Meeting Title: AI-Data Platform Team Standup Date: 2025-07-30 Meeting participants: Sam Roberts, Uttam Kumaran, Vashdev Heerani, Mustafa Raja, Casie Aviles, Miguel de Veyra


WEBVTT

1 00:01:32.040 00:01:33.799 Uttam Kumaran: Hello! Good morning!

2 00:01:33.980 00:01:35.100 Sam Roberts: Good morning!

3 00:01:36.900 00:01:38.020 Uttam Kumaran: How’s everything?

4 00:01:38.646 00:01:42.233 Sam Roberts: It’s going alright. It’s nice to get in and play around a little bit.

5 00:01:42.490 00:01:43.260 Uttam Kumaran: Nice.

6 00:01:44.050 00:01:44.620 Sam Roberts: Yeah.

7 00:01:46.910 00:01:48.200 Uttam Kumaran: Hey? How’s everything?

8 00:01:50.290 00:01:52.780 Vashdev Heerani: Everything is very well. Going very well.

9 00:01:55.040 00:02:01.580 Uttam Kumaran: Cool. Maybe we can just jump into stuff. Yeah, maybe, Sam, I just saw your message. But yeah, like, how’s it going so far.

10 00:02:02.136 00:02:07.823 Uttam Kumaran: I think hopefully. Kind of got a sense of like what we’re using the platform for a little bit and things like that. But

11 00:02:08.060 00:02:08.740 Sam Roberts: Yeah.

12 00:02:08.740 00:02:09.389 Uttam Kumaran: Yeah.

13 00:02:10.080 00:02:10.630 Sam Roberts: Yeah,

14 00:02:11.350 00:02:40.639 Sam Roberts: yeah. So I was exploring the Github and I got super base. And then I finally got access to certain things. And so I was doing a lot of just kind of understanding what’s there? And I think the other. There were 2 things on linear besides just the like overall architecture stuff, and I was like, let me just jump in. I had explored a few tools yesterday or Monday. And one of them looked really promising, and I figured, let’s take a swing at getting it going. And now that I actually have it running locally, I was like.

15 00:02:41.200 00:02:47.450 Uttam Kumaran: Yeah, so what are the features that had? I saw your screenshot? And I saw your note? But it looks like it also has chat stuff, too.

16 00:02:47.450 00:02:52.240 Sam Roberts: It has chapters, so we don’t. I don’t think we have like a a file for chapters, generated.

17 00:02:52.240 00:02:57.510 Uttam Kumaran: We would probably, but what we would probably do is like we would use AI to generate the chapters.

18 00:02:57.510 00:03:04.499 Sam Roberts: Yeah, like, the transcript is the same file format. Basically. From what I’m understanding, I haven’t done a ton with

19 00:03:04.650 00:03:11.109 Sam Roberts: transcripts and subtitles and things before. But yeah, I was able to drop the transcript right in, and it became captions.

20 00:03:11.310 00:03:12.680 Sam Roberts: Oh, great, nice!

21 00:03:12.680 00:03:23.215 Sam Roberts: It’s got some keyboard controls. I was messing around with, you know, basic, like volume up and down, seek. And you can control how much forward and backwards. So it’s defaults to 10, which I think was what someone was requesting in there.

22 00:03:23.460 00:03:24.030 Uttam Kumaran: Oh!

23 00:03:24.340 00:03:28.295 Sam Roberts: And then I was trying to get like speed up and slow down working

24 00:03:29.390 00:03:31.459 Sam Roberts: because I don’t know about you guys. But I’m always watching.

25 00:03:31.460 00:03:34.269 Uttam Kumaran: Oh, I have to watch it. No, I have to watch it like 2 x.

26 00:03:34.270 00:03:43.830 Sam Roberts: Yeah, so I’ve I’ve got it was that was a little broken. It’s a pretty so I don’t know the details of all of this. But this Vidstack player seemed to kind of come out of another

27 00:03:43.950 00:03:54.223 Sam Roberts: like video provider company that I think is shut down, and they kind of open sourced this player. So there’s still like some activity on the github.

28 00:03:55.470 00:04:00.139 Sam Roberts: but it’s not you know. It’s not like it’s just an open source project.

29 00:04:00.140 00:04:00.890 Uttam Kumaran: Point.

30 00:04:00.890 00:04:05.899 Sam Roberts: So there’s still some bugs and things like the. For some reason the speed up slow down.

31 00:04:06.360 00:04:15.813 Sam Roberts: It wasn’t working the way it was supposed to. But it’s pretty easy to just add a function that increases the speed. And so I got that working now.

32 00:04:16.560 00:04:19.410 Sam Roberts: but yeah, it was just nice to get in and click around a little bit and.

33 00:04:19.410 00:04:20.070 Uttam Kumaran: Cool.

34 00:04:20.070 00:04:23.439 Sam Roberts: Make some changes to the code. And yeah, it’s

35 00:04:23.720 00:04:28.600 Sam Roberts: this feels good to get something done a little bit. So cleaning that up a little bit. I was

36 00:04:28.910 00:04:35.499 Sam Roberts: just getting the audio working because it was defaulting to the old player, and it can handle audio files pretty cleanly, too. So

37 00:04:35.650 00:04:37.820 Sam Roberts: hopefully, I’ll be able to get that done.

38 00:04:38.582 00:04:41.120 Sam Roberts: Get something up. But yeah, just.

39 00:04:41.120 00:04:49.230 Uttam Kumaran: And then for the overall architecture stuff like, I didn’t know if you ended up writing something down or putting any notes anywhere. But I’m also happy.

40 00:04:49.740 00:04:55.580 Uttam Kumaran: like, if you want to just create a doc. Because I was, I have some time later today, I was gonna sort of stream, a consciousness.

41 00:04:55.580 00:05:03.489 Sam Roberts: Okay, yeah, I’ll throw. I’ll get a doc. I have like some notes just in, like my local city, and like little things here and there, but I’m still kind of getting a feel, for like

42 00:05:03.700 00:05:05.030 Sam Roberts: everything that’s there.

43 00:05:05.030 00:05:05.549 Uttam Kumaran: Yeah, yeah.

44 00:05:05.550 00:05:07.199 Sam Roberts: Not a ton of like.

45 00:05:07.200 00:05:11.847 Uttam Kumaran: I kinda just wanna put I just wanna put like everything in my brain.

46 00:05:12.180 00:05:12.950 Sam Roberts: That’s actually, probably.

47 00:05:12.950 00:05:16.959 Uttam Kumaran: I’ll kind of probably meet you. I’ll meet you from the bottom, and we’ll meet in the middle somewhere.

48 00:05:16.960 00:05:29.939 Sam Roberts: Yeah, okay, so I’ll I’ll get a doc started and just throw a couple of things that I’ve seen. But you know I’m I’m not too familiar with everything yet, but that yeah, you brain dumping in. There would probably be more helpful than anything for me at this point, because

49 00:05:30.320 00:05:34.149 Sam Roberts: I’m just like I see all the repos. I see, all the super based projects, and I see all the tools and everything.

50 00:05:34.150 00:05:40.460 Uttam Kumaran: Yeah, I’ll give you the I’ll give you like the user stories and the product like the product management overview of like

51 00:05:40.630 00:05:49.469 Uttam Kumaran: every kind of thing that we’re we’ve done. There’s also a lot of stuff that we’ve done and are kind of like sitting or abandoned just because we didn’t go that far into it.

52 00:05:49.470 00:05:50.120 Sam Roberts: Okay.

53 00:05:50.120 00:05:53.519 Uttam Kumaran: Where we’ve done prototypes and stuff. So I just wanna make sure, like all of that work.

54 00:05:54.080 00:06:10.940 Uttam Kumaran: we can pick it back up, or at least we have that reference somewhere. So and then everybody on this team like, I’ll do a 1st pass. I’ll probably get like 60% of it. And then I think the rest of team will be able to be like, Oh, yeah, remember, we did this one day, so we’ll we’ll then have everything documented in one area. So.

55 00:06:11.200 00:06:14.459 Sam Roberts: That’d be that’d be great. Yeah, I’ll throw. I’ll get a notion, Doc, going okay?

56 00:06:15.550 00:06:16.340 Sam Roberts: Better.

57 00:06:17.466 00:06:22.909 Sam Roberts: Yeah. Besides that like, I said, I’ve just been sort of exploring like, I have a meeting

58 00:06:23.040 00:06:30.639 Sam Roberts: later with staff up for like an intro. But I had yesterday with these guys, yeah, just been, you know.

59 00:06:30.850 00:06:33.409 Sam Roberts: digging in, and 1st week is always

60 00:06:34.590 00:06:36.979 Sam Roberts: an uphill kind of climb. But

61 00:06:37.520 00:06:39.639 Sam Roberts: I mean, I’m having fun. So it’s good.

62 00:06:39.640 00:06:41.040 Uttam Kumaran: Cool. Okay, okay.

63 00:06:42.610 00:06:43.510 Uttam Kumaran: Great.

64 00:06:44.960 00:06:48.970 Uttam Kumaran: Cool. I guess. Vash that any updates from your side. What can I help with.

65 00:06:52.510 00:07:00.179 Vashdev Heerani: So I’m getting familiar myself with the Eden project as well as ABC. Platform.

66 00:07:00.958 00:07:05.190 Vashdev Heerani: So kind of getting access from from

67 00:07:05.878 00:07:14.880 Vashdev Heerani: power. Sorry power, not not power. We have a bigquery, and the extra, and the segmentation as well.

68 00:07:15.150 00:07:21.369 Vashdev Heerani: So and also I’m getting S. 3 access and the rail access

69 00:07:22.270 00:07:27.699 Vashdev Heerani: so kind of getting myself familiar with with the platform and the stuff. Yes.

70 00:07:28.510 00:07:29.830 Uttam Kumaran: Okay. Okay. Great.

71 00:07:31.140 00:07:32.000 Uttam Kumaran: Awesome.

72 00:07:33.558 00:07:39.989 Uttam Kumaran: Cool. Okay. Maybe. Mustafa. I know we didn’t get a chance to talk. Maybe let me know. We can walk through client stuff.

73 00:07:43.530 00:07:45.039 Mustafa Raja: Yeah, we can do.

74 00:07:46.277 00:07:48.310 Mustafa Raja: The interlude thing.

75 00:07:48.390 00:07:51.259 Uttam Kumaran: Yeah, so tell me, tell me how it’s going.

76 00:07:52.090 00:07:53.089 Mustafa Raja: Can I share my screen?

77 00:07:53.090 00:07:54.040 Uttam Kumaran: Yeah, please.

78 00:07:54.280 00:07:58.129 Uttam Kumaran: And if you don’t mind giving everybody here like a little bit of a

79 00:07:58.560 00:08:04.680 Uttam Kumaran: 2 second background on the goal when you share and like we’ll accomplish.

80 00:08:05.820 00:08:06.500 Mustafa Raja: Yep.

81 00:08:13.100 00:08:27.849 Mustafa Raja: okay. So so we have. We have a client interlude. They are a design agency. And what they do is that. This they design decks for their clients to pitch to investors.

82 00:08:28.484 00:08:50.939 Mustafa Raja: And they they sort of ha have a manual process around it. They use Claude and the transcripts, or something like that. To get a deck ready for their clients. So we are sort of automating that process for them. And that this is a first.st This is 1st version, sort of a prototype.

83 00:08:51.499 00:08:56.049 Mustafa Raja: For what we are going to give it to give to them.

84 00:08:56.250 00:09:03.059 Mustafa Raja: So yeah, the input, would be transcript and a questionnaire

85 00:09:04.442 00:09:06.380 Mustafa Raja: and we can just save this.

86 00:09:06.530 00:09:08.839 Mustafa Raja: Let the process begin.

87 00:09:15.180 00:09:30.690 Mustafa Raja: So I have a few agents to summarizer, because, the main orchestrator agent does not have context on files because it becomes too much information for it. And it starts to hallucinate. So it only mainly this is only a router agent.

88 00:09:31.541 00:09:36.720 Mustafa Raja: That routes to the relevant agents, and this generates a summary

89 00:09:36.720 00:10:01.919 Mustafa Raja: of according to these files for for this agent to get a gist of it what it’s working with and based on these 2 files, or how many, how, however many files that we would have this agent its job is to do a certain analysis. And get as many metrics as it can

90 00:10:02.010 00:10:06.170 Mustafa Raja: from these documents as we can see here. We have a few metrics.

91 00:10:07.700 00:10:28.380 Mustafa Raja: And we can verify that all of these metrics are actually coming from the files that we are being provided. A sort analysis, and then some notes and everything that this generates the insights. These insights are, get fed into this narrative. Architecture. Agent.

92 00:10:29.128 00:10:39.639 Mustafa Raja: And this the the this agent. What it does is it takes in the files as contacts along with the

93 00:10:43.460 00:10:48.060 Mustafa Raja: insights generated by the insights agent and sort of

94 00:10:48.695 00:10:52.980 Mustafa Raja: get a 1st version of the deck going.

95 00:10:54.030 00:10:56.409 Mustafa Raja: and this this one refines it.

96 00:10:57.689 00:11:09.449 Mustafa Raja: And the Qa. Right now, what? What it only does is it verifies that sort of down the line. We haven’t added a metric that we do not

97 00:11:09.600 00:11:20.519 Mustafa Raja: have in our or sort of the any metric that is, that is not captured by this agent because this captures everything as metric

98 00:11:21.460 00:11:34.429 Mustafa Raja: in it. So we have some sort of verification that we, the agent, did not hallucinate or added, a value. It doesn’t know the source of.

99 00:11:34.730 00:11:42.290 Mustafa Raja: Okay, so this sort of is, the output.

100 00:11:43.180 00:11:44.550 Mustafa Raja: Yeah. So let me know

101 00:11:56.840 00:11:58.300 Mustafa Raja: your mic is off.

102 00:12:00.100 00:12:04.970 Uttam Kumaran: Sorry, I said today we’re gonna get access to some existing outlines.

103 00:12:05.110 00:12:10.119 Uttam Kumaran: It would be great to test our outlines and compare it to the ones that they have.

104 00:12:10.120 00:12:17.249 Mustafa Raja: Yeah, yeah, one more thing. They gave us the transcripts, but

105 00:12:17.900 00:12:20.760 Mustafa Raja: we sort of don’t have access to the raw file.

106 00:12:21.230 00:12:22.700 Mustafa Raja: Let me show what I mean.

107 00:12:28.960 00:12:33.510 Mustafa Raja: Yeah. So so this is sort of the transcript. But this is so.

108 00:12:33.800 00:12:35.210 Mustafa Raja: This is summarized.

109 00:12:35.210 00:12:36.829 Uttam Kumaran: Go to the top, go to the top.

110 00:12:37.540 00:12:41.910 Mustafa Raja: And if you go over here, let me, if we cannot, copy the whole thing.

111 00:12:42.153 00:12:42.640 Uttam Kumaran: Can you.

112 00:12:42.640 00:12:43.699 Mustafa Raja: So what we would do.

113 00:12:43.700 00:12:45.920 Uttam Kumaran: Note in that thread, and ask.

114 00:12:47.710 00:12:48.410 Mustafa Raja: Yeah.

115 00:12:48.410 00:12:52.839 Uttam Kumaran: Can you say, Hey, is there any chance we can get these transcripts.

116 00:12:53.150 00:12:55.650 Uttam Kumaran: or we can get a circle back account

117 00:12:55.860 00:12:57.939 Uttam Kumaran: if we can get the raw files.

118 00:12:58.660 00:13:05.870 Mustafa Raja: Yeah, yeah, yeah, that that would be really helpful. Because, currently our agent our agent is being fed. This

119 00:13:06.210 00:13:06.990 Mustafa Raja: does somebody

120 00:13:07.460 00:13:15.609 Mustafa Raja: the meeting, not the raw transcript, and if and we cannot even copy it properly, if we do, let me show you what happens.

121 00:13:19.810 00:13:21.219 Mustafa Raja: I will assure you.

122 00:13:21.790 00:13:23.960 Mustafa Raja: Nope, hmm.

123 00:13:32.180 00:13:39.019 Mustafa Raja: see, this is sort of what we get if you control a the page. So this isn’t

124 00:13:39.270 00:13:40.270 Mustafa Raja: oh, yeah, we need the.

125 00:13:40.270 00:13:43.499 Uttam Kumaran: Yeah, yeah. So I think they just need to get us an account. I think, with.

126 00:13:43.500 00:13:46.789 Mustafa Raja: Yeah, yeah. Yeah. After the meeting I’ll let them know.

127 00:13:48.050 00:13:51.440 Mustafa Raja: Okay, this was a. This was the demo from me.

128 00:13:51.640 00:13:55.042 Uttam Kumaran: Okay, cool. So one thing also we can do,

129 00:13:56.180 00:14:01.880 Uttam Kumaran: Mustafa is, you can. Even when this when this meeting records.

130 00:14:02.090 00:14:05.450 Uttam Kumaran: you could even probably take this clip, and we can send it as a loom.

131 00:14:06.450 00:14:07.130 Mustafa Raja: Okay.

132 00:14:07.820 00:14:08.420 Uttam Kumaran: Yeah.

133 00:14:09.070 00:14:09.450 Mustafa Raja: Yes.

134 00:14:09.450 00:14:14.290 Uttam Kumaran: Perfect. Okay? And then how is and then let me give you the update on on default.

135 00:14:14.810 00:14:15.460 Mustafa Raja: Yeah.

136 00:14:16.170 00:14:17.900 Uttam Kumaran: So

137 00:14:23.620 00:14:26.072 Uttam Kumaran: So for default.

138 00:14:30.100 00:14:37.130 Uttam Kumaran: I cut this list down to about a hundred 19

139 00:14:37.370 00:14:49.820 Uttam Kumaran: And I’m gonna work today on getting the combinations. But basically I I’m also going. I’m also getting feedback from their sales team. However, I I would be great to probably

140 00:14:50.110 00:14:52.649 Uttam Kumaran: start working on the hardest ones.

141 00:14:52.820 00:14:56.110 Uttam Kumaran: What’s helpful, Mustafa? Should I create tickets

142 00:14:56.540 00:14:58.960 Uttam Kumaran: for each signal that we want to work on.

143 00:15:00.730 00:15:11.419 Mustafa Raja: Yeah, yeah, I understand that we we want to work on some combination. So maybe maybe a ticket on each ticket on each combination would be. I feel good

144 00:15:11.560 00:15:13.097 Mustafa Raja: because I think

145 00:15:15.280 00:15:16.229 Uttam Kumaran: Well, in this way.

146 00:15:16.230 00:15:17.300 Mustafa Raja: This might take.

147 00:15:17.540 00:15:20.549 Uttam Kumaran: Because if you as long as you do it once, I don’t mind.

148 00:15:20.550 00:15:21.210 Mustafa Raja: Yeah, yeah.

149 00:15:21.210 00:15:22.819 Uttam Kumaran: Combination’s easy, right?

150 00:15:25.970 00:15:26.820 Mustafa Raja: Hmm.

151 00:15:26.980 00:15:33.109 Uttam Kumaran: Oh, I so you’ll do. Okay. I see what you mean. Like combination will be the end to end.

152 00:15:33.910 00:15:34.770 Mustafa Raja: Yeah, yeah.

153 00:15:34.940 00:15:38.210 Uttam Kumaran: Okay, okay, so I’ll create some tickets for the combinations.

154 00:15:41.610 00:15:42.220 Mustafa Raja: Yep.

155 00:15:43.703 00:15:45.289 Uttam Kumaran: Okay. Cool.

156 00:15:45.290 00:15:49.020 Mustafa Raja: And other than this, I worked on a ABC ticket.

157 00:15:49.200 00:15:49.700 Uttam Kumaran: Okay.

158 00:15:49.700 00:15:52.510 Mustafa Raja: And would need Amber’s and Casey’s review.

159 00:15:53.510 00:15:54.840 Uttam Kumaran: Okay. Okay. Great.

160 00:15:56.060 00:16:08.729 Uttam Kumaran: Okay. Cool. Yeah. I guess. Probably other question is, for Miguel on either the ambient agents or the new client hubs. If there’s any updates there.

161 00:16:09.867 00:16:15.169 Miguel de Veyra: Yeah, for for the client hub I was actually testing with Casey and

162 00:16:15.907 00:16:31.789 Miguel de Veyra: Mustafa yesterday, and I found some bugs and inconsistencies on. Basically, you know, duplicating it. I’m documenting everything on the doc. Basically, I just remove. For now, I removed all the AI stuff. I’m just filling it up.

163 00:16:31.920 00:16:39.990 Miguel de Veyra: Yeah. Cause remember, we what they call this, we compartmentalize everything. Some of it are not like, not consistent. But also I don’t wanna touch

164 00:16:40.230 00:16:42.050 Miguel de Veyra: some part of it. Because

165 00:16:42.190 00:16:48.550 Miguel de Veyra: if it if that breaks, then everything breaks. So yeah, but everything’s being documented. The ambient agent.

166 00:16:49.188 00:16:58.640 Miguel de Veyra: Yeah, I worked on this like earlier today, I’m it’s basically done. So if you go to AI reminder agent, it’s it’s actually working there.

167 00:16:58.810 00:16:59.400 Uttam Kumaran: Okay.

168 00:16:59.720 00:17:09.690 Miguel de Veyra: What I need to do now, though, is, I think I’ll probably hop on a call with Casey is basically, it’s running for ABC. I just need to figure out basically a way to, you know.

169 00:17:10.240 00:17:14.769 Miguel de Veyra: to run it for everything. And then that should be that that should be it. I can close this out.

170 00:17:15.200 00:17:21.579 Uttam Kumaran: Okay, great. So maybe I’ll I’ll I just joined that channel so I can provide some comments there

171 00:17:21.930 00:17:24.190 Uttam Kumaran: on like the format and or anything.

172 00:17:24.349 00:17:25.519 Miguel de Veyra: Yeah, yeah, yeah, yeah.

173 00:17:25.520 00:17:26.410 Uttam Kumaran: Okay. Great.

174 00:17:26.710 00:17:37.929 Miguel de Veyra: But yeah, if you like, for example, for the last 2, I just sent this out. Now, if you go to like the ABC, you can see that this is the only one yesterday, at least, that didn’t have any replies.

175 00:17:38.080 00:17:38.710 Uttam Kumaran: Okay.

176 00:17:39.870 00:17:40.580 Miguel de Veyra: But yeah.

177 00:17:41.860 00:17:42.880 Uttam Kumaran: Okay, perfect.

178 00:17:42.880 00:17:46.902 Miguel de Veyra: The tea time also. Yeah, I think if

179 00:17:47.700 00:17:57.550 Miguel de Veyra: I talked to Mustafa about it, we have to work on Sunday to basically start from, you know, 12 midnight Eastern up until

180 00:17:58.320 00:18:02.190 Miguel de Veyra: we get somewhere to book, cause I I tried it weekend, and still none.

181 00:18:02.930 00:18:09.458 Uttam Kumaran: Yeah, yeah, my only message is just this, just to email that out to Dan. Like, let’s get as many updates as we can to him.

182 00:18:09.710 00:18:10.739 Miguel de Veyra: Okay. Okay. Sure. Sure.

183 00:18:10.740 00:18:12.690 Uttam Kumaran: Otherwise we’re gonna get a message again.

184 00:18:13.970 00:18:16.630 Miguel de Veyra: Okay. Yeah, I’ll keep him. I’ll send a message.

185 00:18:16.840 00:18:20.768 Uttam Kumaran: Okay? Probably. My only other ask is

186 00:18:21.470 00:18:24.730 Uttam Kumaran: Vashta. There is this ticket that I

187 00:18:25.250 00:18:35.069 Uttam Kumaran: tagged you in, which is like a python ticket that a python pr that Casey did for adding

188 00:18:35.854 00:18:42.207 Uttam Kumaran: more logging and changes for for our zoom

189 00:18:43.270 00:18:46.999 Uttam Kumaran: sort of like pipeline job. I was wondering if you wanna

190 00:18:47.310 00:18:49.609 Uttam Kumaran: take a look and give a review for that.

191 00:18:50.270 00:18:52.680 Uttam Kumaran: I mean, I can. I was gonna do it. But

192 00:18:53.260 00:18:55.230 Uttam Kumaran: since you’re here could be nice.

193 00:18:55.900 00:18:57.749 Vashdev Heerani: Okay. Okay. Okay.

194 00:18:59.811 00:19:05.379 Uttam Kumaran: I just tagged you, and I will send this also here in the zoom chat.

195 00:19:11.100 00:19:13.051 Uttam Kumaran: so yeah, ideally, that’d be great if

196 00:19:13.800 00:19:17.780 Uttam Kumaran: you can take a look at that. And yeah.

197 00:19:20.890 00:19:27.620 Uttam Kumaran: Cool. Okay, what else am I missing?

198 00:19:30.187 00:19:35.910 Uttam Kumaran: Cool. So yeah, I think the I know, Casey, you’re also I saw you’re testing the linear

199 00:19:36.590 00:19:37.610 Uttam Kumaran: ask stuff.

200 00:19:37.900 00:19:41.250 Uttam Kumaran: I guess my question is, gonna be kind of like what you found like,

201 00:19:42.640 00:19:48.159 Uttam Kumaran: are we able to go from linear ask to operating. Should we move off a linear ask like

202 00:19:48.330 00:19:52.430 Uttam Kumaran: the linear ask, thing is nice because it gets put into linear.

203 00:19:53.310 00:19:57.810 Uttam Kumaran: But like we could use another form, or anything that I kind of wanted to get. Your sense.

204 00:19:59.640 00:20:01.249 Casie Aviles: I think this is fine.

205 00:20:02.370 00:20:12.009 Casie Aviles: I’m what I what I do is I I have like a web book that gets triggered whenever a new out of office request comes in, and

206 00:20:12.250 00:20:18.110 Casie Aviles: there’s like an AI step that will extract from the description, since that’s where

207 00:20:18.460 00:20:21.040 Casie Aviles: you know that the days the duration is

208 00:20:21.180 00:20:24.939 Casie Aviles: from, so I think we can go with this, for now.

209 00:20:25.610 00:20:25.975 Uttam Kumaran: Okay.

210 00:20:28.430 00:20:32.309 Uttam Kumaran: And then how is it moving from linear? Ask to operating.

211 00:20:33.389 00:20:39.160 Casie Aviles: That part. I haven’t gotten to that part yet. But yeah, I will work on that next.

212 00:20:39.635 00:20:43.799 Uttam Kumaran: Okay, I feel like, yeah, there’s web hook or something. But

213 00:20:44.650 00:20:50.982 Uttam Kumaran: I mean, we look, we can use a Google form or use another form, if you if you suggest so you can.

214 00:20:51.890 00:20:53.623 Uttam Kumaran: you can let me know

215 00:20:53.970 00:20:54.375 Casie Aviles: Okay.

216 00:20:55.150 00:21:02.939 Uttam Kumaran: But yeah, we spent a lot of time in operating yesterday, so that would be helpful. There is a there is an out of office project there. I think you may have seen that.

217 00:21:03.500 00:21:04.890 Casie Aviles: Yes, I did.

218 00:21:05.120 00:21:09.490 Casie Aviles: I’m planning to yeah. Have the days there, or like, Yeah.

219 00:21:10.420 00:21:11.770 Uttam Kumaran: Okay, okay, perfect.

220 00:21:12.991 00:21:16.908 Uttam Kumaran: Okay? So yeah, I think, otherwise. Yeah, I’m I’m

221 00:21:17.530 00:21:21.910 Uttam Kumaran: I. I think, Sam, if if you let me know when that when you have that doc ready, I’ll just kind of

222 00:21:22.170 00:21:27.159 Uttam Kumaran: brain dump into there. Yeah, if you guys need any help on ABC

223 00:21:27.620 00:21:31.279 Uttam Kumaran: stuff, let me know. Is there anything on ABC that I can help with.

224 00:21:35.825 00:21:44.720 Casie Aviles: We’re just we just fixed a bug that we encountered earlier, but otherwise I think this it’s it’s fine. Now it’s working fine. Now.

225 00:21:45.760 00:21:49.779 Mustafa Raja: Yeah, I feel. On ABC, we are pretty streamlined.

226 00:21:55.600 00:21:57.420 Uttam Kumaran: Okay, cool.

227 00:21:59.950 00:22:05.849 Uttam Kumaran: So yeah, I think I feel like pretty good. Otherwise, like anything else, we wanna discuss.

228 00:22:12.136 00:22:14.213 Mustafa Raja: I may be able to take in

229 00:22:14.700 00:22:19.509 Mustafa Raja: one or 2 internal tickets. So if whenever you have some time.

230 00:22:20.200 00:22:23.280 Mustafa Raja: maybe throw in some in my cycle.

231 00:22:24.220 00:22:28.090 Uttam Kumaran: Yeah. So how did it go with the leads? Work.

232 00:22:30.120 00:22:32.090 Mustafa Raja: Oh, yeah, yeah, it’s in review.

233 00:22:33.550 00:22:34.119 Uttam Kumaran: Yeah, I think, my.

234 00:22:34.120 00:22:37.269 Mustafa Raja: We have. I have the this one in the past.

235 00:22:37.690 00:22:41.590 Uttam Kumaran: Okay, yeah, maybe. Let’s like, let’s take a look now where we’re we’re all on the call.

236 00:22:42.570 00:22:43.200 Mustafa Raja: Yep.

237 00:23:00.640 00:23:25.249 Uttam Kumaran: Nice. Okay? So for context for everybody. Similarly, as we have clients here, we wanted to start having our leads here and our leads are the folks that we’re in sales, conversations with a com like. If I could talk a little bit about like the user stories here, common thing for me is we’re talking to probably anywhere from 10 to 30 potential clients at any time.

238 00:23:25.320 00:23:36.750 Uttam Kumaran: All of them have meetings associated with. They all have like needs for drafting emails, chatting over all the transcripts very, very similar to what we do for clients actually.

239 00:23:38.473 00:23:57.566 Uttam Kumaran: and so what Mustafa worked. However, the source of truth is not our super base is actually gonna be Hubspot. So Mustafa worked on sourcing that data from Hubspot and creating this view. So kind of what what I’m gonna be looking for right now is trying to look for a client that I’m familiar with right now. So I know we’re talking about

240 00:23:58.650 00:24:05.280 Uttam Kumaran: like, let’s do proposal and review for for doobie cars. For example, I should click into here

241 00:24:05.470 00:24:11.619 Uttam Kumaran: and then should see. Okay. Should see the one doobie cars meeting I had.

242 00:24:11.720 00:24:13.910 Uttam Kumaran: If I was to type in here.

243 00:24:14.100 00:24:21.999 Uttam Kumaran: tell me, what do we cars was looking for? I don’t know if the client hub is set up. Yeah, okay.

244 00:24:22.270 00:24:25.569 Uttam Kumaran: so that’s probably gonna be my next

245 00:24:25.970 00:24:31.119 Uttam Kumaran: thing. If maybe we can, I’ll just start to note down some things that we can create related tickets for

246 00:24:35.170 00:24:42.320 Uttam Kumaran: -Oh, pretty new shoes. So Google’s in platform.

247 00:24:42.690 00:24:44.060 Uttam Kumaran: So one is.

248 00:24:44.780 00:24:52.729 Uttam Kumaran: So this is where I think you’re gonna have. We’re gonna have to do something more dynamic for the client hub guys, because we’re gonna have to generate them kind of like

249 00:24:53.570 00:24:56.830 Uttam Kumaran: dynamically, does that make sense.

250 00:25:03.770 00:25:04.710 Sam Roberts: So right now, the

251 00:25:04.920 00:25:09.720 Sam Roberts: clients need to be put in manually. Is that what you’re saying? I think I didn’t quite get all that the other day, but I.

252 00:25:09.720 00:25:14.550 Uttam Kumaran: Yeah, I guess Miguel or Casey or Mustafa. Can you guys talk to like this challenge right now?

253 00:25:22.570 00:25:23.590 Miguel de Veyra: Oh, yeah, this will.

254 00:25:23.590 00:25:24.890 Miguel de Veyra: And yeah, yeah.

255 00:25:24.890 00:25:26.350 Casie Aviles: To manually add this.

256 00:25:27.260 00:25:29.926 Casie Aviles: So yeah, it would be nice to have

257 00:25:31.810 00:25:38.560 Casie Aviles: such that it would be dynamic. So I think we’d we’d need like a way to make it

258 00:25:38.810 00:25:43.360 Casie Aviles: so that the the rag part is automatic, you know.

259 00:25:45.070 00:25:53.826 Casie Aviles: but yeah, otherwise, yesterday we were spending like manual time creating the bots

260 00:25:54.600 00:25:58.109 Casie Aviles: and setting everything up. So I think we should. And then

261 00:25:58.260 00:26:01.749 Casie Aviles: another problem is will be duplicating workflows. So

262 00:26:02.671 00:26:05.959 Casie Aviles: we we can. Yeah, it’s it’s going to

263 00:26:06.750 00:26:13.950 Casie Aviles: like we’re going to reach our limit quite fast. If we add we have a lot of workflows, so

264 00:26:14.280 00:26:16.950 Casie Aviles: perhaps we could consolidate it into one.

265 00:26:17.986 00:26:21.190 Casie Aviles: That’s 1 thing. Maybe we can try to explore.

266 00:26:23.160 00:26:23.690 Miguel de Veyra: Yeah. And then.

267 00:26:23.690 00:26:28.369 Miguel de Veyra: and the other thing is, we have to manually create, like the the slack apps for each.

268 00:26:30.450 00:26:34.310 Uttam Kumaran: Yeah. So basically, Sam, like for every single

269 00:26:34.410 00:26:36.580 Uttam Kumaran: a deal or for a client.

270 00:26:36.720 00:26:51.450 Uttam Kumaran: there are a couple of manual steps in the back end that we have to do like creating a new Nan workflow, creating a new slack app creating a new like super base, dB, potentially

271 00:26:52.280 00:26:55.939 Uttam Kumaran: And so like that stuff probably needs to happen more dynamically.

272 00:26:56.810 00:27:00.499 Uttam Kumaran: Given that we’re gonna be adding more clients and deals and like

273 00:27:00.720 00:27:04.619 Uttam Kumaran: it, just can’t we? It’s it’s not gonna be feasible to

274 00:27:05.170 00:27:16.900 Uttam Kumaran: have to go set out and do that for every single thing. Right? So that’s I think one of the that’s in the a notion, Doc, right now. About like what the current process is.

275 00:27:17.110 00:27:19.879 Uttam Kumaran: And then, ideally, I think I would like to work on.

276 00:27:20.050 00:27:27.429 Uttam Kumaran: What’s something more dynamically, like anytime a new client or a new deal comes in. What are the AI back end pieces that need to get spun up?

277 00:27:27.430 00:27:27.665 Uttam Kumaran: Yeah.

278 00:27:27.900 00:27:35.869 Uttam Kumaran: support that right, for example, ideally like, if I was to go back to another deal with a couple of other meetings. Let’s let’s look at the

279 00:27:37.000 00:27:41.620 Uttam Kumaran: let’s look at the Fan stake one, for example.

280 00:27:43.690 00:27:47.719 Uttam Kumaran: we should see a couple of fancy meetings. Yes, so ideally.

281 00:27:47.900 00:27:51.060 Uttam Kumaran: IA salesperson here, would want to chat

282 00:27:51.600 00:27:56.699 Uttam Kumaran: with. Tell me like what our fans takes highest priorities. The current system we have.

283 00:27:57.040 00:27:59.660 Uttam Kumaran: we would want to basically run a chat

284 00:27:59.900 00:28:03.320 Uttam Kumaran: or a basically a rag chat over all these transcripts

285 00:28:04.007 00:28:19.559 Uttam Kumaran: right? Because these are all the meetings associated with this deal. And so on the sales side, it would be like, Generate me a follow up email based on the last meeting. Generate me like a concise summary for our internal notes on like what this client wants.

286 00:28:19.680 00:28:30.690 Uttam Kumaran: Generate me a note that I can send to our project management team about what to expect if they convert to a client like what the scope of the project is on right. Those are all the things that this this chat bot should support.

287 00:28:32.205 00:28:40.119 Uttam Kumaran: But I would say as part of this ticket that we’re reviewing, I’m actually pretty happy because we’re able to see all the meetings associated with the deal in one place

288 00:28:40.320 00:28:44.620 Uttam Kumaran: previously I would have to go into our dashboard, which is a catch all for all meetings

289 00:28:45.380 00:28:53.719 Uttam Kumaran: and go and find the dance. Take one which maybe it’s

290 00:28:55.410 00:29:00.429 Uttam Kumaran: oh, so right now, if it’s not on this 1st page. The search doesn’t work right? Or

291 00:29:01.380 00:29:02.529 Uttam Kumaran: is that correct?

292 00:29:03.730 00:29:04.469 Mustafa Raja: Yeah, yeah.

293 00:29:06.450 00:29:08.510 Sam Roberts: Oh, yeah. Okay.

294 00:29:11.940 00:29:16.860 Uttam Kumaran: Yeah. So that’s that’s probably another thing. So just to make sure the search works across

295 00:29:17.810 00:29:25.669 Uttam Kumaran: search is working across pages. I mean one thing that I requested was just like should probably just do infinite scroll.

296 00:29:25.800 00:29:28.270 Uttam Kumaran: but like load the 1st 50. So

297 00:29:31.230 00:29:34.590 Uttam Kumaran: I thought that was somewhere in a Pr. Miguel? Or did we ever get to that.

298 00:29:35.658 00:29:42.040 Miguel de Veyra: We decided 50 on the 1st load. Yes, and then just basically a hundred per page.

299 00:29:42.730 00:29:44.030 Uttam Kumaran: Yeah, I just feel like.

300 00:29:44.950 00:29:46.390 Miguel de Veyra: The infinite scroll, deck.

301 00:29:46.670 00:29:51.170 Uttam Kumaran: Yeah. So that’s probably one thing. Otherwise I feel okay. I think the only other thing

302 00:29:53.510 00:29:58.610 Uttam Kumaran: Mustafa, I would suggest, is like, I don’t know what these these stages are, but it’s coming from Hubspot, right? So it’s probably

303 00:29:59.120 00:29:59.670 Uttam Kumaran: awesome.

304 00:30:00.410 00:30:01.930 Mustafa Raja: Yeah, so we have.

305 00:30:02.060 00:30:13.110 Mustafa Raja: I think we have multiple campaigns. And the the ones that that are named are are. The one is, the is, is from the campaign that is active.

306 00:30:14.140 00:30:14.700 Uttam Kumaran: Okay.

307 00:30:14.700 00:30:19.709 Mustafa Raja: And the unknown ones are not active. But I’ll include them too.

308 00:30:21.540 00:30:22.280 Uttam Kumaran: Okay.

309 00:30:22.390 00:30:23.030 Mustafa Raja: There you go.

310 00:30:23.030 00:30:27.629 Mustafa Raja: It’s going to be a backfill in Hubspot. Oh, a backfill in super base Hubspot!

311 00:30:27.880 00:30:32.220 Uttam Kumaran: Okay. Okay. So another thing is, I’m gonna just say, like, we’ll default to all statuses.

312 00:30:33.800 00:30:38.639 Uttam Kumaran: the nice thing is here. You can click in and go directly to the Hubspot related record, which is great.

313 00:30:40.930 00:30:46.830 Uttam Kumaran: I think probably the other thing I will add here is like, bring in the deal owner.

314 00:30:47.500 00:30:49.760 Uttam Kumaran: It’s gonna be me or Robert. But

315 00:30:49.910 00:30:52.320 Uttam Kumaran: that way I could just filter to my deals.

316 00:30:52.670 00:30:53.690 Mustafa Raja: Okay.

317 00:30:54.820 00:30:56.980 Uttam Kumaran: But otherwise like this is pretty good.

318 00:30:57.210 00:30:59.360 Uttam Kumaran: Sam, you kind of get the like.

319 00:31:00.450 00:31:04.650 Uttam Kumaran: The theory behind this sort of like part of the platform.

320 00:31:04.950 00:31:11.850 Sam Roberts: Yeah, I think I’m I mean as much as I think I am. Yeah. Like pulling it in from Hubspot, getting all that information, finding those meetings.

321 00:31:11.850 00:31:23.419 Uttam Kumaran: They see, like everything around streamlining, because going into Hubspot and getting information is such a pain. But, in fact, like, I don’t really need any information from Hubspot. I need the meetings associated with

322 00:31:23.800 00:31:31.959 Uttam Kumaran: deals to be able to chat over. So how quickly can I chat over all the transcripts associated with a deal is the limiting factor, because

323 00:31:32.290 00:31:34.610 Uttam Kumaran: well, I need to generate sows.

324 00:31:35.430 00:31:42.970 Uttam Kumaran: emails, whatever. And so commonly right now we copy paste into our Chat Gbt, locally to do this stuff.

325 00:31:43.575 00:31:52.050 Uttam Kumaran: And that’s a complete pain, because I have to go find the meetings one by one, copy them in and then chat over it versus like

326 00:31:52.410 00:32:01.089 Uttam Kumaran: something pretty easily enabled that we can enable here, additionally, we have other folks on the sales team that are going to start generating those follow ups

327 00:32:01.280 00:32:21.650 Uttam Kumaran: for me and Robert like in coordination. And so this gives them the opportunity to do that. You know, or to ask like, Hey, what should I update or what status update? The next piece will probably also work on, is being able to. Once a meeting happens related to a deal automatically updating properties of the hubs, related. Hubspot

328 00:32:21.970 00:32:43.079 Uttam Kumaran: deal dynamically, like AI will actually suggest, hey? Looks like this is moving to the next phase. I should go up, I should go. Use the Hubspot tool or Hubspot to go, make a change to the deal, or add a note right? So that’s like the level 2 level 3 where the AI starts to handle some of the baseline coordination. Because

329 00:32:43.300 00:32:53.590 Uttam Kumaran: coordinating. If you watch the transcript, you’ll understand what the status is, and what the next steps are all in there. So the next step is for AI to start making those actions themselves.

330 00:32:54.980 00:32:57.009 Sam Roberts: That makes sense. Okay, yeah, no, I’m seeing it.

331 00:32:57.400 00:33:03.250 Uttam Kumaran: Okay, cool. So Mustafa, I think. Probably I have like a couple of like

332 00:33:03.510 00:33:08.889 Uttam Kumaran: really easy ones for this. And then a couple of more difficult ones for deals and platforms. So I’ll assign that to you.

333 00:33:09.000 00:33:13.317 Uttam Kumaran: The other thing that’s that would be helpful is is, would would be

334 00:33:14.270 00:33:14.880 Mustafa Raja: Department.

335 00:33:15.310 00:33:16.550 Uttam Kumaran: Departments.

336 00:33:20.530 00:33:24.606 Uttam Kumaran: I would I would either say departments or the co-pilot kit Demo like

337 00:33:25.170 00:33:31.640 Uttam Kumaran: I don’t know. I was. Gonna ask Miguel like, if you think you’re gonna get to that this week, or whether we should move that off.

338 00:33:34.180 00:33:37.059 Miguel de Veyra: The Copa. Wait. What day is it?

339 00:33:37.530 00:33:38.520 Miguel de Veyra: Wednesday?

340 00:33:38.900 00:33:46.770 Miguel de Veyra: Yeah, I should be able to cause. Yeah, yeah, I should be able to. I’ll I’ll just finish this ambient agent for the minor agent, and then, yeah, I’ll jump to that.

341 00:33:48.250 00:33:51.100 Uttam Kumaran: So then, yeah, I think departments would probably be the best.

342 00:33:54.210 00:34:01.190 Uttam Kumaran: The best thing to work on next, that’ll basically that’ll basically round out.

343 00:34:02.144 00:34:06.700 Uttam Kumaran: Like, that’s the that’s the extent of our meetings. Right

344 00:34:06.950 00:34:10.669 Uttam Kumaran: sales all of the core departments.

345 00:34:11.330 00:34:11.920 Mustafa Raja: Yeah.

346 00:34:14.310 00:34:15.070 Mustafa Raja: Yep.

347 00:34:16.100 00:34:23.870 Uttam Kumaran: Okay, cool and then I think also, Sam, like as you’re poking around this feel free to create like one ticket with all of your notes like.

348 00:34:24.300 00:34:30.059 Uttam Kumaran: you know, on anything easy, low, hanging fruit stuff that we want to clean up as you’re going through front end. So.

349 00:34:30.469 00:34:31.089 Sam Roberts: Totally.

350 00:34:33.840 00:34:34.719 Uttam Kumaran: Okay.

351 00:34:35.270 00:34:40.599 Uttam Kumaran: cool. Alright. And then, yeah, Casey, I think, let me know if the out of office thing is is

352 00:34:40.780 00:34:43.419 Uttam Kumaran: gonna keep going. I would honestly consider

353 00:34:43.880 00:34:47.490 Uttam Kumaran: seeing how much of a dent you can make on the transcript stuff, too, this week.

354 00:34:49.120 00:34:51.470 Uttam Kumaran: The out of office stuff is important. But I think

355 00:34:51.610 00:35:06.260 Uttam Kumaran: what if you get a sense for what needs to happen? We can ticket that we can ticket it out, and and even Vasha or another person can take it. But I think I want I want to see. I think giving taking a look at the spike for a transcript may be a nice break for you, because

356 00:35:06.630 00:35:08.650 Uttam Kumaran: there’s some interesting stuff we can do there.

357 00:35:09.460 00:35:10.709 Casie Aviles: Okay. Sure. Sure.

358 00:35:11.490 00:35:12.050 Uttam Kumaran: Okay.

359 00:35:13.150 00:35:17.799 Uttam Kumaran: Okay, perfect. Alright. Thanks. Everyone. Appreciate it. Slack me if you need anything.

360 00:35:18.250 00:35:19.170 Miguel de Veyra: Thanks. Everyone.

361 00:35:19.170 00:35:20.379 Mustafa Raja: Thank you. Bye-bye.

362 00:35:20.550 00:35:21.020 Sam Roberts: And.