Meeting Title: Uttam Kumaran Date: 2025-03-19 Meeting participants: Caio Velasco, Casie Aviles, Uttam Kumaran


WEBVTT

1 00:01:10.870 00:01:11.570 Casie Aviles: Hey!

2 00:01:12.300 00:01:13.130 Uttam Kumaran: Hey?

3 00:01:16.520 00:01:23.390 Uttam Kumaran: Okay, sorry for the interruption back to where we were. So

4 00:01:24.220 00:01:30.750 Uttam Kumaran: yeah, we were. So right now, what I’m doing is I’m just making sure that these are all dashboards.

5 00:01:31.290 00:01:33.769 Uttam Kumaran: Right? Looks like this is an explore

6 00:01:34.050 00:01:36.473 Uttam Kumaran: like, what is an explore really like?

7 00:01:37.620 00:01:39.680 Uttam Kumaran: hold on! Let me open up. Arc.

8 00:01:55.150 00:02:00.329 Uttam Kumaran: create an explore dashboard. Oh, okay, so this is a dashboard.

9 00:02:01.360 00:02:04.939 Uttam Kumaran: And this is a metrics view. So this is.

10 00:02:07.460 00:02:10.249 Uttam Kumaran: I’m just gonna rename this to dashboard.

11 00:02:14.770 00:02:16.000 Uttam Kumaran: Okay?

12 00:02:16.600 00:02:25.020 Uttam Kumaran: And then this is a metrics, all feedback model metrics.

13 00:02:25.860 00:02:28.960 Uttam Kumaran: And then we also need to create a new folder oop

14 00:02:30.410 00:02:33.499 Uttam Kumaran: create a new folder called models.

15 00:02:33.960 00:02:36.330 Uttam Kumaran: And this model is gonna be

16 00:02:40.080 00:02:48.300 Uttam Kumaran: Oh, the after we do models before oh, okay, here we go. Yeah. So models.

17 00:02:50.010 00:02:53.850 Uttam Kumaran: And then let me just put this up again.

18 00:03:04.158 00:03:06.310 Uttam Kumaran: Okay, here we go.

19 00:03:13.440 00:03:16.299 Uttam Kumaran: Okay. So ABC, home and commercial.

20 00:03:18.990 00:03:22.510 Uttam Kumaran: What’s there? All feedback model does not exist.

21 00:03:23.570 00:03:25.949 Uttam Kumaran: So let’s back.

22 00:03:29.570 00:03:31.070 Uttam Kumaran: Oh, never mind

23 00:03:38.350 00:03:40.820 Uttam Kumaran: metrics. View not found.

24 00:03:46.600 00:03:50.639 Uttam Kumaran: ABC. Conversation, logs, metrics, dashboard.

25 00:03:53.190 00:03:56.930 Uttam Kumaran: What is this? ABC. Conversation? Logs, metrics.

26 00:03:57.840 00:04:09.070 Uttam Kumaran: Oh, oh, oh, what the heck! Oh, never mind nice.

27 00:04:12.660 00:04:19.600 Uttam Kumaran: And then this is not a dashboard. Right? What is this? Oh, oh, nice. Okay.

28 00:04:21.970 00:04:24.250 Uttam Kumaran: How does this look? Better? Right?

29 00:04:24.690 00:04:28.959 Casie Aviles: Yeah, but it’s just this course for each of the metric

30 00:04:30.620 00:04:38.629 Uttam Kumaran: Well, this is a what is this? This is a ABC Brain trust. Eval’s dashboard

31 00:04:39.120 00:04:42.770 Uttam Kumaran: type metrics view. Oh, so this is also metrics.

32 00:04:43.900 00:04:49.329 Uttam Kumaran: So well, I’m gonna put this here. Then

33 00:04:50.030 00:04:52.419 Uttam Kumaran: I’m gonna rename this to metrics.

34 00:04:55.990 00:05:05.810 Uttam Kumaran: And then also, what I’m gonna do is I’m gonna rename this because ABC is redundant here. I’m just gonna rename this. The brain trust evals metric brain trust evals

35 00:05:05.920 00:05:07.130 Uttam Kumaran: metrics.

36 00:05:07.430 00:05:14.550 Uttam Kumaran: And I’m gonna rename this to conversation logs, metrics.

37 00:05:14.980 00:05:17.680 Uttam Kumaran: And then I’m gonna delete this one.

38 00:05:20.740 00:05:25.059 Uttam Kumaran: And this one. I’m also going to rename to remove BBC,

39 00:05:25.750 00:05:29.440 Uttam Kumaran: and I’m gonna rename this to remove ABC.

40 00:05:33.930 00:05:35.550 Uttam Kumaran: Ask to do this.

41 00:06:10.020 00:06:12.160 Uttam Kumaran: Okay? Sorry you were saying, Casey

42 00:06:13.690 00:06:16.580 Casie Aviles: Oh, yeah, I was just mentioning that for

43 00:06:16.790 00:06:20.950 Casie Aviles: the evils. It’s just the scores that we have there.

44 00:06:21.750 00:06:25.359 Casie Aviles: We don’t have much. I think it’s dimensions. It’s just yeah.

45 00:06:27.230 00:06:28.450 Uttam Kumaran: Hmm!

46 00:06:29.550 00:06:30.690 Uttam Kumaran: So

47 00:06:30.970 00:06:36.150 Uttam Kumaran: talk to me about how these scores work like, is this like out of a 1? This is out of one

48 00:06:38.730 00:06:40.430 Uttam Kumaran: like, is this bad

49 00:06:42.200 00:06:46.960 Casie Aviles: Yes, I think it’s 0 to one. That’s how it’s scored

50 00:06:47.850 00:06:48.570 Uttam Kumaran: Okay.

51 00:06:54.850 00:07:02.269 Casie Aviles: But I think the one that’s the most helpful would be the embedding, since Levenstein is just, I think, measuring the edits

52 00:07:05.335 00:07:05.820 Uttam Kumaran: Okay.

53 00:07:06.740 00:07:07.360 Casie Aviles: Which?

54 00:07:07.640 00:07:13.639 Casie Aviles: Yeah, I guess it’s like a naive approach for measuring

55 00:07:15.680 00:07:21.749 Uttam Kumaran: Hmm, okay. So I wonder if I can

56 00:07:22.950 00:07:26.380 Uttam Kumaran: somehow show this out of a out of one. Always

57 00:07:43.920 00:07:53.400 Uttam Kumaran: time series connector, model table dimensions, measures, valid percent of total.

58 00:08:07.740 00:08:10.720 Uttam Kumaran: Okay, maybe not. Okay. But this is fine.

59 00:08:14.890 00:08:15.810 Uttam Kumaran: Okay.

60 00:08:23.920 00:08:29.320 Uttam Kumaran: so what other data do we get like, can I look at a single run

61 00:08:35.559 00:08:39.629 Casie Aviles: Yeah, I don’t think I’m not sure on real how

62 00:08:39.739 00:08:41.819 Casie Aviles: to look at the single run

63 00:08:43.380 00:08:46.650 Uttam Kumaran: Like. If you go to. So what is the Brainforge Eval?

64 00:08:47.150 00:08:49.179 Uttam Kumaran: These are all the columns you get right

65 00:08:50.380 00:08:52.940 Casie Aviles: That’s filtered like there are other columns that

66 00:08:53.737 00:08:56.359 Casie Aviles: I removed for the time being.

67 00:08:57.350 00:08:59.460 Uttam Kumaran: What other columns are there right now?

68 00:09:03.877 00:09:09.139 Casie Aviles: I think if I think you can go back and to the Ui

69 00:09:10.740 00:09:14.719 Uttam Kumaran: Oh, to the brain trust. Ui, okay? Well, what I’m gonna do now is because I just wanna

70 00:09:14.840 00:09:18.430 Uttam Kumaran: I just wanna switch. I’m just gonna make sure you have this Pr.

71 00:09:18.820 00:09:23.441 Uttam Kumaran: and then I just want to switch to help Kaya with something. But let me

72 00:09:24.610 00:09:27.839 Uttam Kumaran: let me just so brain trust Eval models

73 00:09:33.300 00:09:38.449 Uttam Kumaran: name embedding similarity. So input output expected.

74 00:09:39.080 00:09:46.660 Uttam Kumaran: Oh, okay, so why don’t we do this? So I’m gonna go into the brain trust Eval metrics.

75 00:09:47.010 00:09:55.340 Uttam Kumaran: I’m gonna bring in. I’m gonna bring in this input output expected

76 00:09:56.720 00:10:05.179 Casie Aviles: Yeah. But although the the problem with that is, it’s all null at the moment, because the export did not have the input.

77 00:10:05.740 00:10:07.610 Casie Aviles: Yeah. Did not have these

78 00:10:08.840 00:10:10.230 Uttam Kumaran: Oh, it doesn’t have it yet

79 00:10:10.920 00:10:11.800 Casie Aviles: Yeah.

80 00:10:13.130 00:10:13.790 Uttam Kumaran: Why not

81 00:10:19.440 00:10:22.249 Uttam Kumaran: like, is it? But it’s there in the ui, right

82 00:10:28.990 00:10:29.970 Uttam Kumaran: Maybe no.

83 00:10:29.970 00:10:30.540 Casie Aviles: All right.

84 00:10:33.040 00:10:34.589 Uttam Kumaran: It’s gotta be in the ui right?

85 00:10:38.110 00:10:43.330 Casie Aviles: It is there. But it’s kind of it’s weird, because it’s kind of

86 00:10:43.660 00:10:48.469 Casie Aviles: when we export. It’s just that. And we have to like export each of the

87 00:10:49.836 00:10:52.499 Casie Aviles: I I can share my screen. Just so it’s clear

88 00:10:52.500 00:10:54.610 Uttam Kumaran: Oh, you mean you have to get each of the runs.

89 00:10:54.770 00:10:55.910 Uttam Kumaran: Yes.

90 00:10:55.910 00:10:57.890 Uttam Kumaran: Oh, okay, okay. So then let’s

91 00:10:58.860 00:11:00.789 Casie Aviles: So that’s why I commented it out

92 00:11:01.170 00:11:03.020 Uttam Kumaran: Let’s leave it out for now.

93 00:11:03.780 00:11:07.469 Uttam Kumaran: But let’s look at so total evaluations, average similarity.

94 00:11:51.140 00:11:55.889 Uttam Kumaran: Okay, okay, I mean, this is

95 00:11:57.000 00:12:00.659 Uttam Kumaran: pretty good. I mean, the data looks like

96 00:12:01.320 00:12:03.960 Uttam Kumaran: doesn’t look like we’re getting much better, though.

97 00:12:07.020 00:12:09.040 Uttam Kumaran: except on the embedding score. Right?

98 00:12:09.830 00:12:12.879 Uttam Kumaran: But I guess my question is like, why aren’t we

99 00:12:13.900 00:12:17.690 Uttam Kumaran: is the Brain trust running. Every time someone asks a question

100 00:12:19.016 00:12:22.169 Casie Aviles: Yeah, that should be. Yeah. We set it up that way.

101 00:12:23.300 00:12:25.579 Uttam Kumaran: Okay, but we’re not. Okay.

102 00:12:27.400 00:12:30.459 Uttam Kumaran: okay, this makes sense. I think a couple of things one.

103 00:12:31.070 00:12:32.580 Uttam Kumaran: maybe I’ll send it to the

104 00:12:34.260 00:12:36.410 Uttam Kumaran: I’ll send it to the channel. So you have it.

105 00:12:40.950 00:12:48.750 Uttam Kumaran: The one thing you want to add is, I want to add the inputs, real input, real output

106 00:12:49.080 00:12:50.610 Uttam Kumaran: expected output.

107 00:12:55.190 00:12:57.444 Uttam Kumaran: I also want to look at

108 00:13:03.550 00:13:07.830 Uttam Kumaran: I want to look at like the if we can get the user that issued

109 00:13:08.250 00:13:12.420 Uttam Kumaran: the query, I know it’s all gonna be Miguel now. But maybe we can somehow get that.

110 00:13:12.880 00:13:18.860 Uttam Kumaran: Second thing is, how can we join this data with the chat logs?

111 00:13:21.130 00:13:23.170 Uttam Kumaran: 4th thing is

112 00:13:27.210 00:13:34.120 Uttam Kumaran: we need to bring in the question types from the

113 00:13:35.060 00:13:38.079 Uttam Kumaran: Eval golden Winston Golden data sheet.

114 00:13:40.480 00:13:45.729 Uttam Kumaran: Or, if this is possible, to have in brain trust.

115 00:13:47.950 00:13:52.190 Uttam Kumaran: Do those 4 things make sense? I just throw it in the slack thread

116 00:13:53.155 00:13:54.529 Casie Aviles: Yeah, yeah, that makes sense.

117 00:13:54.780 00:14:00.150 Uttam Kumaran: Okay, then let me I’m gonna go ahead and commit all these a lot of changes. So creating

118 00:14:00.380 00:14:05.850 Uttam Kumaran: new ABC home project, cleaning up files.

119 00:14:07.540 00:14:08.270 Uttam Kumaran: Okay.

120 00:14:08.950 00:14:11.459 Uttam Kumaran: Another thing I’m gonna do here is

121 00:14:12.480 00:14:14.363 Uttam Kumaran: I’m gonna go ahead and

122 00:14:19.580 00:14:20.870 Uttam Kumaran: deploy this

123 00:15:34.060 00:15:36.060 Casie Aviles: Okay, I see it now. Also.

124 00:15:36.580 00:15:44.869 Uttam Kumaran: Okay, and then what is this? ABC home and commercial? Oh, shit.

125 00:15:45.270 00:15:52.760 Uttam Kumaran: Okay, fine. That’s fine. ABC, home and commercial real.

126 00:16:02.710 00:16:09.320 Uttam Kumaran: Okay. Take a look at Pr. You can make a change or make any changes you want, and then let me know when I can review

127 00:16:10.130 00:16:10.850 Casie Aviles: Okay.

128 00:16:11.520 00:16:12.290 Uttam Kumaran: Okay. Cool.

129 00:16:12.830 00:16:14.719 Uttam Kumaran: Hey, Kai, are you on? Okay? Cool.

130 00:16:15.790 00:16:16.650 Caio Velasco: Yes.

131 00:16:18.070 00:16:19.510 Uttam Kumaran: Hey? How’s it going

132 00:16:21.270 00:16:22.170 Caio Velasco: Okay.

133 00:16:23.153 00:16:36.189 Uttam Kumaran: Yeah, I mean, all I’m all I saw was, yeah, I just went in. I was working with Annie. I also think she may be creating that Pr, I saw you made one. So I think she’s gonna also work on one

134 00:16:37.520 00:16:43.432 Uttam Kumaran: she’s just going through the whole process because she has. She maybe needs some more columns. But basically, yeah, I just looked in.

135 00:16:44.870 00:16:50.810 Uttam Kumaran: we just pulled this up. And I found in raw, portable, gorgeous tickets. I just found the customer stuff.

136 00:16:51.590 00:16:58.109 Uttam Kumaran: and then I just brought. I just told her like, make a Pr with these 3 customer id email customer name?

137 00:16:59.139 00:17:04.079 Uttam Kumaran: So I think that’s fine. I assume your question is about the second part. Right?

138 00:17:04.619 00:17:08.829 Caio Velasco: Yeah, yeah, no, this part. I I just did. Now, I didn’t know that she was gonna do

139 00:17:09.230 00:17:12.139 Uttam Kumaran: Sorry I should have made that clear. Yeah, she I was. She was okay. See you

140 00:17:12.140 00:17:15.550 Uttam Kumaran: like, can I do this? And I was like you should. I told her, should she should try

141 00:17:16.859 00:17:21.449 Uttam Kumaran: so maybe give her maybe give her a I saw your peer, maybe give her a chance, but like I think, the

142 00:17:21.450 00:17:21.880 Caio Velasco: Okay.

143 00:17:22.319 00:17:26.739 Uttam Kumaran: The next the next piece is definitely a little, probably a little bit harder.

144 00:17:27.438 00:17:31.499 Uttam Kumaran: But I guess, like, yeah, let me know. You read my notes like, what do you think

145 00:17:32.830 00:17:39.576 Caio Velasco: Yeah, no, no, I understand. Like the yeah. Didn’t also understand why customer was not there. And then, when I was

146 00:17:40.590 00:17:45.419 Caio Velasco: learning about recharges. We try to help Robert

147 00:17:45.991 00:18:08.119 Caio Velasco: at the end of the day. I also felt that the only way to connect gorgeous and recharge would be through customer information. At least I didn’t see any other way. So that’s I think it’s covered already. And then, yeah, then I was trying. I was trying to understand, like constellation. Recovery is what? What? That exactly me and how

148 00:18:08.380 00:18:12.790 Caio Velasco: and how do you know that you have to go in a certain place or

149 00:18:13.270 00:18:13.830 Uttam Kumaran: Yeah.

150 00:18:13.830 00:18:16.160 Caio Velasco: Already had knowledge on recharge, or or

151 00:18:16.160 00:18:17.010 Uttam Kumaran: No, I guess like

152 00:18:17.010 00:18:27.700 Caio Velasco: When I spend time learning about it. It took me like quite a bit of time to even understand, like the charges subscription, how they are related to each other. I had to go and recharge website and understand that.

153 00:18:27.700 00:18:28.840 Caio Velasco: Yeah, I

154 00:18:28.940 00:18:32.255 Uttam Kumaran: So let’s talk through this because I just read it briefly. Also, let me

155 00:18:36.700 00:18:38.886 Uttam Kumaran: do you? Yeah, maybe I’ll just put here.

156 00:18:42.970 00:18:49.389 Uttam Kumaran: continue on your Pr, so you can see the full flow.

157 00:19:02.070 00:19:05.890 Uttam Kumaran: Okay? So yeah, I guess I looked at the 1st question, which is basically like.

158 00:19:06.200 00:19:12.089 Uttam Kumaran: Okay, which tickets directly led to cancellation recoveries. Okay, so what that means is like, okay, there’s

159 00:19:12.560 00:19:19.090 Uttam Kumaran: it’s basically like, okay, can we? How do identify which orders were

160 00:19:19.410 00:19:24.540 Uttam Kumaran: cancelled. But then recovered. And then what tickets led to them?

161 00:19:24.780 00:19:29.929 Uttam Kumaran: So it looks like Robert was like cool macros meaning test saved attempt.

162 00:19:30.300 00:19:32.039 Uttam Kumaran: A macro, of course, is like

163 00:19:32.200 00:19:34.589 Uttam Kumaran: just that. It’s just basically like a copy paste.

164 00:19:34.830 00:19:38.950 Uttam Kumaran: So on steroids kind of like copy paste, the same message

165 00:19:39.230 00:19:42.869 Uttam Kumaran: find percent of customers that have real orders after the macro is used.

166 00:19:43.580 00:19:53.590 Uttam Kumaran: Okay, and for order type equals renewal. So when looking at that, I sort of looked at your notes, I think you were pretty close like, I think this is totally right. So basically, you look at

167 00:19:53.590 00:19:54.060 Caio Velasco: Perfect.

168 00:19:54.060 00:19:59.240 Uttam Kumaran: Tickets where the macro, where a macro that is a save attempt is used.

169 00:20:02.220 00:20:04.769 Uttam Kumaran: Okay, I’ll sponsor it later. So

170 00:20:05.290 00:20:05.859 Caio Velasco: And and

171 00:20:05.860 00:20:06.270 Uttam Kumaran: You’re

172 00:20:06.270 00:20:07.299 Caio Velasco: Just to add, to

173 00:20:07.960 00:20:08.310 Uttam Kumaran: Go ahead!

174 00:20:08.310 00:20:18.980 Caio Velasco: No, just what you’re saying like. When when he mentioned renew, I was trying to find this, and I couldn’t. And then, when I was going to recharge stuff. I understood that the subscription tables

175 00:20:19.640 00:20:22.840 Caio Velasco: they are about preview at the end of the day, because that’s the definition.

176 00:20:22.840 00:20:23.490 Caio Velasco: Yes.

177 00:20:23.490 00:20:24.480 Uttam Kumaran: Description. Right?

178 00:20:24.480 00:20:48.720 Uttam Kumaran: Correct. Yes, that’s what I think. That’s what you. That’s where I was like. Even when I talked to Annie, I was like, did you, Google? I was like, what is gorgeous? She’s like, Oh, I see the table. I’m like, no, no, but what is it like? I was like, she said. What do you mean? I’m like Google. What is gorgeous? And of course, of course, gorgeous is subscriptions meaning any order that comes in is part of a subscription right? So that you you nailed it, which is basically, like all we need to do is find out that

179 00:20:48.850 00:20:53.179 Uttam Kumaran: there was an order after the save attempt happened, which.

180 00:20:53.740 00:21:02.339 Uttam Kumaran: like our our short, our hypothesis is that that means that the save attempt worked right, that ultimately we’re trying to find out? Did the save attempt?

181 00:21:02.570 00:21:18.600 Uttam Kumaran: Did the save attempt work, work or not? Bye, Casey, and so that’s ultimately like what we’re trying to find. So I think again, like I think you you did find like this thing in in renewal and tags, but I wouldn’t base it on that, because

182 00:21:18.920 00:21:19.560 Caio Velasco: Yeah, I see.

183 00:21:19.560 00:21:22.830 Uttam Kumaran: Think tags is like probably something that customer service people are doing.

184 00:21:23.790 00:21:27.780 Uttam Kumaran: Instead. I you could probably just look at.

185 00:21:27.980 00:21:39.209 Uttam Kumaran: find all the tickets. That is a test. Save join that to going that to recharge orders.

186 00:21:39.500 00:21:42.500 Uttam Kumaran: And then basically, all you have to do is like the web

187 00:21:42.820 00:21:45.859 Uttam Kumaran: on the join. You can do a where clause on the join, I think.

188 00:21:45.960 00:21:53.340 Uttam Kumaran: which is like, join on the customer, id their customer email, and where there is a order after

189 00:21:53.880 00:21:57.999 Uttam Kumaran: the ticket resolution date. Basically.

190 00:21:59.540 00:22:04.060 Uttam Kumaran: I can’t think of exactly how to do that in my head right now, but like something like that, basically

191 00:22:04.680 00:22:18.890 Caio Velasco: So so again from the beginning. So we there. There was an order placed, and after some time, for some reason they wanted to cancel the subscription. Something like that. And then something happened after that that.

192 00:22:18.890 00:22:21.160 Caio Velasco: Well, that the be recovered

193 00:22:21.160 00:22:29.020 Uttam Kumaran: Yeah. And and actually, what happened is they they did this what’s called a save attempt, which is they sent like, Hey, we’ll give you a free $5 thing

194 00:22:29.210 00:22:34.489 Uttam Kumaran: if you keep. If you keep your subscription right? And what we’re trying to show here is is that effective

195 00:22:34.950 00:22:42.030 Uttam Kumaran: like, are we? Are we saving 20% by using those macros.

196 00:22:43.710 00:22:47.830 Uttam Kumaran: So the macro, the macro, all it is is just a text. Right? It’s like, Okay, they

197 00:22:47.830 00:22:52.479 Uttam Kumaran: click a button, saying, Use this, macro. Use this macro. All we want to know is.

198 00:22:52.630 00:22:57.690 Uttam Kumaran: are they working like are there? Are the Save? Are the save attempts working

199 00:23:00.710 00:23:05.245 Caio Velasco: Oh, okay. So that means that in the or

200 00:23:06.160 00:23:11.699 Caio Velasco: that in the orders table, we have like duplicators of order, because one

201 00:23:11.820 00:23:16.480 Caio Velasco: was an attempt to cancel, and another one was a saved attempt, something like

202 00:23:16.480 00:23:17.520 Uttam Kumaran: So.

203 00:23:18.000 00:23:22.620 Uttam Kumaran: But you’re not going to see that in orders right? If they cancel, you’re not going to see a future order.

204 00:23:24.010 00:23:32.540 Uttam Kumaran: but let’s say they have an active subscription like, if they reach out to cancel of subscription. That means they have an active one.

205 00:23:32.640 00:23:37.859 Uttam Kumaran: so you could infer that if if they reach out and try to cancel, and then an order is placed 2 weeks later.

206 00:23:38.040 00:23:39.900 Uttam Kumaran: That means it worked right

207 00:23:40.320 00:23:43.919 Caio Velasco: Okay, of course they didn’t. They didn’t cancel. At the end of the day

208 00:23:43.920 00:23:49.249 Uttam Kumaran: Cancel. I think there’s probably some edge cases here. Which is why I was like, I was like.

209 00:23:50.200 00:23:53.570 Uttam Kumaran: basically, I think this should. This should. I was like.

210 00:23:54.530 00:24:00.940 Uttam Kumaran: like, I feel like, just fine. Just call this a. v 1, because, of course, maybe they forget to cancel or like

211 00:24:01.050 00:24:07.810 Uttam Kumaran: they wanted to cancel, and if there’s like some, probably some edge cases, but for a version one, all we need to find is like.

212 00:24:08.070 00:24:12.080 Uttam Kumaran: after every ticket for a customer with a

213 00:24:12.880 00:24:25.369 Uttam Kumaran: that was used that like. And this is another thing is like, you don’t even have to know whether they they wanted to cancel, because, again. I don’t know how you get that, but if they use the test, save you know that they’re trying to cancel it right

214 00:24:26.520 00:24:27.340 Caio Velasco: Yes.

215 00:24:27.340 00:24:35.100 Uttam Kumaran: Because otherwise there’s no. You can basically infer that the ticket is a cancellation attempt because of what was tried to mitigate it.

216 00:24:35.490 00:24:39.270 Uttam Kumaran: So then you can say, Okay, any tickets that test save was used.

217 00:24:39.800 00:24:47.879 Uttam Kumaran: It’s a cancellation attempt out of those. What? How many? How many of those customers? Right? So everything here is we’re looking at

218 00:24:48.020 00:24:52.820 Uttam Kumaran: the we’re looking at number of customers that

219 00:24:52.920 00:24:57.450 Uttam Kumaran: had a test save but still ordered. After after that.

220 00:24:58.580 00:25:05.089 Uttam Kumaran: if you once you have the join. Right? You can then do this as like, okay. How many orders after that? How many customers? It’s probably

221 00:25:05.270 00:25:07.729 Uttam Kumaran: probably have to ask Robert, like exactly what?

222 00:25:07.910 00:25:12.550 Uttam Kumaran: What metric is he trying to look at? How many orders the order value, how many customers. But

223 00:25:12.880 00:25:18.840 Uttam Kumaran: basically I think that should that should that should solve it.

224 00:25:19.560 00:25:26.569 Caio Velasco: No perfect, especially because the customer can try like many times, and then we have to group by that. If we wanna find customers

225 00:25:26.840 00:25:34.219 Uttam Kumaran: Yeah. And it’s all it’s also like, I think that’s good version one just to look at, hey, for anyone that tried to cancel. Did they order after cool? Here’s a number.

226 00:25:34.560 00:25:37.350 Uttam Kumaran: Then we can think about like the edge cases and things like that.

227 00:25:39.030 00:25:40.320 Caio Velasco: Okay, hopefully, okay.

228 00:25:42.610 00:25:48.159 Caio Velasco: okay, sounds cool. So should I spend time doing that? Or do you think any can date

229 00:25:48.386 00:25:57.919 Uttam Kumaran: No, I would do that. I think that might. I think I think you should handle it. I think that’s probably gonna be a little bit tough. I think she’s gonna push the Pr to add the customer id stuff right now.

230 00:25:58.383 00:26:01.419 Uttam Kumaran: So I’ll help her to do that, and we’ll get that pushed through.

231 00:26:01.839 00:26:07.110 Uttam Kumaran: I’ll actually have. I’ll actually have her ask you to review. Are you gonna be? Are you gonna be working any longer today?

232 00:26:07.790 00:26:08.800 Caio Velasco: I’m a late right?

233 00:26:08.800 00:26:09.659 Caio Velasco: Much longer. Yeah.

234 00:26:09.660 00:26:10.230 Uttam Kumaran: Yeah. Okay. Okay.

235 00:26:10.900 00:26:13.129 Caio Velasco: So I’ll I’ll I’ll review

236 00:26:13.649 00:26:22.539 Uttam Kumaran: I’ll review this and make sure it gets pushed through. And then I think by tomorrow. You should have the customer information there and then. Yeah, I think this should be fine.

237 00:26:23.140 00:26:26.420 Caio Velasco: Okay, so I’ll close my Pr and

238 00:26:27.910 00:26:28.260 Uttam Kumaran: Perfect.

239 00:26:29.830 00:26:34.309 Caio Velasco: Okay, then I work on this tomorrow because it’s also a good thing for me to learn

240 00:26:34.310 00:26:39.949 Uttam Kumaran: Yeah, try. I mean again, I don’t know exactly how the join is gonna work. But yeah, give it a shot. And we could talk tomorrow, too.

241 00:26:40.320 00:26:41.610 Caio Velasco: Perfect, perfect, nice.

242 00:26:42.010 00:26:44.069 Uttam Kumaran: Hey, dude? Thanks for taking the time. I know we were just

243 00:26:44.070 00:26:44.460 Caio Velasco: Sing it.

244 00:26:44.460 00:26:45.070 Uttam Kumaran: Middle of this

245 00:26:45.440 00:26:49.880 Caio Velasco: This is quite helpful. This is where I learned the most, so

246 00:26:50.138 00:27:06.439 Uttam Kumaran: No, I’m learning a lot to learning how to like walk through the steps. It’s been a while. So yeah, I’m glad. And I’m actually really happy, Annie, because Annie was like, Aren’t I supposed to go through engineering? And I’m like you should push engineering so as approved. But you should try pushing the Pr. Well, you know. Give it a go

247 00:27:06.440 00:27:07.140 Caio Velasco: Nice

248 00:27:07.140 00:27:07.480 Uttam Kumaran: So.

249 00:27:07.480 00:27:08.460 Caio Velasco: Nice. Nice.

250 00:27:08.840 00:27:11.039 Caio Velasco: Okay. Cool. Sounds good.

251 00:27:11.040 00:27:13.199 Uttam Kumaran: Okay, thanks, Kyle. I’ll talk to you tomorrow.

252 00:27:13.200 00:27:15.270 Caio Velasco: Appreciate. Thank you. Talk. Tomorrow.