Meeting Title: AI Service Standup Date: 2026-01-26 Meeting participants: Samuel Roberts, Mustafa Raja, Pranav Narahari, Gabriel Lam, Casie Aviles, Amber Lin, Clarence Stone


WEBVTT

1 00:00:14.040 00:00:14.820 Mustafa Raja: Hey.

2 00:00:15.340 00:00:16.870 Samuel Roberts: Hey, how are you?

3 00:00:17.540 00:00:18.609 Mustafa Raja: Good, how are you?

4 00:00:19.470 00:00:22.760 Samuel Roberts: Doing okay. We’re still getting a lot of snow here.

5 00:00:23.340 00:00:24.100 Mustafa Raja: Oh.

6 00:00:24.110 00:00:30.179 Samuel Roberts: But yeah, I heard there’s a snowstorm or something, right? Yeah, it’s a huge storm, like, all across the country.

7 00:00:30.970 00:00:34.849 Mustafa Raja: We got a lot of snow, I was already out of shoveling this morning.

8 00:00:36.510 00:00:39.580 Mustafa Raja: Yeah, it snowed here also in Pakistan a lot.

9 00:00:40.140 00:00:40.989 Samuel Roberts: Oh, really?

10 00:00:41.380 00:00:44.559 Mustafa Raja: Yeah, it doesn’t usually, but this year it did.

11 00:00:44.850 00:00:46.140 Samuel Roberts: Yeah, that’s crazy.

12 00:00:48.740 00:00:50.120 Samuel Roberts: How was your weekend?

13 00:00:51.310 00:00:54.710 Mustafa Raja: Went out to see some family.

14 00:00:55.210 00:00:59.740 Mustafa Raja: And played some Red Dead Redemption.

15 00:01:02.190 00:01:07.240 Samuel Roberts: Oh, nice! I haven’t touched that. My sister played a lot of that, so I watched a lot of it, but I didn’t play a ton of it.

16 00:01:07.990 00:01:12.940 Mustafa Raja: Actually, I finished it, last night, and…

17 00:01:13.110 00:01:14.000 Samuel Roberts: Oh, nice!

18 00:01:15.370 00:01:18.089 Mustafa Raja: It’s a sad ending, it’s a sad…

19 00:01:18.090 00:01:21.000 Samuel Roberts: Is it? Oh, I don’t… I gotta, I gotta play it, don’t tell me, don’t tell me.

20 00:01:21.130 00:01:21.690 Samuel Roberts: Someday.

21 00:01:22.560 00:01:23.290 Mustafa Raja: Yeah.

22 00:01:23.520 00:01:27.309 Mustafa Raja: You should, it’s, it’s a good game. One and two, both.

23 00:01:28.060 00:01:30.440 Samuel Roberts: Good, good, yeah, I’m pretty sure I have them, I guess…

24 00:01:30.850 00:01:33.030 Samuel Roberts: Haven’t gotten around to them in years.

25 00:01:33.860 00:01:35.389 Mustafa Raja: Yeah, you need to play those.

26 00:01:35.390 00:01:36.080 Samuel Roberts: Yeah.

27 00:01:36.080 00:01:38.589 Mustafa Raja: Did you play The Last of Us or something?

28 00:01:39.330 00:01:42.699 Samuel Roberts: I have not played The Last of Us, actually. I kind of went through a…

29 00:01:42.830 00:01:46.139 Samuel Roberts: A period of not a ton of gaming when a lot of games were coming out, so…

30 00:01:46.490 00:01:49.040 Samuel Roberts: I, a bit behind.

31 00:01:49.520 00:01:51.970 Mustafa Raja: I haven’t either, but I want to play…

32 00:01:52.890 00:01:55.219 Mustafa Raja: I just don’t get the time.

33 00:01:55.560 00:01:57.060 Samuel Roberts: Exactly, exactly.

34 00:01:57.580 00:01:58.800 Samuel Roberts: Hey, everyone.

35 00:02:00.700 00:02:01.530 Pranav Narahari: Hey, guys.

36 00:02:01.800 00:02:02.490 Gabriel Lam: You know what I mean?

37 00:02:02.490 00:02:03.400 Samuel Roberts: this morning.

38 00:02:05.890 00:02:06.500 Pranav Narahari: Cold.

39 00:02:07.620 00:02:14.879 Samuel Roberts: Yeah, yeah, I was already outside shoveling a bunch this morning, so I’m, like, like, ready to be done with the day already.

40 00:02:15.030 00:02:17.720 Samuel Roberts: Ugh, so much snow out there right now.

41 00:02:18.320 00:02:20.049 Samuel Roberts: That’s still coming down, I think.

42 00:02:23.490 00:02:24.019 Pranav Narahari: Oh, boy.

43 00:02:24.220 00:02:24.909 Pranav Narahari: I like them.

44 00:02:25.260 00:02:27.740 Samuel Roberts: Yeah, Cleveland, Ohio, so right on the lake.

45 00:02:29.090 00:02:31.270 Pranav Narahari: Did you, like, get into the negatives there?

46 00:02:32.070 00:02:37.569 Samuel Roberts: Oh, I mean… I don’t actually know if the wind chill certainly hit the negatives. Oh, okay.

47 00:02:38.090 00:02:45.429 Samuel Roberts: But it’s… what is it right now? It’s 14 degrees outside, which honestly is warmer than it was the day before it started snowing. It was, like, 8, I think?

48 00:02:46.350 00:02:49.490 Samuel Roberts: 8 Fahrenheit, for… For those of us internationally.

49 00:02:50.500 00:02:54.329 Samuel Roberts: Just to be, you know… it’s very cold, is the point. Yeah.

50 00:02:54.540 00:02:55.859 Pranav Narahari: Yeah. Oh, man.

51 00:02:57.920 00:02:59.250 Samuel Roberts: Have a decent weekend?

52 00:03:02.320 00:03:04.290 Gabriel Lam: Trying to stay warm.

53 00:03:04.290 00:03:07.940 Samuel Roberts: Yeah, yeah, oh boy, yeah. How much did you guys get there?

54 00:03:08.550 00:03:11.759 Gabriel Lam: I think we got 16 to 20 inches.

55 00:03:12.220 00:03:14.579 Samuel Roberts: Wow. Yeah, that’s more than we got.

56 00:03:14.700 00:03:16.040 Gabriel Lam: Definitely.

57 00:03:17.770 00:03:24.500 Samuel Roberts: Yeah, I saw some of my friends from high school on the group chat sending pictures and stuff, and I was like, oh man, this seems like a lot.

58 00:03:25.770 00:03:26.740 Gabriel Lam: Oh, boy.

59 00:03:27.220 00:03:30.539 Samuel Roberts: Yeah, alright, well, stay warm. Feel better, too.

60 00:03:30.670 00:03:31.819 Samuel Roberts: Thank you. Damn.

61 00:03:32.520 00:03:33.540 Samuel Roberts: Okay, good.

62 00:03:34.730 00:03:38.209 Samuel Roberts: Alright, I guess, let’s… let’s jump in and start the week.

63 00:03:38.750 00:03:47.349 Samuel Roberts: What is, I guess, the… ABC, let’s start there.

64 00:03:50.610 00:03:51.740 Samuel Roberts: Where are we at?

65 00:03:55.260 00:03:58.259 Casie Aviles: I guess I can start… with ABC,

66 00:03:59.620 00:04:02.489 Casie Aviles: So, last week, I just,

67 00:04:03.890 00:04:06.630 Casie Aviles: Tucked on… or, like, I just connected both.

68 00:04:07.000 00:04:11.879 Casie Aviles: Patricia and Denise and… I think…

69 00:04:12.100 00:04:21.500 Casie Aviles: The next action items there are going to fit into our, like, maintenance that we,

70 00:04:21.839 00:04:26.470 Casie Aviles: Allotted for them, so it’s just going to be, like,

71 00:04:27.150 00:04:30.020 Casie Aviles: Fixes to the prompt, and then…

72 00:04:30.880 00:04:32.509 Samuel Roberts: Okay. As well as with.

73 00:04:33.070 00:04:36.260 Casie Aviles: Current, like, error handling.

74 00:04:36.420 00:04:40.829 Casie Aviles: I think that was the main… the main one that Patricia was…

75 00:04:41.340 00:04:48.560 Casie Aviles: concerned about, and additionally, now that we have the admin UI, Ow.

76 00:04:49.140 00:04:51.629 Casie Aviles: working, and we have Janice.

77 00:04:52.950 00:04:56.670 Casie Aviles: chain, or, like, I’ve… walk… walk Jenny’s through.

78 00:04:56.950 00:04:58.959 Casie Aviles: How to use it, so…

79 00:05:00.010 00:05:07.059 Casie Aviles: We should be able to, you know, implement fixes much faster when it’s related to zip codes, but…

80 00:05:07.200 00:05:09.909 Casie Aviles: Yeah, there’s also, like, just some…

81 00:05:10.720 00:05:15.610 Casie Aviles: Improvements that can be done as well, so we spotted a few as we went through.

82 00:05:16.090 00:05:17.950 Samuel Roberts: Right. So those can be…

83 00:05:18.410 00:05:21.090 Casie Aviles: Also part of our, like, maintenance.

84 00:05:21.300 00:05:26.060 Casie Aviles: Additionally, like, in terms of, like, the goals that we have in the Gantt chart.

85 00:05:26.470 00:05:29.640 Casie Aviles: I’m a little bit behind on, like.

86 00:05:30.100 00:05:34.520 Casie Aviles: Some, yeah, some of those… tickets there, so I think…

87 00:05:35.120 00:05:39.540 Casie Aviles: if I were to pick, like, which one was the highest priority, that’s probably the

88 00:05:40.160 00:05:48.729 Casie Aviles: Getting the master agent, deployed, because right now, I think it’s just on…

89 00:05:50.600 00:05:54.799 Casie Aviles: There… I think, yeah, we’re still using the N8N endpoint.

90 00:05:54.800 00:05:55.520 Samuel Roberts: Yes.

91 00:05:56.410 00:05:57.550 Casie Aviles: At the moment.

92 00:05:58.380 00:06:06.380 Casie Aviles: So I think that that’s probably, like, one… one of the… I think that’s high priority, you know, because…

93 00:06:07.220 00:06:12.859 Casie Aviles: some of the issues that we found, like, I haven’t, like I mentioned, like, with the…

94 00:06:13.340 00:06:17.570 Casie Aviles: network errors, timeout errors that Mustafa helped me surface.

95 00:06:18.800 00:06:25.249 Casie Aviles: might be resolved, like, I believe that could be resolved once we have, like, the master agent up and running.

96 00:06:25.780 00:06:26.210 Samuel Roberts: Yeah.

97 00:06:26.210 00:06:32.479 Casie Aviles: So I think that might be what I need to focus on for… this cycle.

98 00:06:33.400 00:06:38.920 Samuel Roberts: Yeah, let’s talk… let’s talk Maestra, because we were doing some… we were looking at the evals.

99 00:06:38.920 00:06:39.460 Mustafa Raja: Yep.

100 00:06:39.460 00:06:42.750 Samuel Roberts: Mustafa ran, so let’s talk about that real quick and see where we’re…

101 00:06:43.400 00:06:49.190 Mustafa Raja: Yeah, so, yeah, so I ran the evals over the weekend,

102 00:06:49.330 00:07:00.709 Mustafa Raja: So, what I found is, the dataset that we have, doesn’t really, work well with the DB, because,

103 00:07:01.040 00:07:09.959 Mustafa Raja: What would happen is we would have the database updated, and the eval would be for something older.

104 00:07:10.390 00:07:23.910 Mustafa Raja: So, for thumbs up, it doesn’t really make sense. It just says that it’s not giving the right answer, but really, it is giving the right answer, because it’s the new updated data.

105 00:07:23.910 00:07:41.290 Mustafa Raja: So, need to do something about that. And then other than that, I had some, what’s it called, rate limits on the model, for some of the rows. Right. So, I might try, I might rerun it, but overall, it was, it was good.

106 00:07:42.640 00:07:46.729 Samuel Roberts: Okay, yeah, I think sorting out exactly what, like… and this might be something…

107 00:07:46.860 00:07:53.510 Samuel Roberts: we need to put a little more thought into, because the… the ZipsDB and stuff will be changing

108 00:07:53.790 00:07:54.430 Samuel Roberts: So, the.

109 00:07:54.430 00:07:55.070 Casie Aviles: Yeah.

110 00:07:55.280 00:07:58.029 Samuel Roberts: The question… like, our golden data set is not…

111 00:07:58.130 00:08:00.580 Samuel Roberts: Gonna be standing still, so we might need to pull…

112 00:08:01.360 00:08:04.809 Samuel Roberts: Some questions that we know are good from, like, the central doc or something.

113 00:08:06.430 00:08:10.850 Samuel Roberts: And then maybe just do some spot checks with the ZipsDB, maybe.

114 00:08:11.430 00:08:25.139 Mustafa Raja: Yeah, and, when we are sending the data to, Snowflake, if we could label, if this question is, from Central Doc or from database, that would be really nice.

115 00:08:25.310 00:08:27.830 Samuel Roberts: Yeah.

116 00:08:28.670 00:08:35.369 Mustafa Raja: filter the right thing, or get the right thing. Okay. And then maybe we could work more on database.

117 00:08:35.510 00:08:37.500 Samuel Roberts: Based on the data we have.

118 00:08:38.200 00:08:41.569 Samuel Roberts: Right, right, that’s an interesting idea, okay. So…

119 00:08:43.610 00:08:48.310 Samuel Roberts: Is there a way to do that currently with the N8N, or we should just probably getting switched.

120 00:08:49.180 00:09:08.440 Mustafa Raja: I think there, there would be, because, for the… when we get the results from, the agent, we, and the results are from the central doc, the result would have, departments at the end, right? So, from which department,

121 00:09:08.440 00:09:09.040 Samuel Roberts: Oh, sure.

122 00:09:09.040 00:09:10.189 Mustafa Raja: came from, right?

123 00:09:10.780 00:09:14.460 Mustafa Raja: So, detecting that, we could, we could label it.

124 00:09:15.460 00:09:28.590 Mustafa Raja: We could see if this is, if this response has, has a department assigned to it or not. If it does, it’s from, Central Doc. If it does not have any department assigned, it’s most likely from database.

125 00:09:29.720 00:09:30.350 Samuel Roberts: Okay.

126 00:09:31.050 00:09:36.839 Samuel Roberts: That’s probably our bad ideas. We probably want to do the same thing with the, yeah, Monster and the NA9 if we can, but…

127 00:09:37.730 00:09:45.790 Mustafa Raja: Yeah, with the Mastra one, we’d have a lot better setup, because the agents are independent there, right?

128 00:09:45.790 00:09:46.490 Samuel Roberts: Right, right.

129 00:09:48.860 00:09:49.480 Samuel Roberts: Okay.

130 00:09:50.340 00:09:53.079 Samuel Roberts: Yeah, I’m thinking maybe we should have a…

131 00:09:53.750 00:09:56.999 Samuel Roberts: chat about all that. Actually, in terms of deploying.

132 00:09:57.000 00:09:57.449 Casie Aviles: Boss for now.

133 00:09:57.450 00:10:02.380 Samuel Roberts: everything, make sure we’re all on the same page there, but we don’t need to take up the time now, so maybe let’s grab some time today.

134 00:10:02.730 00:10:04.210 Mustafa Raja: Yeah, that’d be nice.

135 00:10:04.480 00:10:05.070 Samuel Roberts: Okay.

136 00:10:05.730 00:10:12.099 Samuel Roberts: Yeah, let’s do that, make sure we have a plan for… for that, for the evals, and getting something stood up.

137 00:10:12.670 00:10:15.070 Samuel Roberts: For actually, like, switching over.

138 00:10:16.430 00:10:17.070 Samuel Roberts: Cool.

139 00:10:17.070 00:10:22.620 Mustafa Raja: So the average, the average score for evals was about, 70.

140 00:10:22.720 00:10:28.769 Mustafa Raja: Out of 100. Now, that also includes, the time-limited responses.

141 00:10:28.770 00:10:29.240 Samuel Roberts: Right.

142 00:10:29.240 00:10:39.129 Mustafa Raja: responses where the database would be giving a different answer, but that’s the new updated answer, where it would, score them a lot lower.

143 00:10:39.520 00:10:42.509 Samuel Roberts: Right, okay, so that’s… that’s… that’s encouraging.

144 00:10:42.870 00:10:43.610 Samuel Roberts: Okay.

145 00:10:44.410 00:10:47.900 Samuel Roberts: Yeah, let’s grab some time to really think this through later, because I think…

146 00:10:48.120 00:10:53.859 Samuel Roberts: We can make a better plan for the evals, and then also for getting it out there, and having more data coming in.

147 00:10:54.100 00:10:56.439 Samuel Roberts: Once we get it, deployed.

148 00:10:58.040 00:10:58.600 Samuel Roberts: Okay.

149 00:10:58.600 00:10:59.000 Mustafa Raja: Yep.

150 00:10:59.000 00:11:01.839 Samuel Roberts: Cool. Other ABC thoughts, then?

151 00:11:01.960 00:11:03.060 Samuel Roberts: For… for now?

152 00:11:03.470 00:11:05.060 Samuel Roberts: Or… this meeting, at least.

153 00:11:05.060 00:11:07.630 Amber Lin: We went there to…

154 00:11:07.850 00:11:12.580 Amber Lin: Can you just add the tickets in, or I can try to add the tickets?

155 00:11:14.670 00:11:17.140 Samuel Roberts: I’m sorry, you’re a little quiet, I didn’t catch all that.

156 00:11:17.160 00:11:24.629 Amber Lin: Oh, sorry, I wasn’t on my mic. Could you guys add the tickets in? Or Casey, I can add it based on…

157 00:11:24.790 00:11:29.969 Amber Lin: what you sent, but just want to make sure that we can update the timeline in here.

158 00:11:30.790 00:11:37.460 Samuel Roberts: Yeah, I think, we can do that probably after we meet, and make sure we’re all in… the Gantt is all up to date, Linear’s all up to date.

159 00:11:37.460 00:11:40.599 Amber Lin: Cool, awesome. Yeah, let me know when that happens.

160 00:11:41.020 00:11:42.080 Samuel Roberts: Okay, will do.

161 00:11:45.250 00:11:51.630 Samuel Roberts: Good, okay, so we’ll… we’ll kick that stuff to later, and we can focus on that. I guess that’s good on ABC, then?

162 00:11:53.200 00:11:54.370 Samuel Roberts: Any other thoughts there?

163 00:11:54.370 00:11:55.490 Casie Aviles: It’s another one.

164 00:11:55.870 00:12:01.460 Samuel Roberts: Okay, great. Yeah, so that’ll be a good… we’ll have a good working session later to make sure we’re all on the same page there, update everything, and…

165 00:12:01.740 00:12:03.059 Samuel Roberts: Do that. Okay, great.

166 00:12:03.780 00:12:06.649 Samuel Roberts: Let’s jump to… Lilo, then.

167 00:12:11.140 00:12:15.680 Pranav Narahari: Yeah, on my end, the…

168 00:12:16.270 00:12:20.409 Pranav Narahari: we were working… Sam, I think I kind of updated you, like, end of day on Friday on, like, the…

169 00:12:20.410 00:12:20.990 Samuel Roberts: Yeah.

170 00:12:21.510 00:12:23.129 Pranav Narahari: the Shopify data.

171 00:12:23.990 00:12:24.710 Samuel Roberts: Yeah.

172 00:12:24.710 00:12:39.119 Pranav Narahari: kind of confusing me right now. Basically, what I’m going to do as, like, the last test is getting access to the Shopify dashboard, which Bobby’s, very okay with, based on our conversation on Friday.

173 00:12:39.230 00:12:40.200 Pranav Narahari: Okay.

174 00:12:40.460 00:12:41.679 Pranav Narahari: And then…

175 00:12:41.820 00:12:50.099 Pranav Narahari: I’m hoping there I’ll find that, like, okay, there’s certain filters that I’m not able to see based on, like, the screen recording Bobby showed briefly.

176 00:12:50.760 00:12:51.450 Samuel Roberts: Okay.

177 00:12:51.970 00:12:58.980 Pranav Narahari: And… hoping that, yeah, that helps out a little bit. If not, I’ll just have to, like, dive a little bit deeper into that.

178 00:12:59.170 00:13:12.470 Pranav Narahari: It must just be something small, because… and I know for a fact that we should be able to pull this data, because they’re just providing that API key for a different, like, software tool that they’re using, and they’re able to pull in that data.

179 00:13:13.230 00:13:21.480 Samuel Roberts: Yeah, so it’s… the data’s there somehow, we just gotta get it the right way, or whatever exact, like, query it needs to be to match what they’re seeing. I see what you’re saying.

180 00:13:21.480 00:13:25.019 Pranav Narahari: And with the key that we have, we have the correct permissions, too, so, like…

181 00:13:25.050 00:13:26.660 Samuel Roberts: Oh, certainly, yes, yes.

182 00:13:26.820 00:13:27.460 Pranav Narahari: Yeah.

183 00:13:27.770 00:13:34.089 Samuel Roberts: Good, good, okay. Is there any… have you seen anything online about people talking about, like, API versus dashboards on Shopify?

184 00:13:34.800 00:13:38.179 Pranav Narahari: I’ve heard a lot of people just complaining about the API.

185 00:13:38.180 00:13:39.420 Samuel Roberts: Okay, that’s…

186 00:13:39.420 00:13:39.930 Pranav Narahari: Yeah.

187 00:13:40.190 00:13:41.280 Samuel Roberts: I’m like…

188 00:13:41.290 00:13:47.870 Pranav Narahari: too many fields, or, like, kind of not that intuitive. Like, they’re GraphQL, like…

189 00:13:48.280 00:13:50.930 Samuel Roberts: Yeah, they switched everything over to the GraphQL stuff, didn’t they?

190 00:13:50.930 00:13:52.730 Pranav Narahari: Yeah. So, yeah.

191 00:13:52.970 00:13:58.490 Pranav Narahari: I think it’s just a little bit of a pain, but I think it’s something we can figure out.

192 00:13:58.850 00:13:59.370 Samuel Roberts: Okay.

193 00:13:59.370 00:13:59.980 Pranav Narahari: Cheers.

194 00:14:00.580 00:14:04.180 Samuel Roberts: Okay, yeah, if you need any help debugging that or anything, let me know.

195 00:14:04.240 00:14:04.790 Pranav Narahari: Totally.

196 00:14:04.980 00:14:05.790 Pranav Narahari: Yeah.

197 00:14:05.890 00:14:15.389 Samuel Roberts: I spent a minute trying to just, like, dump the data, and, like, go into the MCP code, and just, like, whatever we were getting at, just have it, like, logging…

198 00:14:15.660 00:14:21.780 Samuel Roberts: Everything, just to see, like, is there’s, you know, is there a way to, like, get closer to it, maybe?

199 00:14:22.240 00:14:29.819 Pranav Narahari: Yeah. Basically, with, like, the data that we’re pulling in right now, I feel like I’ve tried every single combination to figure out

200 00:14:29.870 00:14:46.790 Pranav Narahari: how they’re calculating net sales. I already know they’re doing it in a… like, what they’re calculating isn’t exactly net sales, because they’re not taking into consideration refunds. Oh, sure. I figured that out on Friday, because Bobby told me, like, yeah, actually the number we’re looking for is…

201 00:14:46.870 00:14:50.600 Pranav Narahari: like… Gross sales minus discounts.

202 00:14:50.700 00:14:51.180 Pranav Narahari: Oh.

203 00:14:51.180 00:14:51.770 Samuel Roberts: Okay.

204 00:14:52.250 00:15:00.439 Pranav Narahari: Yeah, so at least I have, like, a definition at this point for, like, all of the different fields, it’s just, for some reason, the data itself is a little bit off.

205 00:15:01.490 00:15:02.110 Samuel Roberts: Okay.

206 00:15:02.110 00:15:05.109 Pranav Narahari: I also wonder if this may be just, like,

207 00:15:06.970 00:15:25.779 Pranav Narahari: maybe even the way that Bobby’s reading it might be different than, like, the way he’s, like, explaining it to me, too. So, like, I think once I’m in the dashboard, I can just ask more relevant questions to Bobby about, like, oh, well, this is, you know, calculating this, is this what you’re looking for? So yeah, I think things should get cleared up.

208 00:15:26.100 00:15:33.890 Pranav Narahari: today, or hopefully tomorrow, I’m still waiting on that access, bobby said that he’d give it to me, like, end of day Friday.

209 00:15:33.890 00:15:34.630 Samuel Roberts: Yeah.

210 00:15:34.630 00:15:40.019 Pranav Narahari: But, yeah, I didn’t see anything in my email, so I just, messaged again in the Slack channel.

211 00:15:40.810 00:15:41.430 Samuel Roberts: Okay.

212 00:15:41.550 00:15:45.750 Samuel Roberts: Yeah, I saw that. I would say, you know, if you’re still having trouble.

213 00:15:45.910 00:15:49.890 Samuel Roberts: Steve, have you checked out the docs and, like, the agent that it has there?

214 00:15:50.400 00:15:56.700 Pranav Narahari: Yeah, yeah, so I used that, because I remember you mentioning it on Friday as well, like, so I tried using that.

215 00:15:56.920 00:16:03.540 Pranav Narahari: It wasn’t really understanding my, like, super specific query, which was… Finding out, like.

216 00:16:03.810 00:16:09.650 Pranav Narahari: the… the sales metric, and then also specifically for new customers. Okay.

217 00:16:09.870 00:16:20.230 Pranav Narahari: And so I think that agent just doesn’t have, like, the… I don’t even know if this documentation exists for, like, the specific filter, which is new customers. Right.

218 00:16:20.410 00:16:21.340 Pranav Narahari: Yeah.

219 00:16:22.180 00:16:26.800 Pranav Narahari: So, like, the way that… yeah, I think it’s just, like, a very specific thing,

220 00:16:27.100 00:16:30.649 Pranav Narahari: that at least I know the API supports, because…

221 00:16:30.850 00:16:31.230 Samuel Roberts: Yeah.

222 00:16:31.230 00:16:36.999 Pranav Narahari: like, they’re able to get it through, like, another third-party, like, SaaS tool, just via the API key, so…

223 00:16:37.200 00:16:40.469 Samuel Roberts: Right. Who knows what they’re… what calculations they’re doing on top of that.

224 00:16:40.470 00:16:44.910 Pranav Narahari: Yeah, exactly. They might just be doing… pulling a bunch of data, and then, like…

225 00:16:45.490 00:16:53.149 Pranav Narahari: But it can’t be too bad, is what I think, so even if I have to, like, just do a bunch of calculations and whatever on top of it, like…

226 00:16:53.440 00:16:55.260 Pranav Narahari: It is what it is, it shouldn’t take too long.

227 00:16:56.180 00:17:01.080 Samuel Roberts: Okay, yeah, keep me updated, let me know if I can, you know, you can bounce ideas off me or anything.

228 00:17:01.330 00:17:10.620 Pranav Narahari: Yeah, sure. And then, also from our meeting, we talked, like, there was a few extra things that they asked for, that they kind of…

229 00:17:11.050 00:17:16.379 Pranav Narahari: put pretty high on, like, the priority list, I would say, so probably something we would want to get to them by Friday this week.

230 00:17:16.609 00:17:24.729 Pranav Narahari: There’s 3 things. One of them was the Claude skills for the chat. I think, Sam, you kind of were already, like, talking to them about that.

231 00:17:25.180 00:17:37.709 Samuel Roberts: Yeah, I’ve got it nearly done. There’s a little weird hiccup there, where I need to add the code execution tool in order to potentially run those, which makes it

232 00:17:38.120 00:17:38.980 Samuel Roberts: gives it…

233 00:17:39.170 00:17:47.459 Samuel Roberts: access to that for, like, every query, which it seems to kind of default to, so I’m still trying to hash that out, but it might not be the worst thing in the world. I’m just… gotta test it.

234 00:17:48.040 00:17:54.349 Pranav Narahari: Gotcha. Okay. Cool. Yeah, that seems like it’s very much gonna be able to be get done by Friday.

235 00:17:54.660 00:18:00.070 Samuel Roberts: Oh, easily, yeah, yeah. As long as I’m… yeah, I’m pretty… pretty confident with this one, like, today or tomorrow.

236 00:18:00.350 00:18:04.560 Pranav Narahari: Nice, nice. And then, the two other ones were… they’re, like.

237 00:18:05.180 00:18:10.170 Pranav Narahari: kind of from their, like, vibe-coded branches, the meta dashboard, and then also the Klaviyo tool.

238 00:18:10.330 00:18:15.130 Pranav Narahari: They both seem fairly straightforward.

239 00:18:15.440 00:18:25.569 Pranav Narahari: in terms of, like, what we need to do, I feel like there’s not really any unknowns. But the meta dashboard, I feel like, is a little bit dependent on the data warehouse, so…

240 00:18:26.140 00:18:29.939 Pranav Narahari: I would say this is probably something we’ll…

241 00:18:30.420 00:18:47.979 Pranav Narahari: maybe not get by… get done by Friday, which I think is fine, because our actual, like, MVP or POC milestone is next Friday, and so I think this can definitely get done by next Friday, as well as, like, everything else that’s in the to-do column.

242 00:18:48.490 00:18:49.300 Pranav Narahari: But, yeah.

243 00:18:49.300 00:18:57.620 Samuel Roberts: Yeah, I think that’s… yeah, I would say it’s worth taking a look at what their vibe-coded stuff is, but yeah, like, once we get the Ducky B…

244 00:18:57.920 00:19:06.010 Samuel Roberts: and some data in there, it’ll make doing those way better. Because right now, it would have to be just, like, fetching from the API every time, and I don’t even know what it’s worth.

245 00:19:06.360 00:19:10.659 Samuel Roberts: Doing that for, like, the two weeks that… or a week and a half at this point that it would be.

246 00:19:11.250 00:19:16.099 Pranav Narahari: Yeah, I agree. I think it’s just a little bit of, like, doing extra work for no reason.

247 00:19:16.430 00:19:17.220 Samuel Roberts: Yeah.

248 00:19:17.220 00:19:20.360 Pranav Narahari: Slack reports are a little bit like that as well.

249 00:19:20.520 00:19:29.709 Pranav Narahari: There’s probably a lot of, like, the calc… I mean, the calculation logic should port over pretty quickly, but everything else, you know, is kind of just…

250 00:19:30.150 00:19:34.440 Pranav Narahari: We’re just doing it as, like, an intermediary while we’re working on the data warehouse setup, so…

251 00:19:34.740 00:19:39.049 Pranav Narahari: Yeah. Yeah. But I don’t think it’s a big deal.

252 00:19:40.440 00:19:43.770 Pranav Narahari: Yeah, I also updated the Gantt, too, so…

253 00:19:44.150 00:19:44.940 Samuel Roberts: Okay.

254 00:19:44.940 00:19:45.630 Pranav Narahari: Yeah.

255 00:19:45.990 00:19:47.989 Pranav Narahari: I feel pretty good about everything.

256 00:19:47.990 00:19:50.569 Samuel Roberts: Yeah, I think we’re on track pretty well. Yeah.

257 00:19:51.120 00:19:53.170 Samuel Roberts: Okay, great.

258 00:19:53.430 00:19:59.509 Samuel Roberts: Yeah, I’ll try to knock out the skill stuff today. Keep me updated with the Shopify stuff, and then…

259 00:19:59.680 00:20:07.170 Samuel Roberts: I’m trying to think… plan for the warehouse setup, at least for dev.

260 00:20:07.660 00:20:10.890 Samuel Roberts: An air bite, so maybe…

261 00:20:11.570 00:20:14.340 Samuel Roberts: We can make a plan for that, Yumi and Casey.

262 00:20:14.670 00:20:17.550 Samuel Roberts: I don’t know, Casey, how’s your… how’s your day looking overall?

263 00:20:17.680 00:20:19.460 Samuel Roberts: Or week, I guess I should say, too.

264 00:20:20.860 00:20:32.460 Casie Aviles: Right now, I think I’ll be doing some ABC. I should have some space. I mean, as of now, it looks like I can have space for Lilo.

265 00:20:33.370 00:20:40.780 Samuel Roberts: Okay. Yeah, I would say, maybe let’s talk later, the three of us, and make a plan for that, like, in terms of assignments and stuff.

266 00:20:44.200 00:20:50.570 Samuel Roberts: Because I don’t… I don’t just want to, like, take the… start running with inducting until we make a plan of attack, in terms of, like, air bit by cloud, mother duck.

267 00:20:50.670 00:20:55.009 Samuel Roberts: just hosting DeafDB, so we’ll… we’ll hash all that out.

268 00:20:56.340 00:21:03.929 Pranav Narahari: Yeah, I think that’s a good call, Sam. I think that should be… probably as soon as, like, you get Claude’s skills up and running, let’s try to, like, sync on. Yeah.

269 00:21:04.090 00:21:09.529 Pranav Narahari: the data warehouse set up so we can have, like, something to show to them on Friday.

270 00:21:10.140 00:21:16.899 Samuel Roberts: Yeah, I would say, even if I don’t get the skills sorted out, let’s try to maybe do that later today, just to make a plan.

271 00:21:17.860 00:21:18.920 Pranav Narahari: That works for me.

272 00:21:19.150 00:21:20.549 Samuel Roberts: Cool. Alright.

273 00:21:21.480 00:21:24.809 Samuel Roberts: I think that’s good, then, on Lilo, unless there’s anything else I didn’t think of.

274 00:21:26.070 00:21:27.740 Samuel Roberts: I think we’re sitting on track there.

275 00:21:29.500 00:21:30.740 Samuel Roberts: Alright, yeah, Gabe.

276 00:21:31.110 00:21:32.260 Samuel Roberts: Internal? Yeah.

277 00:21:32.260 00:21:40.260 Gabriel Lam: I mean, the only update is, wondering if you’ve had a chance to review the PRD, to update.

278 00:21:40.980 00:21:43.380 Gabriel Lam: our linear flows. That’s what we’ve.

279 00:21:43.380 00:21:49.969 Samuel Roberts: Yes, I… I took a look at it. I wouldn’t call it a review, but I looked at it. I will do that. I will owe you a full,

280 00:21:50.230 00:21:55.499 Samuel Roberts: response, I guess, later. Yeah, I appreciate it. Yeah, the only other thing I’m really looking at is…

281 00:21:59.610 00:22:07.269 Gabriel Lam: trying to look at the Slack assistants that we’ve always done, and I’m trying to figure out where they are in GitHub, so I’ll be reviewing that as well.

282 00:22:08.130 00:22:08.550 Samuel Roberts: Okay.

283 00:22:11.060 00:22:13.330 Gabriel Lam: Yeah, that’s it for me. Sorry about that.

284 00:22:13.700 00:22:17.989 Samuel Roberts: No, no, you’re good, you’re good. You know, stay warm, feel better.

285 00:22:18.300 00:22:19.350 Gabriel Lam: Yep.

286 00:22:19.780 00:22:24.519 Samuel Roberts: Cool. Alright, anything else overall, for the week?

287 00:22:25.560 00:22:28.800 Samuel Roberts: I think it’s… I think we got our work cut out for us pretty well this week.

288 00:22:31.670 00:22:34.370 Samuel Roberts: Yeah. Any other residual thoughts?

289 00:22:37.930 00:22:39.980 Samuel Roberts: Alright, cool. So, if…

290 00:22:40.660 00:22:46.940 Samuel Roberts: I was gonna say, if we can get an ABC meeting on the books, and then a Lilo meeting on the books.

291 00:22:47.290 00:22:50.279 Samuel Roberts: If you guys wouldn’t mind, just… I’m not sure,

292 00:22:50.730 00:22:56.300 Samuel Roberts: My day looks pretty open, I got a couple… well, we have a sink later, don’t we? Or no, maybe we’ll just…

293 00:22:56.770 00:22:57.770 Samuel Roberts: Extend that.

294 00:23:00.030 00:23:02.839 Pranav Narahari: Yeah, maybe something later in the day, too, would be…

295 00:23:02.840 00:23:03.640 Samuel Roberts: That’s fine.

296 00:23:03.640 00:23:09.269 Pranav Narahari: Yeah, and I’ll just, I just have that sync there, we can probably just, like, remove it for this week.

297 00:23:09.750 00:23:15.190 Samuel Roberts: Okay, that’s cool, yeah, so if, yeah, if you can get something on the calendar later, and then…

298 00:23:16.310 00:23:20.590 Samuel Roberts: Mustafa or Casey, if one of you want to get something on there for ABC, we can just…

299 00:23:21.530 00:23:23.179 Samuel Roberts: Make a plan for each of those.

300 00:23:24.930 00:23:25.990 Samuel Roberts: Appreciate it.

301 00:23:26.790 00:23:27.970 Casie Aviles: Ugh.

302 00:23:28.580 00:23:30.439 Casie Aviles: Okay, I gave it a look.

303 00:23:32.690 00:23:36.059 Samuel Roberts: Alright, thank you all. I’ll be…

304 00:23:36.200 00:23:40.270 Samuel Roberts: I’ll be here working on the skills stuff, so if you need me for anything, feel free to ping me.

305 00:23:40.470 00:23:42.879 Samuel Roberts: And then, yeah, have a good day!

306 00:23:43.250 00:23:47.000 Samuel Roberts: Stay warm if you’re in the cold, and yeah.

307 00:23:48.590 00:23:49.360 Samuel Roberts: Alright.

308 00:23:50.500 00:23:51.620 Pranav Narahari: Thanks, Sam.

309 00:23:51.930 00:23:53.150 Samuel Roberts: Alright, have a good one, y’all.

310 00:23:54.550 00:23:55.260 Mustafa Raja: Thank you.

311 00:23:55.770 00:23:56.330 Samuel Roberts: Yep.