Meeting Title: Daily AI Team Sync Date: 2025-03-04 Meeting participants: Miguel De Veyra, Casie Aviles, Uttam Kumaran


WEBVTT

1 00:02:04.350 00:02:05.520 Miguel de Veyra: Hello! Hello!

2 00:02:08.660 00:02:09.350 Casie Aviles: E.

3 00:02:10.401 00:02:13.469 Miguel de Veyra: What happened wrong message on to.

4 00:02:14.310 00:02:15.060 Casie Aviles: Hold on!

5 00:02:15.975 00:02:16.380 Miguel de Veyra: Thanks.

6 00:02:16.620 00:02:19.240 Miguel de Veyra: Yeah, I guess you could start. Could you open it up? And

7 00:02:23.960 00:02:25.950 Miguel de Veyra: yeah, wait, let me share screen.

8 00:02:29.650 00:02:36.260 Miguel de Veyra: So I got, I tried following your instructions here, basically uploading it all.

9 00:02:38.270 00:02:43.980 Miguel de Veyra: Where is it here, though I’m still working on some part of it?

10 00:02:45.160 00:02:47.539 Miguel de Veyra: I don’t know. I just did this like couple of hours ago.

11 00:02:50.560 00:02:55.919 Casie Aviles: No, okay. So so for the metrics, like for the Ragas, it’s it won’t work right.

12 00:02:56.630 00:02:57.360 Miguel de Veyra: Not.

13 00:02:57.650 00:02:59.629 Miguel de Veyra: No, no, I think it’s different.

14 00:03:00.520 00:03:08.780 Miguel de Veyra: Basically, the code, the in the process isn’t run so they can’t really check

15 00:03:13.190 00:03:19.650 Miguel de Veyra: better. Yeah, I mean this should work unable to find output.

16 00:03:24.900 00:03:25.750 Miguel de Veyra: nibble

17 00:03:33.000 00:03:36.640 Miguel de Veyra: final output settings.

18 00:03:48.990 00:03:52.400 Miguel de Veyra: I’m gonna maneuver Max output.

19 00:04:12.550 00:04:13.660 Miguel de Veyra: Hello, Casey.

20 00:04:35.530 00:04:36.400 Miguel de Veyra: I’ll wait.

21 00:05:30.800 00:05:31.540 Casie Aviles: Hello!

22 00:05:32.550 00:05:33.560 Miguel de Veyra: Mobile.

23 00:05:33.970 00:05:39.200 Casie Aviles: Internet Zoom Meeting.

24 00:05:42.840 00:05:44.179 Casie Aviles: See you soon.

25 00:05:44.180 00:05:53.180 Miguel de Veyra: Okay, okay, yeah, technically, yeah, it should have been okay. But it’s just this one output variables.

26 00:05:55.010 00:05:56.430 Miguel de Veyra: What’s happening there.

27 00:06:01.280 00:06:02.370 Miguel de Veyra: Unaffectional.

28 00:06:03.530 00:06:09.560 Casie Aviles: Name Yung final output like instead of final output.

29 00:06:10.210 00:06:11.550 Casie Aviles: But it has nothing like.

30 00:06:11.550 00:06:12.370 Miguel de Veyra: Lisa.

31 00:06:12.930 00:06:15.479 Casie Aviles: Yes, agent output, or something like that.

32 00:06:15.860 00:06:17.220 Miguel de Veyra: Hey? Hey? You don’t.

33 00:06:17.530 00:06:21.539 Uttam Kumaran: Hey? Sorry, guys, I was just sort of zoning out.

34 00:06:22.180 00:06:29.110 Miguel de Veyra: I was actually online. And I still missed it. So okay.

35 00:06:29.882 00:06:32.600 Miguel de Veyra: yeah, basically, I’m we’re just trying to, you know.

36 00:06:32.840 00:06:37.919 Miguel de Veyra: figure something out regarding this evals. I talked to the vellum guys earlier today.

37 00:06:38.270 00:06:44.930 Miguel de Veyra: And yeah, just trying to connect the last few dots. I don’t wanna exit it. But yeah.

38 00:06:45.700 00:06:46.910 Miguel de Veyra: And then.

39 00:06:46.910 00:06:48.060 Uttam Kumaran: How is it? Tell me.

40 00:06:49.636 00:06:53.930 Miguel de Veyra: It’s pretty. It’s doing pretty okay, so far.

41 00:06:56.260 00:07:02.139 Miguel de Veyra: Yeah. But wait, let me just exit it again. So basically, I uploaded everything here.

42 00:07:03.290 00:07:09.619 Miguel de Veyra: the at least the ones from, you know, the golden data sheet. And yeah, just trying to get the

43 00:07:10.240 00:07:13.300 Miguel de Veyra: semantic similarity similarity to work.

44 00:07:14.790 00:07:17.750 Uttam Kumaran: Does it like? How is the usability of vellum.

45 00:07:20.060 00:07:22.460 Miguel de Veyra: It’s I don’t know.

46 00:07:22.870 00:07:23.930 Miguel de Veyra: Confusing.

47 00:07:24.380 00:07:25.180 Uttam Kumaran: Really.

48 00:07:25.390 00:07:28.639 Miguel de Veyra: Yeah, maybe I’m just not used to it, but

49 00:07:29.290 00:07:32.735 Miguel de Veyra: I don’t know. It’s a bit. It’s a lot of.

50 00:07:33.930 00:07:41.630 Casie Aviles: Yeah, it was. It was confusing to me. Also, at 1st I was just telling myself, okay, maybe I just need to figure it out. But

51 00:07:42.270 00:07:42.840 Casie Aviles: yeah.

52 00:07:46.160 00:07:48.029 Uttam Kumaran: What do you think? Should we dish it?

53 00:07:51.630 00:07:57.099 Miguel de Veyra: Oh, I don’t think it’s worth 500 bucks a month, that’s for sure.

54 00:07:57.650 00:07:58.999 Uttam Kumaran: Oh, really. Okay.

55 00:08:00.120 00:08:02.400 Miguel de Veyra: Like, a, yeah.

56 00:08:02.540 00:08:06.199 Uttam Kumaran: I mean we oh, I really. The only thing we’re using it for is this.

57 00:08:06.550 00:08:13.600 Miguel de Veyra: For evils. Yeah, then I tried asking them, because it’s a bit confusing to use this even.

58 00:08:14.200 00:08:18.309 Miguel de Veyra: And then they just gave me their documentation like, basically this one

59 00:08:24.350 00:08:30.270 Miguel de Veyra: cause some of it we can’t really use. Cause we, I was talking with Jana like earlier today.

60 00:08:31.180 00:08:38.759 Miguel de Veyra: we can’t really use since we’re it, some of it for it to work, especially like, where is it? The Ragas ones like

61 00:08:41.059 00:08:48.039 Miguel de Veyra: evils like this? 3 we can’t really use because we’re not actually doing the process inside of valium.

62 00:08:49.650 00:08:51.290 Uttam Kumaran: Oh, okay.

63 00:08:51.520 00:09:00.469 Miguel de Veyra: Yeah, we’re doing majority of the process outside of value. That’s why we can’t really use it. Well, we can still use some search on track to get that to work.

64 00:09:01.890 00:09:03.056 Miguel de Veyra: But the

65 00:09:11.620 00:09:13.429 Uttam Kumaran: Do we want to try another one.

66 00:09:14.495 00:09:17.259 Miguel de Veyra: Yeah, definitely. Go see.

67 00:09:17.400 00:09:23.400 Miguel de Veyra: at the end of the day we’re already using Apis to access whatever it. So it should be.

68 00:09:25.510 00:09:28.710 Miguel de Veyra: Theoretically, we can just add it to the Google Code.

69 00:09:31.280 00:09:32.680 Miguel de Veyra: Worst case scenario.

70 00:09:35.310 00:09:35.880 Uttam Kumaran: That’s

71 00:09:48.610 00:09:51.650 Miguel de Veyra: Or we can, you know, build something

72 00:09:55.760 00:09:58.659 Miguel de Veyra: right? I mean the Snowflake. One has it already.

73 00:10:06.010 00:10:13.549 Uttam Kumaran: I mean, really, what we’re running into is because na, then, is not running in code. It’s hard to insert

74 00:10:13.690 00:10:15.079 Uttam Kumaran: the Evals.

75 00:10:15.080 00:10:15.760 Miguel de Veyra: Yeah.

76 00:10:37.210 00:10:40.380 Uttam Kumaran: Yeah. And this is a like, there’s, there’s a lot of feature.

77 00:10:42.400 00:10:46.660 Uttam Kumaran: There’s a lot of feature requests already for to do this. But

78 00:10:52.920 00:10:54.169 Uttam Kumaran: Here, let me.

79 00:10:54.600 00:10:58.100 Uttam Kumaran: I just okay. Let me try this one other thing.

80 00:10:58.630 00:10:59.220 Miguel de Veyra: Okay.

81 00:11:56.540 00:12:01.009 Uttam Kumaran: Just invited. You guys, I’ve I’ve been reading about this company called Brain Trust.

82 00:12:02.060 00:12:08.909 Uttam Kumaran: They I feel like this might work a little bit better.

83 00:12:19.150 00:12:21.079 Miguel de Veyra: Okay, yeah, I’ll explore. This.

84 00:12:36.310 00:12:36.990 Miguel de Veyra: then.

85 00:12:46.600 00:12:48.719 Miguel de Veyra: is valid via SDK.

86 00:12:59.850 00:13:00.789 Uttam Kumaran: Check it out!

87 00:13:01.130 00:13:01.960 Miguel de Veyra: Yep. Done!

88 00:13:01.960 00:13:05.770 Uttam Kumaran: I don’t want us to spend more time on Bell, and like what we can just ditch it.

89 00:13:06.220 00:13:07.040 Miguel de Veyra: Okay.

90 00:13:07.040 00:13:08.320 Uttam Kumaran: Should I just ditch it now?

91 00:13:09.030 00:13:10.440 Miguel de Veyra: Yeah, probably.

92 00:13:11.420 00:13:13.239 Uttam Kumaran: Do you have to change anything on your side?

93 00:13:16.380 00:13:22.120 Miguel de Veyra: Yeah, because technically, everything runs through value in the Google code. So

94 00:13:22.510 00:13:28.370 Miguel de Veyra: yeah, I will just move it out there and then cause it’s live on their end, and then we’ll message them.

95 00:13:30.780 00:13:35.558 Uttam Kumaran: Okay, you can, whatever. I’ll I’ll leave it up. Let me let me look at one

96 00:13:37.850 00:13:40.259 Miguel de Veyra: Yeah, on. When we started.

97 00:13:40.390 00:13:41.190 Uttam Kumaran: Yeah.

98 00:13:41.350 00:13:43.866 Miguel de Veyra: Wait. Let me check, too.

99 00:13:49.580 00:13:51.320 Miguel de Veyra: February 4.th

100 00:13:52.140 00:13:55.110 Miguel de Veyra: Oh, my God, it’s up to the day. Yeah.

101 00:13:57.300 00:14:00.479 Uttam Kumaran: Let me check. Where the fuck do I go to see billing.

102 00:14:05.750 00:14:08.940 Miguel de Veyra: The so.

103 00:14:12.950 00:14:14.290 Uttam Kumaran: Oh, it’s not even

104 00:14:54.070 00:14:57.609 Uttam Kumaran: I don’t even know where I can go to fucking. Cancel this.

105 00:14:58.560 00:15:02.660 Miguel de Veyra: Wait. I think they sent us a link before we might be able to cancel. There.

106 00:15:03.370 00:15:05.169 Uttam Kumaran: I went there. It’s not working.

107 00:15:11.000 00:15:12.050 Miguel de Veyra: The stripe, one

108 00:15:24.630 00:15:25.630 Miguel de Veyra: environmental.

109 00:16:41.480 00:16:44.440 Uttam Kumaran: That is really stupid. They don’t have a.

110 00:16:44.780 00:16:45.949 Miguel de Veyra: And so bottom.

111 00:16:47.130 00:16:48.310 Uttam Kumaran: It’s insane.

112 00:17:10.910 00:17:13.530 Uttam Kumaran: Okay. Brain trust may work a little bit better.

113 00:17:13.810 00:17:14.420 Miguel de Veyra: Okay.

114 00:17:16.159 00:17:20.729 Uttam Kumaran: I don’t think you you actually you you don’t have to do the prompts in there.

115 00:17:20.899 00:17:26.989 Uttam Kumaran: Think you can do it all via the like you don’t need to have their full workflows in there.

116 00:17:27.510 00:17:31.670 Miguel de Veyra: Okay, okay, I’ll I’m checking out now. I’ll keep you posted.

117 00:17:32.870 00:17:46.180 Miguel de Veyra: And then for the other one. Yeah. Casey was primarily the main person who tackled the we were supposed to discuss it earlier. But yeah, I was a bit stuck on the evil side of things. Casey, do you wanna hop on in there.

118 00:17:46.410 00:17:48.640 Casie Aviles: Yeah, sure. Okay.

119 00:17:52.530 00:17:57.339 Casie Aviles: Well, what I did first, st actually, before going into Advarag was just

120 00:17:58.060 00:18:04.560 Casie Aviles: yeah, I listed that I did some tests with. I chatted with the bot, and then I recorded like.

121 00:18:04.760 00:18:10.419 Casie Aviles: how long the responses took. So I was more focused on just you know, if we could optimize the speed

122 00:18:11.742 00:18:17.360 Casie Aviles: so so based on the tests that I have like the here on the left, these are the

123 00:18:18.686 00:18:22.413 Casie Aviles: like the seconds in seconds. This is how it took, how long it took

124 00:18:23.530 00:18:29.810 Casie Aviles: And then, basically, I just wanted to pinpoint like which part of the like the whole workflow is.

125 00:18:30.790 00:18:32.689 Casie Aviles: you know, taking.

126 00:18:33.350 00:18:34.049 Miguel de Veyra: The most.

127 00:18:34.050 00:18:37.469 Casie Aviles: Like, yeah, we could like, which part could we optimize? So

128 00:18:38.190 00:18:42.669 Casie Aviles: yeah, basically, this is like, this is for the entire n, 8, n, 1. So

129 00:18:44.207 00:18:47.339 Casie Aviles: when we when we ask it like this mosquito

130 00:18:47.960 00:18:50.930 Casie Aviles: question, like, it takes a long, a long time. So

131 00:18:52.390 00:18:55.100 Casie Aviles: yeah, those are just some things I’ve found.

132 00:18:56.300 00:19:00.160 Casie Aviles: Yeah, like here. So I listed like the top 10 longest. So it.

133 00:19:00.590 00:19:02.870 Casie Aviles: like the worst, is 95 seconds, but

134 00:19:03.650 00:19:12.240 Casie Aviles: 30 seconds, 50 like that. And then it’s always questions about mosquitoes. So yeah, and then

135 00:19:12.450 00:19:17.559 Casie Aviles: for the lowest ones, it’s just mostly about, you know, trivial questions.

136 00:19:20.330 00:19:23.260 Casie Aviles: And yeah, I guess interesting.

137 00:19:24.950 00:19:26.510 Casie Aviles: Yeah. And then

138 00:19:26.860 00:19:35.910 Casie Aviles: basically, I think there’s this part, the cloud function part. If we if we’re able to like optimize this part, so maybe we can start getting better response times.

139 00:19:39.200 00:19:48.159 Casie Aviles: yeah. So this one, yeah, this is basically what I found out. And by average, we have like, yeah, 14 seconds of run time. So.

140 00:19:51.556 00:19:52.229 Uttam Kumaran: See? Okay?

141 00:19:53.810 00:19:56.179 Uttam Kumaran: So it’s really not. It’s. But

142 00:19:56.420 00:19:59.430 Uttam Kumaran: why is the Google Chat taking that long.

143 00:20:03.310 00:20:10.280 Casie Aviles: I think there’s some overhead here, like with the Google function, the run, the Google Cloud run.

144 00:20:10.480 00:20:13.560 Casie Aviles: So it’s like the function that’s required, for

145 00:20:13.810 00:20:20.780 Casie Aviles: I think it passes like, like, I think it’s an intermediate step between N. 8 N. And then the Google Chat interface.

146 00:20:21.280 00:20:21.970 Uttam Kumaran: Okay.

147 00:20:22.630 00:20:30.089 Casie Aviles: Yeah. So I think if if we manage to improve the time there, maybe we we could start getting closer to like 10 seconds. So.

148 00:20:31.860 00:20:36.009 Miguel de Veyra: Where does Casey, where does vallum

149 00:20:37.160 00:20:40.500 Miguel de Veyra: add in time? Cause, I assume valent should add in time, right.

150 00:20:42.190 00:20:44.860 Casie Aviles: All right. Vellum, I actually, I haven’t.

151 00:20:45.010 00:20:52.330 Casie Aviles: Yeah. I haven’t checked the logs on vellum, but this is like the time I got the

152 00:20:52.780 00:20:56.189 Casie Aviles: message minus the N. 8 n. Time. This is it.

153 00:20:57.563 00:21:00.839 Miguel de Veyra: So it could be. Part of the cloud function. Is Valentine now?

154 00:21:01.450 00:21:02.580 Casie Aviles: Yes, possibly.

155 00:21:02.840 00:21:03.570 Miguel de Veyra: Okay. Okay.

156 00:21:03.640 00:21:05.299 Casie Aviles: And could also be Valentine.

157 00:21:06.660 00:21:08.589 Casie Aviles: But yeah, I didn’t check that part.

158 00:21:09.010 00:21:09.930 Casie Aviles: Excellent.

159 00:21:10.310 00:21:11.390 Casie Aviles: Here it is.

160 00:21:12.440 00:21:15.799 Miguel de Veyra: Okay, yeah, that should be, we’ll check more data.

161 00:21:16.550 00:21:17.220 Casie Aviles: Okay?

162 00:21:17.635 00:21:23.479 Casie Aviles: And yeah. And maybe if we could check the the mosquito one like, why, I’m not sure why, it’s taking very long.

163 00:21:23.480 00:21:42.770 Miguel de Veyra: Yeah, no, no. The the reason is taking it’s very long is because it’s checking the sheets, and then, of course, it won’t find there. No, it will find it there. But the thing is, it’ll find every mosquito tech in every zip code. I think I showed it to you yesterday, right like it gave us everything. But it’s not really the question. The question was that.

164 00:21:42.960 00:21:43.989 Casie Aviles: Oh! So!

165 00:21:43.990 00:21:44.520 Miguel de Veyra: Yeah.

166 00:21:44.520 00:21:50.380 Casie Aviles: We have more. If we have more contacts, the bigger tokens, then it’s going to take longer, right.

167 00:21:50.540 00:21:55.999 Miguel de Veyra: Yeah, yeah, I mean, we just, I think we just have to be more descriptive on the tools. And then what to pass.

168 00:21:56.210 00:21:59.089 Miguel de Veyra: For example, you know, it shouldn’t be asking if

169 00:21:59.270 00:22:03.560 Miguel de Veyra: what’s your mosquito services? Why are you asking the sheets? Right? It’s more on the

170 00:22:04.120 00:22:14.740 Miguel de Veyra: I think it’s just prompt adjustments. But yeah, that’s the reason it’s taking so long because it’s checking both context and then the other. The 1st context, the sheets. It’s getting everything

171 00:22:14.960 00:22:19.069 Miguel de Veyra: like every mosquito tech they have across. You know.

172 00:22:19.180 00:22:23.959 Miguel de Veyra: every city in Austin, I mean in Texas. That’s why it’s slow.

173 00:22:25.070 00:22:25.740 Casie Aviles: Okay?

174 00:22:26.871 00:22:29.440 Casie Aviles: Okay. But yeah, that’s as far as I.

175 00:22:29.610 00:22:33.060 Casie Aviles: You know, the observations that I made. And then.

176 00:22:33.540 00:22:39.570 Casie Aviles: yeah, just for the actual advance around. That that’s still in progress. But I I plan to have it by.

177 00:22:39.780 00:22:43.590 Casie Aviles: I want to have it. I want to test by today and see if we could improve

178 00:22:44.306 00:22:47.959 Casie Aviles: and then yeah, and check against these execution times.

179 00:22:48.610 00:22:53.930 Casie Aviles: I was also briefly checking out, Oh, yeah.

180 00:22:55.000 00:23:01.459 Casie Aviles: And yeah, I think for for this 1st part like it works. Really, it worked really well. But

181 00:23:01.590 00:23:03.080 Casie Aviles: I think we need to like

182 00:23:04.680 00:23:10.450 Casie Aviles: fine tune the settings a bit more, because when it when over here, right? So you can see

183 00:23:11.010 00:23:18.169 Casie Aviles: like started to get it’s not complete. Yeah, like, it’s not complete like it gets cut off. But

184 00:23:19.030 00:23:24.820 Casie Aviles: the good thing is these are already chunked. So you could just export this as a Json file. And then.

185 00:23:25.580 00:23:27.960 Casie Aviles: yeah, we could start vectorizing it.

186 00:23:28.150 00:23:28.960 Casie Aviles: So

187 00:23:29.330 00:23:37.969 Casie Aviles: yeah, it it does the chunking for us. And it uses also ocr, which is good. But yeah, we need to this part not so good.

188 00:23:39.680 00:23:42.010 Casie Aviles: And then it starts turning into.

189 00:23:42.340 00:23:42.810 Uttam Kumaran: Bye.

190 00:23:42.810 00:23:44.092 Casie Aviles: Fine over here.

191 00:23:44.960 00:23:46.960 Casie Aviles: Started. Yeah, this one’s but.

192 00:23:46.960 00:23:48.800 Miguel de Veyra: This is the data cleanup side no.

193 00:23:49.270 00:23:52.509 Casie Aviles: Yeah, we might need to clean this up a bit more. So

194 00:23:52.730 00:23:56.309 Casie Aviles: we get the better results. But this should make it easier to, you know. Jump.

195 00:23:56.310 00:24:04.140 Miguel de Veyra: These are all in, by the way, like every thing here is in. I have a script that basically converts all of this into

196 00:24:04.660 00:24:05.770 Miguel de Veyra: the Json.

197 00:24:07.150 00:24:13.770 Miguel de Veyra: I’ll send it to you, because if you go set 2010, the sheets data. It’s everything. Is there in Json structured Json.

198 00:24:14.560 00:24:23.939 Casie Aviles: Oh, okay, okay, yeah, that’s also the outlook like, it looks like this one.

199 00:24:24.230 00:24:26.860 Casie Aviles: So it gives you a chunk. Id.

200 00:24:29.490 00:24:38.580 Casie Aviles: yeah, so basically, we could just feed this this into our vectorization process. And then, yeah.

201 00:24:39.760 00:24:41.060 Miguel de Veyra: Okay, yeah, that’s good.

202 00:24:42.880 00:24:44.500 Casie Aviles: Yeah. But overall.

203 00:24:44.800 00:24:50.009 Casie Aviles: I guess that’s it for for me for now. But yeah, I want to keep working on

204 00:24:50.940 00:24:54.980 Casie Aviles: figuring out the updates, and also like the execution. Times.

205 00:24:56.640 00:24:59.750 Uttam Kumaran: They? The guy from Bellum just sent like a

206 00:24:59.870 00:25:03.100 Uttam Kumaran: way to run vellum external suites.

207 00:25:05.180 00:25:06.790 Uttam Kumaran: Is this something we can try.

208 00:25:07.660 00:25:08.836 Miguel de Veyra: Alright. Let me check

209 00:26:06.580 00:26:09.040 Miguel de Veyra: I need to check more with them in this way.

210 00:26:10.210 00:26:10.930 Uttam Kumaran: Okay.

211 00:26:12.840 00:26:19.710 Uttam Kumaran: yeah, I I just let’s see, I don’t know. I mean, basically, they’re like we could do this externally outside of Lm.

212 00:26:21.810 00:26:22.509 Miguel de Veyra: Yeah, okay.

213 00:26:28.070 00:26:30.320 Miguel de Veyra: But yeah, anyways, I’ll.

214 00:26:30.320 00:26:37.380 Uttam Kumaran: His. Only his only thing was like we should build it inside of here, but we already tried that, and it’s a lot more complicated than any. Then.

215 00:26:37.540 00:26:40.390 Miguel de Veyra: Yes, it would take too much time. Yeah.

216 00:26:43.400 00:26:48.819 Uttam Kumaran: Okay, try it out if it like, spend like 30 min. If it’s not possible, then let’s just can it.

217 00:26:49.130 00:26:50.310 Miguel de Veyra: Okay. Okay. Sure. Sure.

218 00:26:51.030 00:26:51.660 Uttam Kumaran: Okay.

219 00:26:54.630 00:26:55.520 Uttam Kumaran: Okay. Alright.

220 00:26:55.710 00:27:00.109 Miguel de Veyra: I was sorry sorry. Go ahead.

221 00:27:00.390 00:27:01.499 Uttam Kumaran: No go ahead!

222 00:27:01.500 00:27:06.370 Miguel de Veyra: Oh, what’s the name of the Pm. Again? I thought he was gonna be joining this call.

223 00:27:06.370 00:27:07.200 Uttam Kumaran: Steven.

224 00:27:07.200 00:27:08.210 Miguel de Veyra: Yes, Steven.

225 00:27:11.387 00:27:14.729 Uttam Kumaran: He’s just working on one client right now.

226 00:27:14.730 00:27:15.960 Miguel de Veyra: Oh, okay. Okay. I see.

227 00:27:16.950 00:27:17.630 Uttam Kumaran: Yeah.

228 00:27:18.280 00:27:22.760 Miguel de Veyra: Okay, should I? Message Janice today.

229 00:27:23.030 00:27:24.840 Miguel de Veyra: just to follow up on this stuff.

230 00:27:26.790 00:27:28.960 Uttam Kumaran: I’ll send a note out. Yeah.

231 00:27:29.337 00:27:36.019 Uttam Kumaran: yeah. Yesterday I had to just do a bunch of work for one of our data clients. So I’m sort of moving on past that.

232 00:27:36.550 00:27:37.609 Miguel de Veyra: Okay. Sure. Sure.

233 00:27:37.790 00:27:38.350 Uttam Kumaran: Okay?

234 00:27:38.850 00:27:43.900 Uttam Kumaran: And then, yeah, we I’ll I’ll brief you on what we’re we’re gonna do for the next phase proposal.

235 00:27:45.320 00:27:46.500 Miguel de Veyra: Okay. Sure. Sure.

236 00:27:48.050 00:27:48.700 Uttam Kumaran: Okay.

237 00:27:49.580 00:27:50.770 Miguel de Veyra: Okay. Thanks. Everyone.

238 00:27:51.270 00:27:52.219 Uttam Kumaran: Thanks guys.

239 00:27:52.430 00:27:53.539 Casie Aviles: Thanks. Guys, thank you.