Meeting Title: ABC | backlog grooming Date: 2025-04-28 Meeting participants: Uttam Kumaran, Amber Lin, Miguel De Veyra, Casie Aviles


WEBVTT

1 00:01:21.100 00:01:22.610 Amber Lin: Hello!

2 00:01:24.210 00:01:25.020 Casie Aviles: And.

3 00:01:59.420 00:02:06.406 Amber Lin: Is he? Do you know if Miguel’s coming, I think, he said, yes, I just. We need to wait a little bit more.

4 00:02:06.710 00:02:08.119 Casie Aviles: Yeah. Let me message him.

5 00:02:08.610 00:02:09.050 Amber Lin: Okay.

6 00:02:41.870 00:02:42.700 Uttam Kumaran: Hello!

7 00:02:43.970 00:02:45.090 Amber Lin: Bye.

8 00:02:46.250 00:02:47.190 Uttam Kumaran: Morning.

9 00:02:47.680 00:02:49.079 Amber Lin: Good morning.

10 00:02:52.300 00:02:54.189 Amber Lin: We’ll get started, and then

11 00:02:54.380 00:03:02.520 Amber Lin: when the go comes we can just keep rolling. So this meeting is, just look at the backlog and

12 00:03:03.308 00:03:08.350 Amber Lin: decide on the pace of how we want to move forward.

13 00:03:09.052 00:03:12.848 Amber Lin: Look at what tickets we wanna do this week, and also next week.

14 00:03:13.920 00:03:21.589 Amber Lin: So let me share my screen for linear. Right now we’ve rolled out to all the Csrs.

15 00:03:22.329 00:03:27.279 Amber Lin: There we have improvements here and there. We want to do

16 00:03:30.430 00:03:34.430 Amber Lin: And let’s see.

17 00:03:34.570 00:03:42.800 Amber Lin: So there’s I think the main thing right now is for the document. Update bot

18 00:03:42.950 00:03:47.569 Amber Lin: the trainer, bot to be developed to help them with the formatting.

19 00:03:49.110 00:04:01.040 Amber Lin: I’ll help Annie and I will help do the business reviews. So that’s gonna have it bi-weekly. I’m gonna book. That meeting with you. Fetch and let’s see.

20 00:04:03.720 00:04:23.700 Amber Lin: Yeah. And then there’s quite a quite a few items that they requested throughout the Friday meetings that I put in the backlog. Some of them are here also, so I guess we can go through them and see what is the priority. And what do we want it?

21 00:04:24.220 00:04:26.550 Amber Lin: What do we want to do, and

22 00:04:27.180 00:04:30.959 Amber Lin: how spaced out? We want all these items to be.

23 00:04:38.450 00:04:41.810 Uttam Kumaran: So like. Where are we in the cycle? By the way.

24 00:04:44.160 00:04:44.940 Amber Lin: Here.

25 00:04:46.140 00:04:47.790 Amber Lin: This is the cycle.

26 00:04:48.710 00:04:56.159 Amber Lin: So we have a week, a week more to go, and then we start another cycle.

27 00:04:58.220 00:05:03.219 Uttam Kumaran: So in terms of this cycle looks like we still have to get through all of these tickets right.

28 00:05:04.030 00:05:13.590 Amber Lin: Hmm, yeah, so these are like one off one off things.

29 00:05:13.860 00:05:16.980 Amber Lin: And then this is the.

30 00:05:18.310 00:05:20.960 Amber Lin: These are the items on the data side.

31 00:05:21.450 00:05:23.580 Amber Lin: So yes, we do need to get through them.

32 00:05:24.410 00:05:30.400 Uttam Kumaran: So are we plan? Are we planning for the next cycle right now, or is it just? Is it for.

33 00:05:30.718 00:05:40.600 Amber Lin: Yeah, for the next week, cause we don’t have a we know what we need to do for the trainer. Bot. But look at tickets are not very fleshed out, or.

34 00:05:40.600 00:05:42.400 Uttam Kumaran: Week, or for for the following week.

35 00:05:42.850 00:05:43.970 Amber Lin: For this week.

36 00:05:45.420 00:05:49.950 Uttam Kumaran: But I guess that’s what I’m saying is like we still have tickets that are in the cycle.

37 00:05:51.810 00:05:56.290 Uttam Kumaran: So I guess my question is like.

38 00:05:56.940 00:06:01.709 Uttam Kumaran: if are we? Gonna we’re gonna add more tickets to that. And then are we gonna be able to get all those done.

39 00:06:04.000 00:06:05.010 Amber Lin: Good question.

40 00:06:06.145 00:06:12.359 Amber Lin: Here, let’s look at the current cycle. So these 2 are

41 00:06:12.840 00:06:16.710 Amber Lin: for the data side. I think the Api is

42 00:06:17.590 00:06:21.859 Amber Lin: should be blocked until Tim gives us information.

43 00:06:23.000 00:06:28.760 Amber Lin: These are for me. So we have capacity on the engineering side.

44 00:06:29.340 00:06:32.560 Amber Lin: So that’s we only have tickets for Annie and me.

45 00:06:33.120 00:06:33.770 Uttam Kumaran: Okay.

46 00:06:34.420 00:06:43.420 Amber Lin: Yeah. So I wanted good question. I wanted to have some progress on the trader bar and be able to present something on Friday.

47 00:06:44.910 00:06:51.860 Uttam Kumaran: So which tickets are related to the trainer. Bot.

48 00:06:55.630 00:07:03.369 Uttam Kumaran: And then, I guess, like, Let, I guess, in this meeting. Let’s just go through all. Let’s just go through all these, because it’s still, I guess I don’t know

49 00:07:03.700 00:07:08.970 Uttam Kumaran: what each of these are, so let’s just go one by one from 3 0, 6 down.

50 00:07:09.480 00:07:11.150 Amber Lin: Sure. Let’s do that.

51 00:07:12.720 00:07:13.299 Uttam Kumaran: Can go.

52 00:07:13.300 00:07:18.849 Amber Lin: Actually from here do we have Github admin for all the teams.

53 00:07:20.600 00:07:25.489 Uttam Kumaran: Yes, but only like right now. It’s just me and Robert our admin.

54 00:07:25.650 00:07:28.649 Uttam Kumaran: So I guess, like for this ticket, like what?

55 00:07:29.250 00:07:31.520 Uttam Kumaran: Why like? Why do we need this.

56 00:07:32.190 00:07:42.660 Amber Lin: This is because last time we were gonna grant Tim access and we had to go through you. So if you give one of us access we wouldn’t have to ask you for that anymore.

57 00:07:43.090 00:07:48.800 Uttam Kumaran: So right now, like, Marianne has this. So this is gonna go through operations team.

58 00:07:49.390 00:07:50.120 Uttam Kumaran: So.

59 00:07:50.120 00:07:50.640 Amber Lin: Okay.

60 00:07:50.990 00:07:55.650 Uttam Kumaran: Yeah, any any any pool? Access goes through Ops.

61 00:07:56.140 00:07:57.790 Uttam Kumaran: Now, so

62 00:07:58.060 00:08:06.859 Uttam Kumaran: and probably mark this as Don. Yeah, she has. She and a couple of other people have admin. But they have a process because we need, we’re gonna start basically tracking access to everything

63 00:08:07.060 00:08:08.580 Uttam Kumaran: for compliance. So.

64 00:08:08.990 00:08:10.370 Amber Lin: So I’ll say this is done

65 00:08:11.220 00:08:20.115 Amber Lin: right. Next one. This one is to. We want to check the errors, error rate and the quality scores.

66 00:08:20.880 00:08:36.019 Amber Lin: I was discussing with Casey last Friday of our error. Rates are constantly 0, which is not the case because we do get feedback, saying that it’s not right. So what we’re talking about is okay. Can we incorporate

67 00:08:36.159 00:08:43.740 Amber Lin: the thumbs down feedback or their detailed feedback into a factor of a quality? Scores and error rates.

68 00:08:44.510 00:08:51.779 Uttam Kumaran: Okay, maybe let’s okay, can we? Can we write that in the ticket? If you could just like edit this and just say, Yeah, can we include?

69 00:08:55.110 00:09:00.490 Uttam Kumaran: So can we include thumbs down as part of the yeah.

70 00:09:00.990 00:09:02.860 Uttam Kumaran: And then maybe we could just make

71 00:09:03.640 00:09:07.660 Uttam Kumaran: this the title like thumbs down in quality score.

72 00:09:08.660 00:09:09.330 Amber Lin: Hmm.

73 00:09:11.030 00:09:12.950 Uttam Kumaran: That way. It’s really clear.

74 00:09:15.845 00:09:20.049 Uttam Kumaran: Okay, so yeah, let’s so let’s talk about

75 00:09:20.699 00:09:23.399 Uttam Kumaran: I guess what? Let’s come in back and do

76 00:09:24.240 00:09:28.749 Uttam Kumaran: come back and do the priorities, or we can do it now. But what is like.

77 00:09:29.690 00:09:32.040 Uttam Kumaran: how big of a priority is this?

78 00:09:32.440 00:09:34.620 Amber Lin: I don’t think this is more of our

79 00:09:35.240 00:09:51.920 Amber Lin: improve internal improvements. They’re fine with the quality scores, errors. I don’t think they’re looking at it that much. It’s more of a if we want to be accurate in measuring our improvements, then we should be more improve. These.

80 00:09:52.700 00:09:57.020 Uttam Kumaran: So what? So what is like? What actually needs to happen on the engineering side

81 00:09:57.230 00:09:59.109 Uttam Kumaran: to make this happen, Casey?

82 00:09:59.220 00:10:05.310 Uttam Kumaran: And then can we put like the acceptance criteria here, like what? What changes need to be made.

83 00:10:06.410 00:10:13.830 Casie Aviles: And where? Yeah, yeah, because, I think what we could do is to just include the

84 00:10:14.925 00:10:18.250 Casie Aviles: the thumbs up thumbs down feedback. So

85 00:10:18.964 00:10:21.159 Casie Aviles: what we could do with that is to

86 00:10:21.920 00:10:36.989 Casie Aviles: modify the quality score, so I could include it, or as part of the prompt of the because we have, like another Llm. Chain to generate the quality score, so we could include the feedback

87 00:10:37.590 00:10:39.999 Casie Aviles: as part of its considerations.

88 00:10:41.840 00:10:44.330 Amber Lin: Hmm cool.

89 00:10:47.050 00:10:54.340 Uttam Kumaran: So include as but see, even this is like, not like, what does this mean? Include as part of quality score Llm. Prompt like

90 00:10:59.760 00:11:03.759 Uttam Kumaran: So you’re so there’s so how does the quality score work right now?

91 00:11:04.570 00:11:11.350 Casie Aviles: So we use another Llm. To generate the quality score. So on top of the

92 00:11:11.530 00:11:14.313 Casie Aviles: scores that we get with the

93 00:11:14.840 00:11:15.840 Uttam Kumaran: Brain, trust.

94 00:11:15.840 00:11:22.329 Casie Aviles: Yes, it takes those as input as text input and then it should ask.

95 00:11:22.640 00:11:25.289 Casie Aviles: generate a quality score from that.

96 00:11:26.450 00:11:27.939 Casie Aviles: So it’s an Llm.

97 00:11:30.210 00:11:34.530 Uttam Kumaran: Okay, so include thumbs down. So can you. So, Casey, can you link

98 00:11:34.840 00:11:38.820 Uttam Kumaran: wherever is this? Is this gonna be a change in N, 8 n.

99 00:11:39.990 00:11:41.069 Casie Aviles: Yes! Yes!

100 00:11:41.850 00:11:46.140 Uttam Kumaran: Okay, so can you link, yeah, if we can just link where this change

101 00:11:46.440 00:12:01.810 Uttam Kumaran: has to happen. So part. So I guess like for. And this is a really good like example. And I’ll probably start joining a couple couple of other teams for grooming this next few weeks. But part of the goal for grooming is less about

102 00:12:02.010 00:12:06.469 Uttam Kumaran: planning, meaning less about like when it’s gonna happen.

103 00:12:06.670 00:12:07.130 Amber Lin: Hmm.

104 00:12:07.480 00:12:22.920 Uttam Kumaran: And who’s gonna take it more about what it is and like what the actual scope is, and a good barometer for a ticket is anyone on the team should be able to look at the ticket, understand? And actually go execute it

105 00:12:23.060 00:12:27.739 Uttam Kumaran: right? We’re probably pretty far from that. But like, that’s a good north star.

106 00:12:29.760 00:12:47.550 Uttam Kumaran: so one of the things that you’ll see when we’re and we probably won’t get to everything today, which is okay. One of the things you’ll see is during grooming. This is where we want to make it really clear what needs to happen for this to be considered done. And then how do we go test this right? So

107 00:12:48.270 00:13:00.760 Uttam Kumaran: Casey mentioned, include thumbs down feedback as part of quality score looks like there’s some modification that has to happen in 8 n, and then I think, Casey, I want to know how, what, how we go test this

108 00:13:05.240 00:13:10.209 Casie Aviles: Hmm, yeah. So I guess for testing what I would do is I would just

109 00:13:10.410 00:13:14.539 Casie Aviles: go to the bot and chat and then send a thumbs down feedback.

110 00:13:19.990 00:13:24.980 Casie Aviles: and I I’ll check like, what’s the quality score after that.

111 00:13:25.890 00:13:26.520 Uttam Kumaran: Okay?

112 00:13:31.200 00:13:35.580 Uttam Kumaran: And then there’s no deployment like needed, right? Because it’s this is on any. Then.

113 00:13:36.840 00:13:37.340 Casie Aviles: Yes!

114 00:13:38.090 00:13:38.660 Uttam Kumaran: Okay?

115 00:13:39.890 00:13:43.560 Uttam Kumaran: So then you and then you check quality score. Okay, okay, great

116 00:13:44.280 00:13:53.980 Uttam Kumaran: cool. Alright. This is in a better spot. Now I feel like, though in terms of priority, I think we yeah, we should just set the prior if we if we don’t know, just set it in the middle.

117 00:13:54.681 00:13:59.358 Uttam Kumaran: And then, yeah, Casey, I think it’s for up to you to provide an estimate.

118 00:14:00.320 00:14:01.400 Amber Lin: Somewhere.

119 00:14:01.750 00:14:03.770 Amber Lin: Okay. I think around here.

120 00:14:04.710 00:14:07.189 Casie Aviles: I think we can do it. 2 points.

121 00:14:07.560 00:14:11.660 Amber Lin: Okay, awesome sounds good.

122 00:14:11.910 00:14:15.390 Amber Lin: So next one.

123 00:14:17.130 00:14:24.020 Amber Lin: So I would say, this is also lower priority. I’ll explain. This ticket of

124 00:14:24.930 00:14:41.179 Amber Lin: I just saw that in our feedback feedback sheet there was one that they put thumbs down because they wanted to put in feedback a good feedback, but we didn’t have a feedback box for thumbs up data so that would confuse it.

125 00:14:41.310 00:14:52.799 Amber Lin: Moving forward. If if we include thumbs down as a feedback, quality score feature, if we don’t have somewhere where people do a feedback or thumbs up, then they’re gonna put it in thumbs down.

126 00:14:53.110 00:14:56.420 Amber Lin: So that’s the reason why I want to have this.

127 00:15:01.000 00:15:05.650 Uttam Kumaran: Add detailed, okay, cool, cool. I think this is fine.

128 00:15:06.473 00:15:07.500 Amber Lin: I think.

129 00:15:08.040 00:15:12.070 Uttam Kumaran: In terms of the so in terms of acceptance. Criteria. This makes sense.

130 00:15:12.620 00:15:15.100 Uttam Kumaran: What is this on your side, Casey, like.

131 00:15:17.730 00:15:18.070 Casie Aviles: So.

132 00:15:18.070 00:15:18.700 Uttam Kumaran: Yeah.

133 00:15:19.660 00:15:24.040 Casie Aviles: Yeah, I think I’ll have to edit the code here. So yeah.

134 00:15:26.450 00:15:32.090 Uttam Kumaran: So this is editing the code, where, oh, in in the for the for the Google script.

135 00:15:32.090 00:15:33.680 Casie Aviles: Yeah, yeah.

136 00:15:34.952 00:15:37.487 Uttam Kumaran: Okay, the other. The other piece is

137 00:15:41.050 00:15:43.130 Uttam Kumaran: The what’s it called?

138 00:15:44.236 00:15:47.569 Uttam Kumaran: Tim is gonna work on creating a staging.

139 00:15:47.870 00:15:48.790 Amber Lin: Hmm.

140 00:15:48.790 00:15:50.429 Uttam Kumaran: A testing version.

141 00:15:50.980 00:15:56.190 Uttam Kumaran: But we it may not happen this before we get this done.

142 00:15:56.590 00:15:59.909 Uttam Kumaran: because basically anything that has a Ui change.

143 00:16:00.860 00:16:04.099 Uttam Kumaran: I wanna go test before we like roll it out.

144 00:16:05.126 00:16:11.519 Uttam Kumaran: So probably one ticket that we can add in the backlog is around establishing this new development

145 00:16:12.730 00:16:14.659 Amber Lin: We have it actually.

146 00:16:14.660 00:16:19.179 Uttam Kumaran: This new like development, like agent, or whatever chat

147 00:16:19.360 00:16:21.800 Uttam Kumaran: that way, we can test any of these front facing changes.

148 00:16:22.340 00:16:27.650 Amber Lin: Hmm stay tuned appointments.

149 00:16:27.940 00:16:33.310 Uttam Kumaran: You could just say, state, yeah, staging deployment, yeah, testing agent.

150 00:16:33.570 00:16:41.560 Uttam Kumaran: And then basically the goal of this is, we want to be able to test user facing changes

151 00:16:41.880 00:16:46.060 Uttam Kumaran: before rolling things out to the production agent.

152 00:16:52.870 00:16:53.690 Uttam Kumaran: So

153 00:16:53.810 00:16:59.260 Uttam Kumaran: the way this is gonna work technically is we’re gonna have everything’s gonna get merged into a development branch

154 00:16:59.510 00:17:08.430 Uttam Kumaran: that’s Branch is going to be connected to another agent. We’ll go be able to test once we’re like comfortable. We can then merge that into master.

155 00:17:23.169 00:17:28.859 Amber Lin: And acceptance criteria is that it’s folk

156 00:17:31.709 00:17:37.239 Amber Lin: that changes we make can be seen in the testing.

157 00:17:37.690 00:17:41.719 Uttam Kumaran: Yeah, basically, we want to actually be able to test the whole thing end to end

158 00:17:41.890 00:17:49.229 Uttam Kumaran: right? Like, I don’t want to make a change to Google because I’m they’re like guaranteed that we’re gonna mess something up at some point.

159 00:17:49.773 00:17:51.959 Uttam Kumaran: And we’re gonna have to do some rollback.

160 00:17:52.210 00:17:56.170 Amber Lin: So as much as possible, I wanna be able to test the whole system.

161 00:17:56.840 00:18:02.110 Uttam Kumaran: In an environment that maybe just a couple of us have access to.

162 00:18:06.290 00:18:07.870 Amber Lin: And since

163 00:18:11.140 00:18:15.790 Amber Lin: oops, I guess leveraging works.

164 00:18:16.700 00:18:17.550 Amber Lin: Okay?

165 00:18:18.340 00:18:24.589 Amber Lin: Oh, well, okay, so, is this all tim is there anything, we need to do.

166 00:18:25.452 00:18:30.899 Uttam Kumaran: Yeah, I basically our team, like Casey and Miguel just need to work with Tim on establishing this.

167 00:18:31.310 00:18:32.050 Amber Lin: Okay,

168 00:18:38.280 00:18:46.379 Amber Lin: I’ll put this in there. Sign, say, like green, 2, 3 points.

169 00:18:46.700 00:18:47.929 Amber Lin: What you guys think.

170 00:18:50.310 00:18:51.390 Uttam Kumaran: Up to Casey.

171 00:18:53.280 00:18:56.089 Casie Aviles: Yeah, I guess we could just put 3 for now.

172 00:18:56.370 00:18:59.270 Amber Lin: Okay, sounds good

173 00:19:02.440 00:19:03.410 Amber Lin: and.

174 00:19:04.180 00:19:10.660 Uttam Kumaran: Yeah. And then one thing is the goal for points. So the best way to think about points is, if

175 00:19:11.170 00:19:20.189 Uttam Kumaran: ideally the the way you can, you can estimate is you should know what the scope of work is right? So that ticket seems like. There’s probably a little bit of ambiguity still.

176 00:19:20.300 00:19:29.160 Uttam Kumaran: So my my sense would be whatever you think, it’s gonna take go, one run higher, typically, like, yeah.

177 00:19:29.160 00:19:29.600 Amber Lin: Hmm.

178 00:19:30.026 00:19:33.869 Uttam Kumaran: Because you’re basically adding a buffer for the ambiguity.

179 00:19:35.700 00:19:43.490 Uttam Kumaran: Otherwise, like our goal here is to be is to be accurate, not to be fast, right

180 00:19:43.690 00:19:56.489 Uttam Kumaran: like we can. Only we are, we are, we’re we’re we’re as we have to assume that we’re we’re all working as fast as we can. But the goal of of these are that we hit the mark, meaning, if that’s a 5, then it should take about

181 00:19:56.630 00:20:00.149 Uttam Kumaran: like whatever to 2 days or so right to get done.

182 00:20:01.490 00:20:03.359 Uttam Kumaran: That’s what we want to go for.

183 00:20:04.420 00:20:05.000 Amber Lin: Hmm!

184 00:20:05.490 00:20:06.680 Amber Lin: Sounds good.

185 00:20:09.460 00:20:18.630 Amber Lin: And back on this. What would Casey? What would you say? The estimate is the feedback box.

186 00:20:21.110 00:20:24.420 Casie Aviles: Okay, yeah. Let’s set it to 3.

187 00:20:24.980 00:20:32.210 Amber Lin: Oh, okay, sounds good, and the

188 00:20:32.690 00:20:39.080 Amber Lin: employment with him, I would say, is, is pretty high or medium.

189 00:20:44.230 00:20:45.780 Casie Aviles: Yeah, I guess. Medium.

190 00:20:46.460 00:20:47.000 Amber Lin: Okay?

191 00:20:50.460 00:21:00.780 Amber Lin: yeah. Now, this one. So for this back. Then they said there was 2 Fridays ago. They said there were some issues with the

192 00:21:01.570 00:21:08.149 Amber Lin: Google sheets, where, with the inspectors and postal codes where

193 00:21:08.500 00:21:14.860 Amber Lin: it was a bit limit, it was limited to pests, and we wanted it to.

194 00:21:15.750 00:21:23.110 Amber Lin: We wanted to check what the error came from of. If it was because we had a limitation in the prompting

195 00:21:23.790 00:21:26.589 Amber Lin: that restricted it to Pest.

196 00:21:28.030 00:21:29.970 Amber Lin: I don’t know if that was very clear.

197 00:21:32.750 00:21:34.739 Miguel de Veyra: I guess. Sorry. I thought it was 11.

198 00:21:35.720 00:21:38.570 Uttam Kumaran: Oh, all good, I guess, for this one.

199 00:21:38.570 00:21:38.970 Uttam Kumaran: Hmm!

200 00:21:40.040 00:21:42.849 Uttam Kumaran: Can do. We have these like errors somewhere.

201 00:21:45.150 00:21:49.320 Amber Lin: I don’t know where it is. They brought it up in the meeting.

202 00:21:49.910 00:21:54.059 Amber Lin: I think it was May. Let me see like problem.

203 00:21:54.260 00:21:55.280 Amber Lin: It’s okay.

204 00:21:55.280 00:22:04.820 Uttam Kumaran: Well, that’s like for this, like, yeah, this, I think this is good, and that we want to do this investigation. But team definitely needs, like a couple of examples of those errors, to go like.

205 00:22:05.135 00:22:05.450 Amber Lin: Good.

206 00:22:05.450 00:22:06.630 Uttam Kumaran: Basically test.

207 00:22:09.220 00:22:16.559 Uttam Kumaran: I think this is a super high prior. This is like a very high priority, though, because everybody I met with asked about the inspectors.

208 00:22:17.770 00:22:19.740 Amber Lin: Okay. Fine errors.

209 00:22:22.340 00:22:25.339 Uttam Kumaran: I, I mean ideally, yeah. And this is like another

210 00:22:25.610 00:22:33.949 Uttam Kumaran: thing we talked about, which is like Janice recording issues in that one, Doc. And then there’s also the spreadsheet, Doc. All of those errors should

211 00:22:34.160 00:22:35.350 Uttam Kumaran: get here.

212 00:22:35.900 00:22:38.950 Uttam Kumaran: Get get into here somehow, right like

213 00:22:39.080 00:22:44.550 Uttam Kumaran: either they should be individual tickets, or they should all be listed under one ticket. But.

214 00:22:45.048 00:22:50.810 Uttam Kumaran: we want to mitigate as much stuff outside of linear, for, like what tasks we’re working on.

215 00:22:52.090 00:22:56.180 Uttam Kumaran: Denise may find it easiest to put in a spreadsheet, but then that should be the only place

216 00:22:57.270 00:22:57.950 Uttam Kumaran: good, too.

217 00:22:57.950 00:22:59.060 Amber Lin: Yeah, absolutely.

218 00:23:00.770 00:23:04.220 Amber Lin: So let’s say, this is high priority.

219 00:23:04.330 00:23:09.560 Amber Lin: How long would it take? I know it’s a bit ambiguous right now.

220 00:23:10.450 00:23:15.389 Uttam Kumaran: Yeah, I don’t. I don’t think we can point this without knowing what the problems are.

221 00:23:19.280 00:23:20.540 Amber Lin: Okay. Let me see.

222 00:23:20.540 00:23:25.520 Uttam Kumaran: I mean, that’s my that’s like my perspective, like I I don’t think we like.

223 00:23:25.660 00:23:26.970 Uttam Kumaran: I think.

224 00:23:27.220 00:23:28.520 Amber Lin: Okay, go ahead.

225 00:23:28.910 00:23:32.560 Uttam Kumaran: Yeah, I just don’t think that there’s I would. I want to see like what the

226 00:23:32.910 00:23:36.580 Uttam Kumaran: what the prompt was that they asked, and then what it received.

227 00:23:37.310 00:23:42.480 Uttam Kumaran: because otherwise it may be like a maybe a 5 min fix. It may be like a few day fix so.

228 00:23:42.620 00:23:56.659 Amber Lin: Yeah, cool sounds good. I think it’s mostly when they ask, Okay, what are the technicians for this service? And it says we don’t have that information. Maybe when they ask for certain handy or other things, they don’t have that information, but I will add that

229 00:23:58.650 00:23:59.970 Amber Lin: I will put

230 00:24:03.500 00:24:06.529 Amber Lin: I’ll assign it to me, for now and then I will.

231 00:24:08.920 00:24:09.680 Amber Lin: You guys.

232 00:24:09.680 00:24:14.169 Uttam Kumaran: Well, this is where I don’t. Wanna. I don’t wanna move it to the cycle until we have

233 00:24:14.500 00:24:17.199 Uttam Kumaran: things done right. So let’s go back to this one.

234 00:24:18.379 00:24:19.519 Uttam Kumaran: This one.

235 00:24:19.810 00:24:23.330 Uttam Kumaran: It’s not assigned this one like, are you going to

236 00:24:23.745 00:24:31.679 Uttam Kumaran: fix? Do the deploy the fix? Because then it’s not gonna be assigned to you, I think what we should do is this, still needs requirements.

237 00:24:32.550 00:24:35.860 Uttam Kumaran: Right? So you can. I think we should move this back to

238 00:24:37.470 00:24:48.080 Uttam Kumaran: requirements started right? Because that that’s ultimately the status of the ticket is. There’s still a couple of requirements needed. Once we have the requirements, we can point it and then move it in.

239 00:24:49.000 00:24:50.029 Amber Lin: Sounds good.

240 00:24:51.130 00:24:52.679 Uttam Kumaran: Thanks sorry for being annoying.

241 00:24:52.680 00:24:58.960 Amber Lin: No, this is good. I just wanted a way to remind myself that this needs to be done.

242 00:25:00.780 00:25:08.110 Uttam Kumaran: Yeah, I think I think a good way of doing that is, if you mark if we’ve marked it as high priority. And then you can basically look at

243 00:25:08.830 00:25:13.079 Uttam Kumaran: either. Look at, look at the high priority ones that are in requirements started.

244 00:25:13.200 00:25:25.490 Uttam Kumaran: Or if you know that we want to get this done this week, then you can set a due date. And that way, when you filter on like, hey, what needs to actually get put into the cycle. You can just look at the high priority ones. But

245 00:25:26.480 00:25:27.100 Uttam Kumaran: yeah.

246 00:25:27.700 00:25:29.320 Amber Lin: Sounds good. That’s great.

247 00:25:30.020 00:25:33.439 Amber Lin: Okay, I will speed through. I’m.

248 00:25:33.440 00:25:40.389 Uttam Kumaran: We’re not. Gonna we’re we’re we’re not gonna be able to get through everything on on just the 3.rd So we’ll just we can keep doing more more of these.

249 00:25:40.550 00:25:43.249 Amber Lin: But this is the this is the level of like.

250 00:25:43.900 00:25:48.420 Uttam Kumaran: Still, we need to have on these tickets on, because what we’re gonna find is

251 00:25:48.530 00:26:03.059 Uttam Kumaran: if we execute these in the sprint, then our lack of detail is gonna is gonna cause us to either take longer or gonna get the wrong thing done. The grooming is gonna be like this. It’s like, basically like one by one by one, by one, we go like and.

252 00:26:03.060 00:26:03.710 Amber Lin: And.

253 00:26:04.350 00:26:06.589 Uttam Kumaran: Like, we just basically check. So we can.

254 00:26:06.820 00:26:16.019 Uttam Kumaran: we can do, we can, we can continue, or we can also just do another session later like. But this is the sort of level of granularity that I want to see.

255 00:26:16.020 00:26:16.540 Amber Lin: Thank.

256 00:26:16.990 00:26:17.660 Uttam Kumaran: Yeah.

257 00:26:19.170 00:26:21.470 Amber Lin: Yeah, I added.

258 00:26:22.484 00:26:32.400 Amber Lin: I added another 30 min to this. I don’t know if you guys have time. We also have an AI grooming session right after this. So just

259 00:26:32.620 00:26:39.979 Amber Lin: we can, we can continue for another 30 min on ABC stuff. I do want to get through the trainer bot stuff first.st

260 00:26:39.980 00:26:40.660 Uttam Kumaran: Okay.

261 00:26:41.040 00:26:50.149 Amber Lin: Yeah, so okay, we have 4 here. So I guess real quick.

262 00:26:51.000 00:27:00.700 Amber Lin: This was something that Steven suggested of having a pending animation while Annie is thinking, as goal is to help improve.

263 00:27:01.240 00:27:04.920 Amber Lin: insert experience mostly.

264 00:27:05.440 00:27:09.989 Amber Lin: So, Casey, how long would you think that would take.

265 00:27:10.170 00:27:12.739 Casie Aviles: I guess I’d start 1st with

266 00:27:13.900 00:27:17.970 Casie Aviles: how to do this. And honestly, I don’t know yet, because I don’t know if

267 00:27:18.130 00:27:21.460 Casie Aviles: Google Chat supports the animation thing.

268 00:27:21.600 00:27:22.960 Amber Lin: Hmm, okay.

269 00:27:26.040 00:27:31.239 Uttam Kumaran: So. So I would. I would just say like, if it let’s say it does support. What would it be?

270 00:27:32.769 00:27:35.950 Casie Aviles: It’s it’s going to be a change in the code again.

271 00:27:38.170 00:27:41.839 Uttam Kumaran: Think it’s like a that’s like a 5 point thing or a smaller.

272 00:27:43.926 00:27:47.939 Casie Aviles: I guess if I know how to do it, then it’s going to be

273 00:27:48.100 00:27:51.889 Casie Aviles: faster. I just have to edit the code and push it.

274 00:27:53.260 00:27:56.579 Uttam Kumaran: Then I would, I would say, estimate this as a 5.

275 00:27:57.070 00:28:01.090 Uttam Kumaran: And then one thing we have to we have to put in here is that?

276 00:28:01.360 00:28:03.050 Uttam Kumaran: Yeah, I would put put 5.

277 00:28:04.060 00:28:08.600 Uttam Kumaran: And then I think, Casey, you should. I think you probably have to do

278 00:28:08.960 00:28:12.280 Uttam Kumaran: as part of this ticket. It’s going to be going and doing discovery

279 00:28:14.298 00:28:16.640 Uttam Kumaran: on like how this is happening. Right?

280 00:28:16.880 00:28:23.490 Uttam Kumaran: So this is where it’s like, this is a good example of, we can go do the research before, but I think it’ll probably take you

281 00:28:24.350 00:28:28.330 Uttam Kumaran: like half an hour right to probably go find out whether this is possible.

282 00:28:28.910 00:28:31.810 Uttam Kumaran: and then executing the ticket, so.

283 00:28:33.140 00:28:35.999 Amber Lin: Yeah, I would say, this is not.

284 00:28:36.240 00:28:43.240 Amber Lin: This is not a problem. So this is an additional ad. So it will be either low or medium.

285 00:28:44.750 00:28:46.239 Uttam Kumaran: I would just put it as low.

286 00:28:46.850 00:28:48.499 Amber Lin: Okay, sounds good.

287 00:28:49.290 00:28:51.490 Amber Lin: Okay. Next one.

288 00:28:53.390 00:28:56.319 Amber Lin: Oh, move the assignee.

289 00:28:57.520 00:28:58.540 Amber Lin: So.

290 00:29:00.440 00:29:01.500 Amber Lin: And

291 00:29:04.570 00:29:14.270 Amber Lin: yeah, this ticket is a bit long. Mostly the goal is to see, we talked about last time of what is contributing

292 00:29:14.850 00:29:24.089 Amber Lin: 2. 1 was contributing to long situation times and 2. What type of questions take, how long?

293 00:29:26.280 00:29:26.960 Amber Lin: One.

294 00:29:35.780 00:29:36.410 Uttam Kumaran: So Anna.

295 00:29:36.410 00:29:42.460 Uttam Kumaran: analyze and reduce bot execution, bring brought response and consistent discussions. Okay, cool. What makes execution time long?

296 00:29:42.620 00:29:44.210 Uttam Kumaran: What types of questions?

297 00:29:49.110 00:29:52.019 Uttam Kumaran: Yeah, this is a, this is a really, this is a good one.

298 00:29:55.200 00:29:59.169 Amber Lin: We still wouldn’t say, this is top top priority. It’s more of a.

299 00:29:59.170 00:30:03.100 Uttam Kumaran: Yeah. I asked every I asked everybody how they felt about the execution time.

300 00:30:03.210 00:30:04.699 Uttam Kumaran: They felt pretty good

301 00:30:04.850 00:30:16.079 Uttam Kumaran: as long as this is like not 30 seconds. We’re doing pretty well, just like I would like us to constantly keep this in mind. I would set this as I would just set this as a medium.

302 00:30:16.320 00:30:20.700 Uttam Kumaran: and then estimate that’s probably for makeup going. Kate.

303 00:30:20.700 00:30:28.850 Amber Lin: Quite a bit of time. Patrick did some of this before, so when Patrick did, he found out how, what?

304 00:30:29.020 00:30:32.499 Amber Lin: What? I guess, what type of

305 00:30:32.900 00:30:44.649 Amber Lin: action the bot was doing that caused a long time and think. We all found out that when it goes to the spreadsheet it was pretty fast with the central talk once took a long time, if I remember it correctly.

306 00:30:45.211 00:30:51.079 Amber Lin: But it was not as granular of what type of questions asked, took. How long.

307 00:30:52.710 00:30:53.350 Casie Aviles: Okay.

308 00:30:53.940 00:30:56.169 Amber Lin: But it was under here.

309 00:30:56.300 00:30:57.539 Amber Lin: What you guys think.

310 00:30:59.880 00:31:01.119 Casie Aviles: Yeah, it’s interesting.

311 00:31:01.350 00:31:05.510 Miguel de Veyra: This one probably will need to break down into smaller tickets.

312 00:31:05.510 00:31:07.469 Uttam Kumaran: It seems like a 5 or an 8.

313 00:31:07.470 00:31:09.359 Miguel de Veyra: Yeah, but we don’t.

314 00:31:09.620 00:31:14.109 Uttam Kumaran: Because part of this is going to be Annie doing some of the analysis, and then the team.

315 00:31:14.250 00:31:17.830 Uttam Kumaran: I would leave it as an 8, and then, considering.

316 00:31:18.330 00:31:21.579 Uttam Kumaran: consider how to break this down, maybe a little bit more. But.

317 00:31:22.100 00:31:22.420 Amber Lin: Yeah.

318 00:31:22.420 00:31:23.110 Uttam Kumaran: Ticket.

319 00:31:24.510 00:31:28.570 Amber Lin: Cool next.

320 00:31:31.180 00:31:39.649 Amber Lin: This was the oh, by the way, once the we checked with Tim a few times, that I don’t think a button is deployed yet.

321 00:31:40.140 00:31:42.720 Amber Lin: So this is a bit blocked

322 00:31:44.709 00:31:49.430 Amber Lin: and the window service is what they sent a while ago.

323 00:31:49.430 00:31:53.859 Uttam Kumaran: Okay. But then I guess this, this is 2 different. So this is already done.

324 00:31:54.520 00:31:55.200 Amber Lin: Hmm.

325 00:31:56.420 00:31:59.580 Uttam Kumaran: Is this already done like our side is done.

326 00:32:02.230 00:32:06.250 Amber Lin: We need to add the offer to the button.

327 00:32:06.620 00:32:09.469 Amber Lin: So the button code is there.

328 00:32:09.927 00:32:19.779 Amber Lin: I don’t know if we have this specific offer we have. Oh, by the ways, General, by the ways when you click that button. But we don’t have this specific offer.

329 00:32:20.220 00:32:26.960 Uttam Kumaran: So how does this work, Casey Miguel? Like, what like? What’s a how does this sound?

330 00:32:26.960 00:32:28.549 Uttam Kumaran: How does the logic work for this.

331 00:32:29.780 00:32:32.714 Casie Aviles: Yeah. So for the oh, by the way, triggers,

332 00:32:33.785 00:32:37.919 Casie Aviles: right now for the button, what we did is that it would read from

333 00:32:38.070 00:32:48.139 Casie Aviles: the input output, and then it would generate, based on that, it would generate the most related. Oh, by the way, offer

334 00:32:49.300 00:32:53.809 Casie Aviles: so it’s it’s the button on top of a an 8 N. Workflow.

335 00:33:11.040 00:33:13.820 Amber Lin: Yeah. So we right now have

336 00:33:13.930 00:33:27.629 Amber Lin: generating it, depending on the context. What they wanted was, now they kind of wanted to test out one offer, so that the Csrs get accustomed to the groove

337 00:33:27.830 00:33:35.059 Amber Lin: didn’t seem like it was a pressing priority, because they were pretty happy with just having the oh, by the ways.

338 00:33:36.000 00:33:40.569 Amber Lin: I think, having the button, there would be helpful. But it’s not there yet.

339 00:33:41.390 00:33:49.449 Uttam Kumaran: I guess there’s there’s like, Mo, there’s like multiple things happening here I’m like, is this stuck on Tim? Is this us like it’s not what. I don’t know what

340 00:33:49.790 00:33:51.780 Uttam Kumaran: what’s going on with it like, who’s

341 00:33:51.960 00:33:55.119 Uttam Kumaran: is this stuck on Tim? Which means we already did work on it.

342 00:33:55.380 00:33:56.050 Amber Lin: And

343 00:33:58.420 00:33:59.170 Uttam Kumaran: Like.

344 00:33:59.340 00:34:01.439 Uttam Kumaran: Then why is this in the backlog.

345 00:34:04.240 00:34:22.179 Amber Lin: The button. Feature is one thing, so, having the feature of a button is done. And now to deploy, this is the second layer, but once we have the button we want to. Now experiment of having a single offer, so we can track

346 00:34:22.350 00:34:23.610 Amber Lin: the

347 00:34:25.659 00:34:28.700 Amber Lin: The upsells results essentially.

348 00:34:29.489 00:34:30.409 Uttam Kumaran: Okay?

349 00:34:33.249 00:34:43.949 Uttam Kumaran: So then I would put something. So then, this is mainly like we have to change some logic on our end in order to just have the the button trigger. One offer.

350 00:34:44.710 00:34:45.230 Amber Lin: Thank you.

351 00:34:45.230 00:34:48.269 Uttam Kumaran: And every time there’s a new offer we have to change this.

352 00:34:51.480 00:34:54.010 Amber Lin: Yeah, I think that’s how it I can see that

353 00:34:54.010 00:34:57.051 Amber Lin: how it works if we do it this way.

354 00:34:57.390 00:35:03.930 Uttam Kumaran: Like, why can’t? Why can’t their their team control? What offer gets put into this like? Why are we in the middle of that.

355 00:35:04.540 00:35:10.029 Miguel de Veyra: Yeah, we can just create another doc, though. The oh, by the something like an Oh, by the way, offers Doc.

356 00:35:10.620 00:35:11.160 Miguel de Veyra: and then we can.

357 00:35:11.160 00:35:13.890 Uttam Kumaran: But why can’t you? Why can’t we just keep in the same, Doc.

358 00:35:15.040 00:35:19.510 Miguel de Veyra: Because it will run through everything again like just for the oh, by the way, bottom.

359 00:35:20.700 00:35:24.040 Uttam Kumaran: But why don’t you just Rajax? The one piece you need

360 00:35:25.140 00:35:27.410 Uttam Kumaran: or like? Cut the one piece you need?

361 00:35:29.840 00:35:40.740 Uttam Kumaran: This is what I’m saying is like, I don’t want to maintain another piece of code that all it does is trigger. This logic, which means, if they change every day, then we have to take on a task every day.

362 00:35:40.860 00:35:46.610 Uttam Kumaran: so why can’t? If it’s in the dock, why can’t we just get it out of the dock? Just get that one piece out of the dock.

363 00:35:51.560 00:35:54.030 Miguel de Veyra: Okay, yeah, we’ll look at the stock specific.

364 00:35:54.030 00:36:00.820 Uttam Kumaran: But do you know what I mean like? Why, we that team like they should indicate which offer is the one they want to push

365 00:36:01.030 00:36:09.359 Uttam Kumaran: our our agent should be able to. We either need to just like, give it just that one section, or we say, only push

366 00:36:10.420 00:36:16.179 Uttam Kumaran: some piece. But yeah, I don’t want. I don’t want us to be in the middle of updating what offer they want to do. Because.

367 00:36:16.470 00:36:22.130 Uttam Kumaran: yeah, that’s like, this is where I want to avoid as much like hard coded.

368 00:36:22.370 00:36:25.109 Uttam Kumaran: That is possible because we’re gonna get into a bind.

369 00:36:30.080 00:36:36.410 Amber Lin: Okay, so I will. Let’s put this into requirements.

370 00:36:36.840 00:36:46.190 Amber Lin: Let’s think about how we’re gonna do that, is it gonna be one dog. How. Let’s see how.

371 00:36:46.710 00:36:47.490 Amber Lin: So?

372 00:36:49.967 00:36:53.180 Amber Lin: I don’t know what’s out here. I’m just gonna put it there.

373 00:36:53.370 00:36:58.240 Amber Lin: I would say that we go revisit this and do the priority and estimates.

374 00:37:06.890 00:37:09.150 Uttam Kumaran: So what is? Yeah, what is a special test.

375 00:37:10.558 00:37:13.550 Amber Lin: Last time. We we already

376 00:37:15.960 00:37:17.770 Amber Lin: we already did part of it.

377 00:37:19.090 00:37:21.180 Amber Lin: It was last time of

378 00:37:22.070 00:37:39.019 Amber Lin: we needed to send a rodent tech for this specific second story wasp service. And then I asked, Janice, is there any other special cases you want to include? And this is where, she responded. If all we’re all rotary services, we’ll need to send a rodent tech. So this is like a

379 00:37:39.290 00:37:41.200 Amber Lin: bot behavior.

380 00:37:41.780 00:37:43.040 Amber Lin: What behavior? I think.

381 00:37:43.040 00:37:45.450 Uttam Kumaran: So then, like for this? Yeah, like.

382 00:37:46.390 00:37:49.599 Uttam Kumaran: so, how? What are the steps to resolve? Something like this?

383 00:37:55.880 00:38:01.540 Uttam Kumaran: So I get the acceptance criteria. But what is like the steps to resolve? Basically, yeah.

384 00:38:04.650 00:38:07.090 Amber Lin: My assumption that we add something in the prompt.

385 00:38:07.720 00:38:13.100 Uttam Kumaran: Miguel and Casey like, what? How do you like? Given a ticket like this like? What are the steps to resolve this.

386 00:38:16.000 00:38:19.009 Casie Aviles: Yeah, I think it’s just going to be a prompt if

387 00:38:19.280 00:38:22.850 Casie Aviles: yeah, prompt adjustment. If we just want the bot to

388 00:38:23.757 00:38:26.260 Casie Aviles: say something that it doesn’t say.

389 00:38:31.460 00:38:32.040 Uttam Kumaran: Okay.

390 00:38:32.940 00:38:33.480 Amber Lin: Cool.

391 00:38:33.880 00:38:36.410 Uttam Kumaran: And then why do you? How do you end up testing this.

392 00:38:38.244 00:38:47.059 Casie Aviles: Yeah, I guess. Just chat the bot again and recreate like the the exact input output

393 00:38:47.660 00:38:49.760 Casie Aviles: questions that they tested with.

394 00:38:51.510 00:39:02.179 Uttam Kumaran: So this is a good example of like when once it’s tested, and once it’s done, it would be great for us to communicate directly with them and just share. Hey, here’s what the new output is. Is this fine?

395 00:39:02.810 00:39:06.369 Uttam Kumaran: Right? So part of like the acceptance criteria is not

396 00:39:06.540 00:39:13.009 Uttam Kumaran: the fact that we’ve resolved it. But the fact that we also communicated them and they approved the new. They improve the new fix

397 00:39:13.490 00:39:19.460 Uttam Kumaran: so, and maybe we can turn these into tech boxes. Amber. These the acceptance criteria.

398 00:39:19.460 00:39:20.130 Amber Lin: Yeah.

399 00:39:21.600 00:39:23.989 Uttam Kumaran: It’s easy. Yeah, cool.

400 00:39:24.930 00:39:31.800 Uttam Kumaran: So yeah, that this is like a good example of like, I think it’s narrow. But I want us to just resolve these and communicate back

401 00:39:33.810 00:39:39.579 Uttam Kumaran: cause we’re only gonna get be able to take a couple of these per week. So I want them to really help us prioritize like, what support.

402 00:39:47.270 00:39:52.910 Amber Lin: Casey, how big of a ticket do you think this is somewhere here.

403 00:39:55.980 00:39:57.779 Casie Aviles: I think we could put as 3.

404 00:39:59.040 00:40:06.909 Amber Lin: Oh, okay, 3 points is that it takes half a day. If I remember correctly.

405 00:40:08.212 00:40:11.140 Casie Aviles: Yeah, we’ll need to have them approve as well, right?

406 00:40:11.820 00:40:16.519 Casie Aviles: So as part of yeah, that’s why I added a bit.

407 00:40:17.950 00:40:23.539 Amber Lin: Sounds good. We have 20 min. Let’s do the trainer bot.

408 00:40:24.250 00:40:26.670 Uttam Kumaran: Yeah, I have to hop soon.

409 00:40:26.870 00:40:27.330 Amber Lin: Yeah.

410 00:40:27.330 00:40:29.379 Uttam Kumaran: Well, let’s talk about the trainer bought, and then.

411 00:40:29.540 00:40:30.170 Amber Lin: D.

412 00:40:32.290 00:40:33.190 Uttam Kumaran: Yeah.

413 00:40:34.830 00:40:36.069 Amber Lin: It’s the

414 00:40:39.520 00:40:44.649 Amber Lin: that one feedback to suggestion. But

415 00:40:46.770 00:40:48.180 Amber Lin: You see, I’m a girl.

416 00:40:49.980 00:40:51.260 Amber Lin: Where do you think.

417 00:41:00.040 00:41:07.710 Casie Aviles: Okay, so I do, we need to break this down into smaller tickets.

418 00:41:09.241 00:41:10.357 Amber Lin: Depends on you.

419 00:41:11.140 00:41:12.599 Amber Lin: How would it be done?

420 00:41:13.360 00:41:18.197 Casie Aviles: Oh, okay. So I believe the 1st thing that we need to do is to just

421 00:41:18.630 00:41:20.109 Casie Aviles: create a new bot

422 00:41:23.380 00:41:30.310 Casie Aviles: there. So there’s 2 parts to that which is the N. 8 n. Workflow, and then the ui.

423 00:41:31.970 00:41:34.690 Casie Aviles: which is the Google chat ui.

424 00:41:52.900 00:41:57.054 Casie Aviles: Yeah, I think those are the main parts for this.

425 00:41:58.380 00:42:01.220 Casie Aviles: am I? Am I missing something, Miguel? Or.

426 00:42:01.669 00:42:08.419 Miguel de Veyra: And wait. The Y says the next value, add is proactively generate actionable content changes.

427 00:42:08.980 00:42:25.880 Miguel de Veyra: So in the spreadsheet. Didn’t we talk about last week to just basically, if there’s like an error, this bot should trigger. And then it basically, what happens is that it’s gonna look into that, doc, and then suggest the change or like, Hey, we don’t have this. Maybe we can add something like this.

428 00:42:26.500 00:42:29.199 Miguel de Veyra: I think that’s what we talked about last week for this ticket.

429 00:42:30.020 00:42:30.560 Amber Lin: But.

430 00:42:30.970 00:42:34.380 Miguel de Veyra: Like in the spreadsheet. There’s gonna be like a new column.

431 00:42:36.960 00:42:40.620 Miguel de Veyra: Named. But basically suggestion.

432 00:42:41.970 00:42:51.189 Miguel de Veyra: And then that’s where this bought. So I don’t think necessarily, we need a ui for this. It’s we already have the end any end workflow? I I think it’s just a matter of

433 00:42:52.130 00:42:56.600 Miguel de Veyra: like adding it to that work to that total.

434 00:42:56.600 00:42:57.320 Amber Lin: Spreadsheet.

435 00:42:57.750 00:42:58.370 Uttam Kumaran: Yeah, yeah.

436 00:42:58.370 00:43:05.160 Uttam Kumaran: and what what about for? Like one of the is there another ticket in terms of like actually going and updating that, Doc.

437 00:43:07.010 00:43:13.809 Uttam Kumaran: because one of the things that Janice and them are struggling with is actually going in and finding where to update making the update.

438 00:43:17.090 00:43:18.950 Uttam Kumaran: I assume that’s like another ticket.

439 00:43:19.250 00:43:23.232 Amber Lin: Yes, that would be another ticket. I do.

440 00:43:23.980 00:43:24.580 Amber Lin: It’s.

441 00:43:24.580 00:43:25.389 Uttam Kumaran: Yeah. Okay.

442 00:43:25.850 00:43:26.800 Uttam Kumaran: Okay. Cool.

443 00:43:26.960 00:43:31.550 Amber Lin: Yeah. So I would say, it’s in spreadsheet.

444 00:43:32.630 00:43:36.339 Uttam Kumaran: So I would go ask, like, I think this is where

445 00:43:36.490 00:43:41.630 Uttam Kumaran: need to understand? This is like a very much like a human in the loop problem. So you need to

446 00:43:41.770 00:43:49.449 Uttam Kumaran: like, I think we could. We could do a bunch of things. But I just don’t know what the ABC team is like interested in right like do they want?

447 00:43:49.870 00:43:54.519 Uttam Kumaran: If there is a feedback like, where do they want that? Do they want them to spreadsheet

448 00:43:55.040 00:43:58.180 Uttam Kumaran: like? Do they want that to. Yeah.

449 00:43:58.550 00:44:00.920 Uttam Kumaran: this is where I think it’s probably best to just like.

450 00:44:01.080 00:44:03.690 Uttam Kumaran: see what they think about this solution.

451 00:44:04.340 00:44:04.990 Amber Lin: Okay.

452 00:44:06.430 00:44:07.170 Amber Lin: But.

453 00:44:07.170 00:44:09.900 Uttam Kumaran: I feel like if they’re okay with it. Then, yeah, this makes sense.

454 00:44:10.380 00:44:11.550 Amber Lin: Sounds good.

455 00:44:11.780 00:44:20.310 Amber Lin: I think this is lower priority than the formatting and helping them find the updates. But I just wanted to put it there.

456 00:44:20.450 00:44:21.560 Uttam Kumaran: So it’s saying, Yeah.

457 00:44:21.560 00:44:25.100 Uttam Kumaran: I would. I would put this as high and put the other one as high, too.

458 00:44:25.990 00:44:31.140 Amber Lin: Alright, I’ll put the other one as urgent, in my opinion.

459 00:44:31.520 00:44:32.060 Amber Lin: But.

460 00:44:32.060 00:44:32.620 Uttam Kumaran: Okay.

461 00:44:32.981 00:44:37.679 Amber Lin: What would you guys say is, the estimates would be like 5 ish.

462 00:44:41.260 00:44:41.920 Casie Aviles: Yeah.

463 00:44:42.470 00:44:44.780 Miguel de Veyra: 5. Yeah, I think maybe 5 makes sense.

464 00:44:45.460 00:44:48.220 Amber Lin: Sounds good next one.

465 00:44:49.190 00:44:53.159 Amber Lin: There’s also this, want us to look at.

466 00:44:53.690 00:45:01.879 Amber Lin: helping helping them format the current document and helping them add new content.

467 00:45:04.590 00:45:08.860 Amber Lin: So maybe not even building trainer. By ui, so

468 00:45:09.080 00:45:13.190 Amber Lin: what do you guys think? How are we gonna get? How are we gonna get this done?

469 00:45:16.240 00:45:16.960 Amber Lin: Of.

470 00:45:17.437 00:45:20.300 Miguel de Veyra: Sorry just to clarify this one

471 00:45:20.600 00:45:24.169 Miguel de Veyra: like change the doc in Google, Doc, right?

472 00:45:24.860 00:45:28.580 Miguel de Veyra: It’s only gonna suggest changes. That’s bot friendly.

473 00:45:36.570 00:45:40.879 Uttam Kumaran: Well, like, let’s talk about like, what the problem is like. The problem is that

474 00:45:41.000 00:45:43.040 Uttam Kumaran: they have a change they want to make.

475 00:45:43.310 00:45:49.139 Uttam Kumaran: and one they don’t know where in the dock to go to make it 2. They don’t know what’s the proper format.

476 00:45:49.460 00:45:58.820 Uttam Kumaran: 3. They don’t know whether their answer has all the necessary information, right?

477 00:45:59.990 00:46:07.349 Uttam Kumaran: So it’s a garbage in garbage out problem, like if they go in and just write some garbage, then the bot is gonna return that garbage to all the users.

478 00:46:08.200 00:46:10.499 Amber Lin: What was the 3rd point you mentioned? Sorry.

479 00:46:10.860 00:46:14.739 Uttam Kumaran: Whether it has all. It’s whether it has all the necessary information.

480 00:46:18.570 00:46:21.659 Amber Lin: Sounds good. I think each of them could be. It should be a ticket.

481 00:46:28.280 00:46:34.900 Uttam Kumaran: But this is where, like how much of this can AI take care of versus like, how much of this do you want the user to take care of.

482 00:46:37.260 00:46:39.080 Uttam Kumaran: Like, yeah, so.

483 00:46:42.060 00:46:46.510 Amber Lin: The format is something we can help with.

484 00:46:47.430 00:46:53.969 Amber Lin: Alright, that will be the guidelines that we’re gonna create

485 00:46:55.010 00:46:57.119 Amber Lin: and then helping them format it.

486 00:46:58.262 00:47:02.200 Amber Lin: Whether it has all the necessary information

487 00:47:02.930 00:47:06.339 Amber Lin: could be a back and forth chat.

488 00:47:06.340 00:47:08.700 Uttam Kumaran: Yes, yes.

489 00:47:11.830 00:47:15.610 Uttam Kumaran: so this is one. This is the one thing that Scott wanted to

490 00:47:16.160 00:47:19.319 Uttam Kumaran: sort of chat about which I have a bunch of

491 00:47:19.570 00:47:24.520 Uttam Kumaran: in a bunch of sort of like random

492 00:47:24.730 00:47:28.380 Uttam Kumaran: notes from him. He I think we should have a meeting with him

493 00:47:28.680 00:47:34.050 Uttam Kumaran: because he has some ideas on this, and can probably provide the best requirements.

494 00:47:34.520 00:47:35.070 Amber Lin: Yeah.

495 00:47:35.070 00:47:35.580 Amber Lin: Yes.

496 00:47:35.580 00:47:37.829 Uttam Kumaran: Of which you can go. Probably then

497 00:47:38.690 00:47:40.870 Uttam Kumaran: probably get a better ticket out of.

498 00:47:41.030 00:47:43.080 Uttam Kumaran: So I’m just gonna put some.

499 00:47:44.930 00:47:56.069 Amber Lin: Yeah, let me create. I’m gonna convert this into. Put this in a project. And each of this needs to be a separate ticket. I’m gonna make one about this. So you can put all the requirements in there.

500 00:47:57.840 00:48:07.639 Uttam Kumaran: Yeah. So I I just put some of his notes in in the comments here.

501 00:48:08.000 00:48:08.720 Amber Lin: Okay.

502 00:48:10.070 00:48:16.000 Uttam Kumaran: So basically, he there just wants it to be like

503 00:48:16.860 00:48:26.389 Uttam Kumaran: he. He gave an example of like a back and forth. I want to create a new services. Great. I can help with that. Is this this service? Okay, is a new service. Okay? I want to update this. Okay, what do you want to change

504 00:48:26.510 00:48:34.370 Uttam Kumaran: when it’s done, it produces the knowledge text and then ask, does this look right? And then it could go insert that right? So probably multiple pieces to this, but

505 00:48:35.330 00:48:41.399 Uttam Kumaran: I I put it. I put it in the in the comment on on on the previous ticket, or whatever.

506 00:48:41.780 00:48:42.340 Amber Lin: Hmm

507 00:48:51.040 00:48:59.619 Amber Lin: sounds good. And lastly, it would be a where to put the update.

508 00:49:00.720 00:49:10.479 Amber Lin: How do you guys? We think we should approach this to me it sounds like a control F find in Central Doc problem. Is there any way AI can help with that?

509 00:49:14.211 00:49:17.270 Miguel de Veyra: So far we haven’t no

510 00:49:18.430 00:49:20.980 Miguel de Veyra: cause. Yeah, yeah, for now. No.

511 00:49:21.770 00:49:23.869 Uttam Kumaran: Wait. Why, what do you mean? What are you talking about?

512 00:49:24.320 00:49:26.430 Miguel de Veyra: Like we we can’t.

513 00:49:26.430 00:49:29.480 Uttam Kumaran: The bot can’t tell you where to put the dot. Put the information.

514 00:49:29.480 00:49:37.760 Miguel de Veyra: I mean, you can. Okay, okay, yeah. We can tell you we can. The bot can tell you where to put it, but not change it, or put it there.

515 00:49:38.590 00:49:41.529 Uttam Kumaran: Are you sure? Are you? Okay? Then there’s a 2 proms right.

516 00:49:41.530 00:49:42.150 Miguel de Veyra: Here.

517 00:49:42.470 00:49:44.980 Uttam Kumaran: So are you sure the 4th one is not possible?

518 00:49:46.860 00:49:54.809 Miguel de Veyra: Yeah, we discussed it before that. If we need to basically find it in the find and replace in like programmatically.

519 00:49:56.880 00:50:00.860 Miguel de Veyra: I don’t think I think we run a spike here before. It’s not

520 00:50:01.060 00:50:03.650 Miguel de Veyra: possible. Via Google, Doc. We need something else.

521 00:50:05.260 00:50:09.569 Uttam Kumaran: But I guess this is like, this is what I’m having a trouble understanding like

522 00:50:10.230 00:50:15.970 Uttam Kumaran: you’re telling me. It’s not possible for us to just like, make an edit in the document. And then.

523 00:50:16.340 00:50:19.030 Uttam Kumaran: just like of recreate the document

524 00:50:23.100 00:50:29.820 Uttam Kumaran: like, I’m not saying it has to happen through Google Api. But it’s just a text. It’s a giant text file, right? Why can’t

525 00:50:31.610 00:50:41.150 Uttam Kumaran: can I give you an example in cursor? I can go to cursor, have an agent say, Go, make this update to a piece of code, and it goes and makes the update in line. How is this any different than that.

526 00:50:45.850 00:50:50.439 Miguel de Veyra: So for so far on what we research, we haven’t really

527 00:50:50.910 00:51:01.220 Miguel de Veyra: check. If I I mean, we know it’s not possible via Google Api. So if there’s like another, if it’s probably possible to you, if we use another way.

528 00:51:02.650 00:51:14.679 Uttam Kumaran: Yeah, because because doing this via cursor is like, I mean, I, this is how cursor works. Right? I just put in. I want to make this change to the code, and it goes and finds the line. And then it makes the change right? This is the same, basically the same thing.

529 00:51:18.570 00:51:19.990 Uttam Kumaran: Sorry. Go ahead, Casey.

530 00:51:20.260 00:51:23.950 Casie Aviles: Yeah, I think one of the solutions that we explored for that is.

531 00:51:24.160 00:51:27.439 Casie Aviles: you basically take the entire context.

532 00:51:27.560 00:51:29.710 Casie Aviles: So I mean the entire document.

533 00:51:30.290 00:51:33.810 Casie Aviles: and then the bot would just edit that

534 00:51:33.920 00:51:40.619 Casie Aviles: and then spit out the same document with the edits. I think that’s 1 of the implementations we did.

535 00:51:40.840 00:51:47.120 Casie Aviles: But like, for example, I think what Miguel is talking about is like directly. You know the inline stuff.

536 00:51:47.270 00:51:49.670 Casie Aviles: We’re not sure happily do that.

537 00:51:50.020 00:51:52.379 Casie Aviles: So I yeah, that might be more

538 00:51:53.140 00:51:55.890 Casie Aviles: complex with how that works.

539 00:51:57.610 00:52:01.390 Uttam Kumaran: Okay, okay, that makes sense. I just wanna push on the fact that, like

540 00:52:02.030 00:52:08.939 Uttam Kumaran: the way you do it, sure. But the fact that it’s not possible. I don’t agree with like I mean, cause cause we’re

541 00:52:09.520 00:52:15.720 Uttam Kumaran: we’re able to do that with code right now, like we go into a repo, and I can say, find this piece and make an edit.

542 00:52:15.880 00:52:26.240 Uttam Kumaran: So I think there’s some mechanism. But you’re right in that. That’s that’s let’s separate that out at minimum. I think we can get to like. Here’s what you need to change. And here’s where you can go. Put it.

543 00:52:27.010 00:52:29.109 Uttam Kumaran: The bot being able to go do. That

544 00:52:29.540 00:52:32.029 Uttam Kumaran: is a stretch is a nice stretchful, you know.

545 00:52:34.180 00:52:35.380 Uttam Kumaran: Okay, cool.

546 00:52:36.620 00:52:43.430 Amber Lin: I will separate out into tickets, like all 4 of these should be individual tickets.

547 00:52:45.340 00:52:45.980 Uttam Kumaran: Okay.

548 00:52:46.090 00:52:49.590 Uttam Kumaran: And then Scott wanted to meet about this, so I would just grab.

549 00:52:49.590 00:52:53.999 Amber Lin: I have a I have a ticket to make a make a meeting with him.

550 00:52:55.170 00:53:00.900 Amber Lin: but I just wanted this to look a little bit more organized before I talk to him, to know what we have

551 00:53:01.730 00:53:03.160 Amber Lin: before we talk to him.

552 00:53:03.930 00:53:09.400 Amber Lin: Probably say, like Tuesday or Wednesday, we’ll meet with Scott if I can grab hot.

553 00:53:10.710 00:53:13.909 Uttam Kumaran: Okay, I have to jump to go work on some other stuff.

554 00:53:14.340 00:53:14.819 Amber Lin: Yeah, cool.

555 00:53:15.246 00:53:20.599 Uttam Kumaran: But I know we have probably more grooming to do here, and more stuff for

556 00:53:20.740 00:53:24.579 Uttam Kumaran: the normal AI team. But this is base. This is sort of like

557 00:53:24.930 00:53:34.730 Uttam Kumaran: the individual grooming process that I want to go towards, where basically, we have the goal, like what what the problem is, what the goal is what the acceptance criteria steps to resolve.

558 00:53:36.890 00:53:42.540 Uttam Kumaran: and that’s really clear before we can make an estimate. And before we can assign, and before we move it into cycle.

559 00:53:42.700 00:53:46.350 Uttam Kumaran: So it’s gonna take a couple of more sessions like this to go through that.

560 00:53:48.360 00:53:51.569 Uttam Kumaran: But that’s a that’s what we’ll do. So.

561 00:53:52.210 00:53:52.820 Amber Lin: Yeah.

562 00:53:54.780 00:53:55.730 Amber Lin: Okay, cool.

563 00:53:56.160 00:54:00.649 Amber Lin: Great. I will book another session for us, maybe sometime this week.

564 00:54:02.240 00:54:07.909 Uttam Kumaran: Yeah, we, I mean, we have. We have the stand up tomorrow. So we can just continue to do this tomorrow or

565 00:54:08.390 00:54:10.349 Uttam Kumaran: whatever we want to do.

566 00:54:13.260 00:54:23.529 Uttam Kumaran: Yeah, I I just wanna make sure all the tickets for ABC. And for this all have are really rich in requirements, and that we can actually plan out due dates.

567 00:54:24.052 00:54:30.189 Uttam Kumaran: And then I sent some notes like, I’m I maybe if you guys have time like, I’m still really curious on

568 00:54:30.440 00:54:34.540 Uttam Kumaran: when the internal agent work is gonna like, cause I know.

569 00:54:34.750 00:54:39.030 Uttam Kumaran: I know we landed a bunch of data in last week. I saw there’s stuff in there.

570 00:54:39.390 00:54:41.259 Uttam Kumaran: So I just want to know, like, when

571 00:54:42.100 00:54:44.000 Uttam Kumaran: when we’re gonna be able to see

572 00:54:44.590 00:54:47.869 Uttam Kumaran: some of the agents operational in in slack.

573 00:54:49.570 00:54:50.180 Casie Aviles: Okay.

574 00:54:52.080 00:54:54.919 Uttam Kumaran: So you can just what update any

575 00:54:55.230 00:55:00.319 Uttam Kumaran: status isn’t linear, and then you can slack me. I’ll be on. I’ll be on slack all day. I just have to jump to another meeting.

576 00:55:00.320 00:55:02.230 Amber Lin: Okay. Sounds good.

577 00:55:02.610 00:55:03.160 Casie Aviles: Okay.

578 00:55:04.460 00:55:09.059 Amber Lin: Surprise, we can jump to the AI meeting link.

579 00:55:09.270 00:55:10.669 Amber Lin: If you guys want.

580 00:55:13.940 00:55:14.690 Amber Lin: Okay.

581 00:55:14.690 00:55:18.840 Casie Aviles: So are we going to continue this or go move to the internal work.

582 00:55:19.446 00:55:24.669 Amber Lin: Yeah. Moved. Let’s move to internal, so that our Zoom agent can have a better summary.

583 00:55:25.150 00:55:25.840 Casie Aviles: Okay.

584 00:55:26.170 00:55:27.999 Amber Lin: Yeah, I’ll see. You guys, there.