav Narahari: Yeah, so I just did a…

1290 02:27:05.700 ⇒ 02:27:12.999 Pranav Narahari: a test, and it says, no, not every termite technician can perform Chilona maintenance, you must schedule a Cholona-approved technician.

1291 02:27:14.300 ⇒ 02:27:19.959 Pranav Narahari: Okay, yeah, let’s… I mean, the question itself isn’t even asking for who.

1292 02:27:21.370 ⇒ 02:27:22.260 Casie Aviles: Yeah, it’s not even on.

1293 02:27:22.260 ⇒ 02:27:22.590 Pranav Narahari: Okay.

1294 02:27:22.590 ⇒ 02:27:25.029 Casie Aviles: So, who the technicians are.

1295 02:27:27.340 ⇒ 02:27:29.330 Pranav Narahari: Yes, I think that’s updated then.

1296 02:27:29.950 ⇒ 02:27:33.220 Pranav Narahari: Yeah, so on this one, do you want to just respond to Janiece and say.

1297 02:27:36.080 ⇒ 02:27:37.610 Pranav Narahari: Alright, I can just respond right now.

1298 02:27:37.840 ⇒ 02:27:39.050 Casie Aviles: Oh, okay, yeah.

1299 02:28:23.570 ⇒ 02:28:29.750 Pranav Narahari: Alright, this one we can put into the done… or… Yeah, I think it’s…

1300 02:28:29.920 ⇒ 02:28:37.419 Pranav Narahari: We can put it into, for Janice’s review… I got it.

1301 02:28:38.280 ⇒ 02:28:38.970 Casie Aviles: Okay.

1302 02:28:39.920 ⇒ 02:28:41.050 Pranav Narahari: 4 or more.

1303 02:28:41.690 ⇒ 02:28:50.990 Casie Aviles: I have a question for 2550. Do we have, like, the… this WDI pricing sheet, or is that something…

1304 02:28:51.520 ⇒ 02:28:55.139 Casie Aviles: we don’t have yet, so I can ask Janice as well.

1305 02:28:55.250 ⇒ 02:28:56.499 Casie Aviles: To provide that.

1306 02:28:57.590 ⇒ 02:29:00.790 Pranav Narahari: Let me check in the… Google Drive.

1307 02:29:18.070 ⇒ 02:29:20.070 Pranav Narahari: Yeah, I don’t see it here.

1308 02:29:21.610 ⇒ 02:29:23.680 Casie Aviles: Okay, I’ll just ask Janiece.

1309 02:29:24.280 ⇒ 02:29:26.390 Pranav Narahari: Yeah.

1310 02:29:27.140 ⇒ 02:29:29.660 Pranav Narahari: I would say let’s ask Mustafa first.

1311 02:29:31.380 ⇒ 02:29:32.360 Casie Aviles: Oh, okay.

1312 02:29:32.750 ⇒ 02:29:33.420 Pranav Narahari: No.

1313 02:29:37.090 ⇒ 02:29:40.390 Mustafa Raja: Terry, what, what are we talking about here?

1314 02:29:41.900 ⇒ 02:29:46.620 Pranav Narahari: Oh, I didn’t even realize you’re still here. Okay, cool. The… the WDI pricing sheet?

1315 02:29:49.920 ⇒ 02:29:56.359 Pranav Narahari: Is that in the… I don’t see it in the Universal Docs, but is that something we’re embedding from, like, you know, a different Google Sheet or something?

1316 02:29:57.610 ⇒ 02:29:59.079 Mustafa Raja: Can you send me a link?

1317 02:30:01.830 ⇒ 02:30:07.040 Pranav Narahari: Yeah, you want to send the ticket? Or, Casey, you want to share your screen and show which one we’re talking about?

1318 02:30:08.520 ⇒ 02:30:10.049 Casie Aviles: Yeah, I can also shoot that.

1319 02:30:10.690 ⇒ 02:30:12.040 Pranav Narahari: Yeah, let’s just do that.

1320 02:30:14.430 ⇒ 02:30:18.930 Casie Aviles: Yeah, I also sent it in the chat, it’s ABC2550.

1321 02:30:24.040 ⇒ 02:30:27.559 Mustafa Raja: Yeah, I think, yeah, I don’t have much context on this.

1322 02:30:28.130 ⇒ 02:30:31.440 Pranav Narahari: Okay, so then we’ll just ask Janiece then for.

1323 02:30:31.500 ⇒ 02:30:38.030 Mustafa Raja: And if it’s not in the folder, It’s not being embedded.

1324 02:30:38.590 ⇒ 02:30:40.149 Pranav Narahari: Okay. Cool.

1325 02:30:41.370 ⇒ 02:30:44.259 Pranav Narahari: Did you see Janice’s response in the external chat?

1326 02:30:49.070 ⇒ 02:30:51.979 Pranav Narahari: She just said, yes sir, so I think we’re good to go on…

1327 02:30:53.700 ⇒ 02:30:59.149 Mustafa Raja: Yes, she said yes, sir, but I asked her to point out the templates.

1328 02:31:00.980 ⇒ 02:31:05.050 Pranav Narahari: We’re getting templates from Gusha, we want to move away from it. Yeah, so…

1329 02:31:06.790 ⇒ 02:31:14.829 Mustafa Raja: Yeah, I asked if she can help us point, point which one of the Universal Docs are templates, and she said, yes, sir.

1330 02:31:15.300 ⇒ 02:31:18.079 Pranav Narahari: Yeah, I’m gonna be more clear here.

1331 02:31:33.250 ⇒ 02:31:36.240 Samuel Roberts: Alright, anything else, you need from me before I jump?

1332 02:31:36.620 ⇒ 02:31:38.310 Samuel Roberts: I gotta go baby duty.

1333 02:31:39.750 ⇒ 02:31:40.540 Samuel Roberts: Yeah.

1334 02:31:42.040 ⇒ 02:31:45.620 Pranav Narahari: No, I think, I think we’re good. Okay.

1335 02:31:45.740 ⇒ 02:31:48.810 Pranav Narahari: Yes, yeah, once you update the… oh, you…

1336 02:31:49.060 ⇒ 02:31:51.980 Pranav Narahari: Did you update it? Yeah, perfect.

1337 02:31:51.980 ⇒ 02:31:56.770 Samuel Roberts: Take a look at those proposed tickets from earlier, too. I think we can sink tomorrow on them a little bit.

1338 02:31:57.580 ⇒ 02:32:01.589 Pranav Narahari: Yeah, I think the idea there was I wanted to see a POC first.

1339 02:32:01.820 ⇒ 02:32:04.150 Pranav Narahari: Before we kind of break down the tickets.

1340 02:32:04.150 ⇒ 02:32:05.110 Samuel Roberts: The theming?

1341 02:32:06.500 ⇒ 02:32:09.009 Pranav Narahari: Yeah, so, like, for that rep- for the reports?

1342 02:32:09.180 ⇒ 02:32:15.200 Pranav Narahari: like, it’s kind of… both these POCs should be… Probably just one POC.

1343 02:32:15.440 ⇒ 02:32:18.079 Pranav Narahari: like, you could probably just do it in one POC,

1344 02:32:18.180 ⇒ 02:32:22.630 Pranav Narahari: with just the kind of the questions that I generated, right? Like, if we can effectively answer those…

1345 02:32:22.630 ⇒ 02:32:24.640 Samuel Roberts: I’m talking about tickets for the theme report stuff.

1346 02:32:24.860 ⇒ 02:32:27.380 Samuel Roberts: Not just the… the questions.

1347 02:32:28.780 ⇒ 02:32:30.149 Samuel Roberts: The POC I gotta work on tomorrow.

1348 02:32:31.140 ⇒ 02:32:37.870 Pranav Narahari: Yeah, I’m guessing those tickets are based on, like, your technical approach, right? And so we just need to make sure the technical approach is right.

1349 02:32:38.930 ⇒ 02:32:40.330 Pranav Narahari: So you want me to put…

1350 02:32:40.330 ⇒ 02:32:43.960 Samuel Roberts: as well, in addition to those questions and bottleneck?

1351 02:32:45.120 ⇒ 02:32:46.290 Pranav Narahari: Sorry, can you repeat that?

1352 02:32:47.360 ⇒ 02:32:51.659 Samuel Roberts: the POCs you’re talking about that you mentioned in the Slack, where the questions

1353 02:32:52.910 ⇒ 02:32:58.190 Samuel Roberts: It was before the questions, but now we have the questions to test and develop, like… That… that out.

1354 02:32:58.580 ⇒ 02:33:02.210 Samuel Roberts: You want something for the theme report generator, too?

1355 02:33:06.160 ⇒ 02:33:15.789 Pranav Narahari: I think what I’m noticing here is just the… the questions that are being asked are kind of…

1356 02:33:17.420 ⇒ 02:33:20.359 Pranav Narahari: They’re essentially just, like, gonna be the same thing that…

1357 02:33:21.450 ⇒ 02:33:26.109 Pranav Narahari: we’re going to ask the AI to do for generating a report.

1358 02:33:26.300 ⇒ 02:33:29.739 Pranav Narahari: Like, that generating the report portion of things, I think, is…

1359 02:33:29.840 ⇒ 02:33:36.429 Pranav Narahari: Actually, extremely simple, because what’s the difficult part is actually being able to answer those questions.

1360 02:33:36.800 ⇒ 02:33:46.360 Samuel Roberts: Sure, okay. Yeah, then I… yeah, that’ll be tomorrow. I’ll test those questions. I… I’m worried the org chart is a blocker, because we can’t impersonate people without their emails.

1361 02:33:46.510 ⇒ 02:33:51.620 Samuel Roberts: But I’ll find that out in the morning. But if you can poke them for that, that would be great.

1362 02:33:52.210 ⇒ 02:34:01.190 Mustafa Raja: Oh, sorry, the org chart, we need that, so we can explicitly ask about the calendars and stuff like that, right?

1363 02:34:01.190 ⇒ 02:34:03.230 Samuel Roberts: That’s exactly what I’m talking about, Mustafa.

1364 02:34:03.590 ⇒ 02:34:06.119 Mustafa Raja: Okay, okay. So the People’s API didn’t work?

1365 02:34:09.810 ⇒ 02:34:17.140 Samuel Roberts: The People’s API isn’t gonna give us everything that we’re gonna need from that, I think. I mean, we could just guess from that, but I can try that in the morning. That’s why I’m not sure if it’s a full.

1366 02:34:17.140 ⇒ 02:34:18.920 Mustafa Raja: Yeah, oh yeah, okay, great, yeah.

1367 02:34:19.290 ⇒ 02:34:22.579 Mustafa Raja: I’m just interested in that selection also.

1368 02:34:23.070 ⇒ 02:34:27.910 Samuel Roberts: Yeah, yeah, yeah, that’s something I can try out in the morning. Yeah.

1369 02:34:28.530 ⇒ 02:34:30.130 Samuel Roberts: And then test those questions.

1370 02:34:30.690 ⇒ 02:34:31.260 Mustafa Raja: Yep.

1371 02:34:33.100 ⇒ 02:34:37.609 Mustafa Raja: Also, let me know if the memory looks good to you, once you get… Yes.

1372 02:34:37.610 ⇒ 02:34:49.730 Samuel Roberts: I got that downloaded, it worked, I was able to… yeah, sorry, I did respond to that earlier, but I did see, in my quick testing, that it was working there. I think we’ll still want to look at persisting that volume.

1373 02:34:50.070 ⇒ 02:34:53.539 Samuel Roberts: So that it stays… But we can talk about that later.

1374 02:34:54.050 ⇒ 02:34:59.690 Mustafa Raja: Yeah, yeah, okay, let’s talk about that later, because I do really want to talk about that.

1375 02:35:00.360 ⇒ 02:35:01.279 Samuel Roberts: Yeah, definitely.

1376 02:35:01.280 ⇒ 02:35:02.729 Mustafa Raja: Once you have good time.

1377 02:35:03.240 ⇒ 02:35:06.060 Samuel Roberts: Yeah, I think tomorrow, if you want to grab some time early, we can.

1378 02:35:06.060 ⇒ 02:35:07.549 Mustafa Raja: Yeah, yeah, yeah, definitely.

1379 02:35:08.240 ⇒ 02:35:09.480 Samuel Roberts: Okay, sounds good.

1380 02:35:10.600 ⇒ 02:35:11.969 Samuel Roberts: Alright, thank you guys.

1381 02:35:12.920 ⇒ 02:35:13.500 Casie Aviles: Thank you.

1382 02:35:13.500 ⇒ 02:35:14.649 Mustafa Raja: Thank you. Have a good day.

1383 02:35:15.060 ⇒ 02:35:17.480 Samuel Roberts: Morning. Alright, or tomorrow, yeah. Alright.

1384 02:35:18.410 ⇒ 02:35:23.039 Casie Aviles: Yeah. For Pranav, I have, like, a question for you.