Meeting Title: AI Team Standup Date: 2025-05-27 Meeting participants: Mustafa Raja, Uttam Kumaran, Miguel De Veyra, Casie Aviles, Awaish Kumar


WEBVTT

1 00:01:13.540 00:01:14.310 Mustafa Raja: Hey!

2 00:01:19.630 00:01:20.370 Miguel de Veyra: Hey? Mate?

3 00:01:20.520 00:01:21.360 Miguel de Veyra: Morning.

4 00:01:22.210 00:01:23.780 Mustafa Raja: Yeah. Good morning.

5 00:03:02.030 00:03:06.140 Miguel de Veyra: Hey, Mustafa! What’s the name of the consolidation thing that you made.

6 00:03:08.144 00:03:10.360 Mustafa Raja: Can you ask again.

7 00:03:10.360 00:03:12.819 Miguel de Veyra: What’s the name of the consolidation workflow?

8 00:03:13.605 00:03:14.250 Mustafa Raja: Yeah.

9 00:03:14.987 00:03:17.429 Miguel de Veyra: Let me check. I don’t remember.

10 00:03:17.430 00:03:19.190 Casie Aviles: Okay, sure, reusable, component.

11 00:03:19.190 00:03:22.280 Mustafa Raja: Yeah, yeah, yeah, it’s in the reusable components. Folder.

12 00:03:23.000 00:03:23.820 Miguel de Veyra: Okay.

13 00:03:25.720 00:03:28.800 Mustafa Raja: Inside internal AI internal. AI, right? Okay.

14 00:03:30.870 00:03:34.609 Mustafa Raja: yeah. And then it will be slack bot, even processor.

15 00:03:36.240 00:03:37.910 Miguel de Veyra: Okay, there you go. Nice side.

16 00:05:02.170 00:05:07.540 Miguel de Veyra: Oh, okay. I still might need to use aws for this.

17 00:05:10.520 00:05:11.720 Casie Aviles: For the images.

18 00:05:11.720 00:05:15.080 Miguel de Veyra: Yeah, cause slack. It’s uploaded to slack. Right?

19 00:05:16.353 00:05:20.019 Miguel de Veyra: You’re using that as URL. I don’t have that right now.

20 00:05:21.220 00:05:22.689 Casie Aviles: So it has to be uploaded first.st

21 00:05:23.390 00:05:23.980 Casie Aviles: Yes.

22 00:05:23.980 00:05:25.110 Miguel de Veyra: Yeah.

23 00:05:29.100 00:05:29.840 Uttam Kumaran: Hey, guys.

24 00:05:32.060 00:05:34.950 Miguel de Veyra: Or I don’t know. I’ll find a workaround, hey, Utah!

25 00:05:35.930 00:05:36.640 Mustafa Raja: Hey!

26 00:05:37.022 00:05:38.169 Uttam Kumaran: What’s the discussion?

27 00:05:39.556 00:05:44.599 Miguel de Veyra: Yeah, basically, we’re just trying to find a way.

28 00:05:44.780 00:05:49.109 Miguel de Veyra: because for cool parts, the image upload. How it will go from, you know.

29 00:05:49.460 00:05:53.699 Miguel de Veyra: from react into node because it’s working on slack. So I was hoping

30 00:05:53.950 00:06:00.170 Miguel de Veyra: I could cut some time there. But yeah, slack has, you know, if you upload an image on slack. It stays there

31 00:06:00.900 00:06:08.089 Miguel de Veyra: right? There’s a public URL. But if if it’s from react, there’s nothing like that, I guess I can upload. Yeah, I’ll find a way.

32 00:06:09.600 00:06:14.950 Uttam Kumaran: Yeah, I mean, is the image, Api, not just like plug and play? Or is it like a little bit more complicated.

33 00:06:15.300 00:06:17.230 Miguel de Veyra: It’s a bit more complicated.

34 00:06:18.140 00:06:19.180 Uttam Kumaran: Okay.

35 00:06:19.180 00:06:22.239 Miguel de Veyra: I need, because the the model needs the

36 00:06:22.360 00:06:26.179 Miguel de Veyra: the model needs the image to be uploaded somewhere.

37 00:06:26.470 00:06:37.119 Miguel de Veyra: So I’m thinking, instead of having to go through all that hell in aws, and we I’m not sure we want to save it. It’s just gonna be saved locally, temporarily, on

38 00:06:37.450 00:06:42.120 Miguel de Veyra: on Heroku, and then we’ll just, you know, we’ll just delete it after a certain amount of time.

39 00:06:43.220 00:06:43.940 Miguel de Veyra: I think that.

40 00:06:43.940 00:06:50.289 Uttam Kumaran: I feel like, yeah, I mean, but did you like there’s there’s a lot of apps that are doing this like, did you have a look at like

41 00:06:50.600 00:06:57.340 Uttam Kumaran: simple image, like AI. Recognition apps, and just like, look at some of the open source ones that exist.

42 00:07:00.200 00:07:01.019 Miguel de Veyra: Because there’s no.

43 00:07:01.020 00:07:16.180 Uttam Kumaran: There’s no need to like completely reinvent this. It’s like, it’s like such a. It’s a pretty basic feature. I feel like there’s either going to be a like an Api that wraps open AI and supports this. Or there’s gonna be some like boilerplate code. We can just copy.

44 00:07:17.030 00:07:18.170 Miguel de Veyra: Yeah, yeah, and like.

45 00:07:18.170 00:07:21.100 Uttam Kumaran: I don’t want to do a ton of back end work for this right now.

46 00:07:21.789 00:07:27.720 Miguel de Veyra: Yeah, I mean the the AI will process the image. It’s just getting that image into the AI.

47 00:07:28.520 00:07:33.059 Uttam Kumaran: No, no, that’s that’s what I mean. Like I I get that. You need a back end server.

48 00:07:33.060 00:07:33.470 Miguel de Veyra: I know.

49 00:07:34.290 00:07:34.940 Uttam Kumaran: Blood.

50 00:07:35.770 00:07:38.359 Uttam Kumaran: Well, you need to store the image somewhere right.

51 00:07:38.360 00:07:39.100 Miguel de Veyra: Yes, yes.

52 00:07:39.660 00:07:44.209 Uttam Kumaran: So okay, that’s what I guess that’s what I mean. But like, I guess what I’m asking is like.

53 00:07:45.120 00:07:51.390 Uttam Kumaran: if you just like, I I guess all my questions on most things are going to be like, did you Google? It first.st

54 00:07:51.795 00:07:55.850 Miguel de Veyra: Yes, yes, I’ve been working on basically Openai with this.

55 00:07:56.330 00:07:58.250 Uttam Kumaran: No, no, no, but like less about

56 00:07:58.400 00:08:11.060 Uttam Kumaran: asking AI how to do it like. There will be some people that have done this and already uploaded the code to Github to do this like, if you Google like, maybe you can try right now, if you Google, like simple

57 00:08:11.230 00:08:19.139 Uttam Kumaran: image, open AI recognizer like, there’s a lot of very basic

58 00:08:19.290 00:08:22.509 Uttam Kumaran: notebooks that already do this, that I played with.

59 00:08:23.050 00:08:24.860 Miguel de Veyra: Okay, yeah, sure. I’ll look into it.

60 00:08:26.930 00:08:30.539 Uttam Kumaran: I mean do like this is what I guess. What I’m trying to say is like

61 00:08:30.700 00:08:44.640 Uttam Kumaran: anytime where we try to do something, and it’s like we can’t figure it out. I really want us to to just try to see who’s done it in public already, because these all these things we’re trying. I’ve seen in many forms. It’s that’s the way.

62 00:08:44.900 00:08:50.629 Uttam Kumaran: you know. We sort of even have the idea to work on it. So I want us to find those open source

63 00:08:50.770 00:08:56.950 Uttam Kumaran: like for Nan, for example, it’s like, you can just Google, how to do something in na, then. And that way, we start with

64 00:08:57.400 00:09:04.330 Uttam Kumaran: their workflow. We can just copy that and start from there right versus like having to come to the same conclusions they did.

65 00:09:06.500 00:09:08.129 Uttam Kumaran: So that’s what I’m hoping like.

66 00:09:08.380 00:09:16.110 Uttam Kumaran: But even for this, like I don’t I. This is where I like. I want to know at the end of the day, like, if we’re able to finish work because

67 00:09:16.210 00:09:20.310 Uttam Kumaran: I we can’t spend 5 days on image recognition, right? And so

68 00:09:20.760 00:09:29.359 Uttam Kumaran: if it’s taking more than 2 days, I told Amber stop working on it. And so it’s clear to me that it’s taking more than 4 or 5 days.

69 00:09:29.670 00:09:35.349 Uttam Kumaran: This is where I’m getting a little bit frustrated because nobody’s giving me insight into like how long things are taking

70 00:09:36.326 00:09:46.089 Uttam Kumaran: and like, we’re just working on things and sort of like not considering how long things are going to take, or we’re just working on it for weeks and weeks and weeks like it can’t happen this way.

71 00:09:49.790 00:09:54.949 Uttam Kumaran: Like we, we agree that the image. Recognition was not going to be part of phase one.

72 00:09:55.540 00:09:59.969 Uttam Kumaran: But then, now we’re working on it. Right? So like did Amber make that call, or like

73 00:10:00.170 00:10:02.949 Uttam Kumaran: who decided to work on the image recognition stuff.

74 00:10:05.070 00:10:09.479 Miguel de Veyra: Amber told me last week to basically, you know, if we have time, let’s work on that.

75 00:10:10.090 00:10:12.360 Uttam Kumaran: But like we spend now, like how many days on it.

76 00:10:14.216 00:10:17.299 Miguel de Veyra: This is probably the second day that I worked on it fully.

77 00:10:17.810 00:10:21.260 Uttam Kumaran: Okay, I’m just like concerned that we’re like

78 00:10:21.710 00:10:28.410 Uttam Kumaran: spending like we’re like, Go work on it. And then it’s like we go down like a week long rabbit hole, and then we don’t get anywhere right. So

79 00:10:28.690 00:10:35.520 Uttam Kumaran: I mean, I don’t know. I if if it’s if we don’t like, I don’t want to work on things until we have a clear understanding of how long it’s gonna take.

80 00:10:35.770 00:10:38.820 Uttam Kumaran: Which means we need to do the research upfront like

81 00:10:39.140 00:10:48.389 Uttam Kumaran: we need like, I I really need to start to know how long things are really gonna take to get delivered because we’re we’re constantly struggling with that right? So

82 00:10:49.760 00:10:54.600 Uttam Kumaran: like if you the the only point of reference I have. If you pull up that notion, Doc.

83 00:10:54.760 00:10:59.540 Uttam Kumaran: for the AI proof of concept for pool parts.

84 00:10:59.640 00:11:04.020 Uttam Kumaran: we said that there’s a chance that we don’t. This doesn’t make it in.

85 00:11:04.120 00:11:06.620 Uttam Kumaran: And so, like.

86 00:11:07.000 00:11:12.659 Uttam Kumaran: I’m I just want to know, like, does Amber know, that this is taking longer, and this may not happen like

87 00:11:12.770 00:11:16.739 Uttam Kumaran: I can’t be hearing about this right now and then. Sort of make a call at the end, like.

88 00:11:17.080 00:11:18.020 Uttam Kumaran: you know.

89 00:11:19.060 00:11:22.350 Miguel de Veyra: Yep, yep, okay. I’ll speak with Amber.

90 00:11:23.370 00:11:30.079 Uttam Kumaran: Okay. And then the other thing is like, I guys, I really need to get an update like, sort of end of day with like how things are going.

91 00:11:30.260 00:11:59.200 Uttam Kumaran: This is part of the, you know. Part of the reason that you know. I think the team was struggling before is that I just need to know, like, how long things are. Gonna take. I think Mustafa has been doing a really good job of that like sending looms at the end of the day. I I think, Casey, I think we usually just chat here and there, but I think for Casey and Miguel just like a like a few minute loom on just like what happened today and where you’re stuck.

92 00:11:59.560 00:12:00.119 Casie Aviles: 12, along.

93 00:12:00.120 00:12:03.180 Uttam Kumaran: Because, as you, as you guys know, like we’re gonna

94 00:12:03.400 00:12:05.280 Uttam Kumaran: we’re, gonna I’m gonna be able to

95 00:12:05.570 00:12:18.909 Uttam Kumaran: find the opportunity to help or to like, say, like, Okay, let’s scratch that like, we just need those iteration cycles like, I’m not I. I only have this 30 min block to talk to you guys. So I need that context.

96 00:12:29.400 00:12:34.522 Casie Aviles: Yeah, regarding that, I just want to ask about zoom,

97 00:12:35.360 00:12:40.539 Casie Aviles: I think I just have a video limit at the moment.

98 00:12:41.580 00:12:44.689 Uttam Kumaran: Okay, you can go in and and press. Just request upgrade.

99 00:12:46.290 00:12:50.850 Casie Aviles: Okay, yeah. I think I was upgraded before, but I got downgraded.

100 00:12:52.350 00:12:56.249 Uttam Kumaran: Yeah, if you just go in and hit request upgrade, you’ll I’ll give you the upgrade

101 00:12:57.810 00:13:05.840 Uttam Kumaran: But like we only meet for 30 min, and I don’t want to put another meeting at the end of the day to do this. But I’m not getting the context. I need

102 00:13:06.020 00:13:08.490 Uttam Kumaran: to like make decisions as like the Pm.

103 00:13:09.440 00:13:17.640 Uttam Kumaran: Cause I need to know how long things are gonna take. And every day I’m coming to the meeting with like, Oh, it’s gonna take longer. It’s gonna take longer, like it’s not. That’s not

104 00:13:18.150 00:13:20.070 Uttam Kumaran: we. We can’t operate that way.

105 00:13:21.970 00:13:28.049 Uttam Kumaran: so I just need a lot more clarity right for the image upload. This project we planned 2 weeks ago.

106 00:13:28.230 00:13:33.150 Uttam Kumaran: right? And so I just like we can’t work on things where the scope is unlimited.

107 00:13:33.300 00:13:44.899 Uttam Kumaran: Right? So at this point, like, I’m gonna say, stop working on the image upload until Amber makes a decision right? And I’m gonna give the feedback to her as well that we can’t work on things that are open ended

108 00:13:45.010 00:13:50.050 Uttam Kumaran: for for a feature like the image upload. We need to do the research upfront

109 00:13:50.440 00:14:03.659 Uttam Kumaran: like, I need to know what options there are like, did we Google, did we talk to Openai like, did we? Did we look at what opportunities there are to speed this up, and then what is a reasonable timeline for delivery?

110 00:14:03.780 00:14:07.270 Uttam Kumaran: We can’t start something and then figure all that out. It’s not possible.

111 00:14:08.120 00:14:13.240 Uttam Kumaran: Right? So like I, there’s things that we know how long it’s gonna take.

112 00:14:13.400 00:14:16.059 Uttam Kumaran: And so it’s easy for us to say that’s only going to be 2 days.

113 00:14:16.370 00:14:25.439 Uttam Kumaran: But I’m not looking for just like tell me what day I’m looking for the answer right like, how confident are we that this will actually get done right now.

114 00:14:25.780 00:14:35.579 Uttam Kumaran: What I’m hearing Miguel, is that we’re not going to get this done like it’s not clear what the path forward is. I heard we need heroku. We need a, we need image upload. We need all these things.

115 00:14:35.700 00:14:45.039 Uttam Kumaran: That’s that’s too big of a scope. Right? And so if if you were to say, Okay, cool, that’s all we need. Let me go work on it for 4 days, and we don’t get paid.

116 00:14:45.250 00:14:51.699 Uttam Kumaran: Then it’s like completely useless. So all that work needs to happen upfront. And then you need to come to

117 00:14:52.150 00:14:58.799 Uttam Kumaran: your Pm. Which is amber and tell her, hey? I did some research. This ticket is gonna take 4 days of work.

118 00:14:59.310 00:15:00.539 Uttam Kumaran: Should I work on it

119 00:15:00.830 00:15:05.080 Uttam Kumaran: like that’s a decision you have to put back on your project manager. She has the responsibility

120 00:15:05.680 00:15:09.960 Uttam Kumaran: right? Like you have to give her the ability to make that call.

121 00:15:12.830 00:15:13.910 Miguel de Veyra: Okay. Noted.

122 00:15:15.390 00:15:33.709 Uttam Kumaran: That’s the that’s the distinction, right like. If if you have it all in your head and you’re executing and making decisions, then you’re not giving her any grace to go like she has to communicate to the client she has to communicate to me right, and so put that decision on her. But what you owe her is you owe her

123 00:15:33.840 00:15:38.139 Uttam Kumaran: like, hey? These are all I did like 1 h or 2 h of research.

124 00:15:38.330 00:15:55.259 Uttam Kumaran: It’s gonna take 4 or 5 days, and your goal is to estimate accurately, not to say I’ll get in 2 days, and then it’s taking 5 days. That’s what’s happening right. And so this is where, like, I, I really want you guys to hold yourself accountable to

125 00:15:55.750 00:15:58.490 Uttam Kumaran: really spending time and doing the research like, I think

126 00:15:58.660 00:16:04.810 Uttam Kumaran: one of the positive things that I saw yesterday from Mustafa is before he started working on the

127 00:16:05.680 00:16:07.939 Uttam Kumaran: you know, some of the sales and 8 N. Work.

128 00:16:08.070 00:16:12.209 Uttam Kumaran: I got a call and was like, Hey, can I just walk you through this architecture before I go for it?

129 00:16:12.420 00:16:14.550 Uttam Kumaran: That’s the sort of stuff I need to hear.

130 00:16:14.870 00:16:25.040 Uttam Kumaran: you know, before, before I can help and make a call. If we’re only talking these 30 min per day, and I’m not getting enough information. Then I just have to. I just have to make the meetings longer.

131 00:16:25.370 00:16:28.929 Uttam Kumaran: and we have to talk through every single ticket. I don’t want to do that. So

132 00:16:30.890 00:16:43.369 Uttam Kumaran: does that make sense for everyone. We’re small team. It’s only us. So like we owe it to each other to help each other out and and get some of this stuff out, so I don’t want us to make the same mistakes that we did, you know, a month or 2 ago.

133 00:16:53.490 00:16:54.439 Uttam Kumaran: How do we feel.

134 00:16:54.440 00:16:56.159 Mustafa Raja: Yeah, this feels good.

135 00:16:57.110 00:16:57.800 Uttam Kumaran: Okay.

136 00:17:01.360 00:17:03.235 Uttam Kumaran: okay, cool. Maybe.

137 00:17:04.380 00:17:09.769 Uttam Kumaran: Mustafa, do you want to? Maybe run, stand up, and maybe we just go through open tickets.

138 00:17:11.109 00:17:12.759 Mustafa Raja: Yeah, I can do that.

139 00:17:13.310 00:17:13.849 Uttam Kumaran: Okay.

140 00:17:14.256 00:17:15.390 Mustafa Raja: Let me share my screen.

141 00:17:19.069 00:17:19.960 Mustafa Raja: Sure.

142 00:17:21.220 00:17:22.040 Mustafa Raja: Okay.

143 00:17:34.653 00:17:40.649 Mustafa Raja: yeah. So let’s continue. So

144 00:17:41.142 00:17:50.829 Mustafa Raja: I guess, start we’re starting from the rightmost. So this is the ticket. And I put in some of some comparisons of the cost.

145 00:17:51.480 00:17:53.360 Mustafa Raja: I hope you saw it.

146 00:17:54.160 00:18:16.459 Mustafa Raja: Yes, yeah. So so this one, this one. This one is a lot cheaper than this one. These 2 serve the same purpose. But this one would give us data, that which post is getting how much of reactions or how much engaging is this post? So we can rank

147 00:18:16.900 00:18:32.309 Mustafa Raja: the posts. And then we can put limit that. Okay, we need only top 10 of these posts. And we would get the users that are interacting with them. With this, we do not have this data. So we would have to

148 00:18:33.640 00:18:39.639 Mustafa Raja: fetch every user without any kind of sorting that we did here. So.

149 00:18:39.640 00:18:47.510 Uttam Kumaran: I think the the I think the Linkedin one is probably the best for now, because I think we will want to reference those Urls.

150 00:18:48.105 00:18:54.110 Uttam Kumaran: We can. We can let we can let Robert make the call on like if he wants less or more information.

151 00:18:54.310 00:18:57.320 Uttam Kumaran: I think your solution now makes sense.

152 00:18:58.660 00:19:12.260 Uttam Kumaran: yeah, I feel, I feel okay for context for Miguel and Casey. This, we’re we’re searching. We we build like a little thing that search. You search for an event, and it scrapes the top Linkedin posts. That reference those events

153 00:19:12.550 00:19:20.400 Uttam Kumaran: so that we can then retarget those people find out that they’re interested in like a certain event, we can go to that event and message them to meet up.

154 00:19:20.760 00:19:23.169 Uttam Kumaran: So it’s helping the sales team do that.

155 00:19:26.080 00:19:26.690 Mustafa Raja: Okay.

156 00:19:26.690 00:19:27.210 Uttam Kumaran: Okay.

157 00:19:27.620 00:19:34.700 Mustafa Raja: Yeah. But this one is a lot more expensive than this one, though. Okay. So coming back.

158 00:19:37.260 00:19:40.439 Mustafa Raja: Okay, these are mine. Actually sorry.

159 00:19:40.740 00:19:43.489 Mustafa Raja: Yeah, this is better. Yeah.

160 00:19:46.220 00:19:50.389 Mustafa Raja: So we have this from Megan.

161 00:19:53.188 00:19:54.430 Miguel de Veyra: Sorry! Can you scroll up.

162 00:19:54.650 00:19:55.190 Mustafa Raja: Yeah.

163 00:19:56.930 00:19:58.130 Miguel de Veyra: Yeah, I think this is done.

164 00:19:58.610 00:20:00.549 Mustafa Raja: Okay, this is done. So.

165 00:20:00.550 00:20:02.590 Miguel de Veyra: The fixes was the last part of this.

166 00:20:02.730 00:20:03.580 Mustafa Raja: Okay. So I’m moving.

167 00:20:03.580 00:20:06.989 Uttam Kumaran: Yeah, the voice, the voice agent is fine for now to not worry about that.

168 00:20:06.990 00:20:09.779 Miguel de Veyra: Yeah, yeah, we’ll close this, yeah.

169 00:20:09.780 00:20:11.270 Uttam Kumaran: What was the fix for this.

170 00:20:12.809 00:20:20.960 Miguel de Veyra: I added, I basically removed because before the height I had was a hundred percent on everything, because we don’t have to scroll down.

171 00:20:21.290 00:20:26.750 Miguel de Veyra: It seemed like it was affecting. the actual website.

172 00:20:26.940 00:20:34.480 Miguel de Veyra: not just the sandbox one. So I had to add parameters on, basically only on this chat, box, not anywhere else.

173 00:20:35.200 00:20:35.830 Uttam Kumaran: Okay.

174 00:20:36.470 00:20:42.560 Miguel de Veyra: And then the other thing, I added the fix here, which I didn’t, I should have added here, but

175 00:20:43.750 00:20:47.409 Miguel de Veyra: the the record analysis, and

176 00:20:47.840 00:21:04.519 Miguel de Veyra: the one we did for the standardization it for cause. I duplicated that workflow and then just changed up the details. I just realized, yeah, the production Urls were not the same. That’s why it was hit or miss. Sometimes it’s working, sometimes it’s not, but yeah, that’s not fixed.

177 00:21:05.590 00:21:19.579 Uttam Kumaran: Okay, yeah, I I clicked around. I saw that yesterday, too. I think maybe just like, if you have a few minutes today for everyone on the team. You could just like, go to that demo site and just click around and and send any feedback

178 00:21:20.261 00:21:27.300 Uttam Kumaran: just to make sure everything’s working. I don’t mind like, I think, Miguel, once once we get it out, I think me and you can work on like.

179 00:21:27.570 00:21:36.179 Uttam Kumaran: for example, after like 10 responses that I wanted to collect an email and like we’ll work on like that. But that’s fine, because that’s just like improving the chat bot.

180 00:21:36.594 00:21:40.850 Uttam Kumaran: I just wanna make sure like, at least it’s just working. There’s something there.

181 00:21:40.850 00:21:41.240 Miguel de Veyra: Yeah.

182 00:21:41.240 00:21:46.249 Uttam Kumaran: And then we can. We can iterate on the experience and send the the ui, and whatever.

183 00:21:47.000 00:21:47.960 Miguel de Veyra: Okay. Yeah.

184 00:21:49.180 00:21:49.870 Mustafa Raja: So should I.

185 00:21:49.870 00:21:56.189 Uttam Kumaran: For example, like we may we, we may end we may end up. We may end up not doing just the N. 8 N. Block we may like.

186 00:21:56.520 00:22:02.659 Uttam Kumaran: there’s there’s some like really nice chat, chat, bot ui frameworks that we can use.

187 00:22:03.163 00:22:06.070 Uttam Kumaran: But yeah, you can mark this one as done.

188 00:22:06.340 00:22:08.230 Mustafa Raja: Yeah, okay.

189 00:22:09.490 00:22:10.030 Uttam Kumaran: Cool.

190 00:22:12.040 00:22:13.449 Mustafa Raja: So now this one

191 00:22:18.384 00:22:21.299 Mustafa Raja: well, Miguel, I guess this is your ticket.

192 00:22:21.300 00:22:22.310 Mustafa Raja: I think.

193 00:22:24.370 00:22:25.480 Miguel de Veyra: Wait.

194 00:22:28.600 00:22:32.280 Miguel de Veyra: I think, was this about the chat thing, either.

195 00:22:32.580 00:22:36.019 Mustafa Raja: Maybe it’s the linear linear Mcp thing.

196 00:22:36.530 00:22:39.289 Miguel de Veyra: Yeah, but we moved it. Internal review was this.

197 00:22:39.290 00:22:40.080 Casie Aviles: What we discussed.

198 00:22:40.080 00:22:40.870 Miguel de Veyra: Yesterday, Casey.

199 00:22:40.870 00:22:44.629 Uttam Kumaran: Well, I think we were supposed to like. I asked if we could have a meeting about this.

200 00:22:46.022 00:22:49.740 Miguel de Veyra: Yeah, yeah. I think this. We will have a meeting later this week, as he’s mentioned.

201 00:22:49.740 00:22:54.769 Uttam Kumaran: Can we? Can we book that like while we’re on this call? Yeah.

202 00:22:56.675 00:22:57.010 Miguel de Veyra: What.

203 00:22:57.010 00:23:03.580 Uttam Kumaran: For for me like more, I mean probably tomorrow for me.

204 00:23:03.900 00:23:04.830 Miguel de Veyra: Okay, yeah.

205 00:23:05.770 00:23:07.319 Miguel de Veyra: Let me check your schedule.

206 00:23:08.730 00:23:11.070 Uttam Kumaran: Tomorrow or early Friday.

207 00:23:11.640 00:23:13.349 Uttam Kumaran: You want to do early Friday.

208 00:23:14.743 00:23:20.670 Miguel de Veyra: Wait. Wait. Wait for? Oh, yeah. Yeah. Okay. Sure. I’ll schedule one early Friday.

209 00:23:20.920 00:23:21.540 Uttam Kumaran: Okay.

210 00:23:21.540 00:23:29.310 Miguel de Veyra: You’re free on wait. Sorry. 1039, 8 am. Our time. 8 pm. Our time

211 00:23:30.500 00:23:35.010 Miguel de Veyra: cause. After that I think you’re yeah. You’re full.

212 00:23:35.560 00:23:40.970 Uttam Kumaran: I’m going to this like event here on Thursday, so I don’t know. I’ll see if I can get in and out, but I’m not sure.

213 00:23:40.970 00:23:42.969 Miguel de Veyra: Okay. But yeah, I’ll just schedule it either way.

214 00:23:43.390 00:23:43.980 Uttam Kumaran: Okay.

215 00:23:51.160 00:23:52.630 Awaish Kumar: These both are done.

216 00:23:53.173 00:23:56.399 Mustafa Raja: Okay. So so I should move it. Move these to done.

217 00:23:57.490 00:23:58.340 Awaish Kumar: Yep.

218 00:23:58.830 00:24:01.580 Mustafa Raja: Okay, moving this to them.

219 00:24:02.470 00:24:06.570 Uttam Kumaran: And they’re all these are all in the same same bucket. Right? Wish.

220 00:24:08.005 00:24:10.770 Awaish Kumar: Yeah, in the same bucket. Just include the

221 00:24:11.130 00:24:17.240 Awaish Kumar: like, segregated by the source like notion. And the Google sheets.

222 00:24:17.670 00:24:18.529 Uttam Kumaran: Okay. Okay.

223 00:24:21.980 00:24:23.660 Mustafa Raja: Okay. Now, this one.

224 00:24:23.660 00:24:26.179 Uttam Kumaran: And so. And how does the notion one work again?

225 00:24:26.870 00:24:27.789 Uttam Kumaran: Here we go.

226 00:24:27.790 00:24:28.245 Awaish Kumar: No.

227 00:24:29.160 00:24:30.440 Uttam Kumaran: For anyone, yeah.

228 00:24:33.280 00:24:36.080 Awaish Kumar: Was it related to my ticket, or for the board? Yeah.

229 00:24:36.080 00:24:39.530 Uttam Kumaran: Yeah, like, like, I guess, like no, no, for the for the ingestion.

230 00:24:40.030 00:24:48.609 Awaish Kumar: Yep. So what I’m like for this, basically for these, this ticket, we were interested in moving knowledge base document.

231 00:24:48.850 00:25:00.039 Awaish Kumar: And the Pr summaries document. So basically, I’m picking up those 2 docs and this uploading them as a txt file in the S 3.

232 00:25:01.550 00:25:05.629 Uttam Kumaran: And what’s the are you using the notion, SDK, or how are you doing it?

233 00:25:05.630 00:25:07.460 Awaish Kumar: Yeah, yeah, I’m using notion. SDK.

234 00:25:08.730 00:25:12.860 Uttam Kumaran: And so if I want, if we want to add more files to.

235 00:25:12.860 00:25:13.630 Awaish Kumar: We can.

236 00:25:13.920 00:25:20.189 Uttam Kumaran: The ingestion. So like, do you have like a, is it a list? Or like, yeah? And.

237 00:25:20.190 00:25:22.520 Awaish Kumar: So yeah, it’s just a updating configuration.

238 00:25:22.520 00:25:27.940 Uttam Kumaran: Miguel also worked on Miguel. You worked on some notion ingestion stuff, too. Right?

239 00:25:27.940 00:25:28.620 Miguel de Veyra: Yeah, yeah.

240 00:25:28.620 00:25:29.200 Uttam Kumaran: More.

241 00:25:30.060 00:25:31.010 Uttam Kumaran: Okay.

242 00:25:33.290 00:25:37.119 Awaish Kumar: Okay. So like, I have wrote a script. And right now I’m just

243 00:25:37.560 00:25:47.230 Awaish Kumar: sending 2 pages. But, like it, it is just like configuration. Now, like, we can just update the configuration, add more clients and more pages.

244 00:25:47.390 00:25:52.730 Awaish Kumar: just update the Vinyl file, and it will populate the data to S 3.

245 00:25:53.610 00:25:57.909 Uttam Kumaran: And then another question is like, does, how is it? It’s all coming in as Markdown

246 00:25:59.040 00:26:02.199 Uttam Kumaran: or like, how how does it come up in the text. We’re not sure yet.

247 00:26:03.430 00:26:06.429 Awaish Kumar: It’s like all the text in the notion, Doc, it’s just

248 00:26:06.780 00:26:14.290 Awaish Kumar: grabs the all the para plain text from the paragraphs create a txt file.

249 00:26:15.490 00:26:18.170 Uttam Kumaran: And then does it do sub pages or no?

250 00:26:21.870 00:26:23.370 Awaish Kumar: What that means.

251 00:26:24.180 00:26:28.450 Uttam Kumaran: Like, you know, in notion you could have one page, and then within it you could have a sub page.

252 00:26:28.910 00:26:30.240 Awaish Kumar: Oh, sub pages.

253 00:26:31.760 00:26:36.150 Awaish Kumar: Yeah, it. Like, we can update the script to do that right now.

254 00:26:36.670 00:26:39.280 Awaish Kumar: it just takes the current page. Yeah.

255 00:26:41.370 00:26:46.530 Awaish Kumar: if we have a sub page, we can just give it as a sub page in our configuration.

256 00:26:46.990 00:26:49.439 Awaish Kumar: in vimal file, like we can say

257 00:26:49.580 00:26:54.339 Awaish Kumar: page one, page. 2 things like like that just listed with the Id

258 00:26:54.956 00:27:05.000 Awaish Kumar: in the Vinyl file, it will work. But if we want to ha that script, handle it automatically. Then, yeah, we need to update the script a little bit.

259 00:27:06.830 00:27:13.089 Uttam Kumaran: Okay? And then Miguel, like you have you do is we did this notion thing right? Like what happened to that stuff.

260 00:27:13.320 00:27:23.589 Miguel de Veyra: I sent it to them. But the thing, the the difference between what I did and what needed to be done here is the one I did was on database pages.

261 00:27:25.315 00:27:25.670 Uttam Kumaran: Okay.

262 00:27:25.670 00:27:33.160 Miguel de Veyra: So it’s it’s it’s different, like, I think, Casey, if I remember correctly, the one I did wasn’t working for normal pages right?

263 00:27:35.860 00:27:36.800 Casie Aviles: But it’s still useful.

264 00:27:36.800 00:27:37.430 Casie Aviles: Sounds good.

265 00:27:40.010 00:27:42.260 Uttam Kumaran: Where? Where is that code? Miguel?

266 00:27:42.605 00:27:45.710 Miguel de Veyra: It’s also published on the AI back end thing.

267 00:27:46.620 00:27:48.759 Uttam Kumaran: Can you? Can you create a ticket?

268 00:27:49.421 00:27:53.638 Uttam Kumaran: Or or mustafa? You can create one that’s just like

269 00:27:55.490 00:27:59.900 Uttam Kumaran: Improve notion in ingestion script.

270 00:28:00.100 00:28:03.020 Uttam Kumaran: and we can add a couple of things here.

271 00:28:03.772 00:28:09.079 Mustafa Raja: Okay? do I create it in this to do something.

272 00:28:09.080 00:28:11.370 Uttam Kumaran: Put in the in the ready for development.

273 00:28:11.370 00:28:12.970 Mustafa Raja: Ready for development. Okay.

274 00:28:13.460 00:28:21.480 Uttam Kumaran: Okay, and you can. And you can add so you can just say, improve notion ingestion script.

275 00:28:23.420 00:28:27.859 Uttam Kumaran: And then you could just put like a couple of bullet points here, and then I can fill this out.

276 00:28:28.527 00:28:34.499 Uttam Kumaran: One is, we want to add the ability to scrape sub pages.

277 00:28:37.590 00:28:42.780 Uttam Kumaran: Second, we want to add the ability to scrape databases.

278 00:28:47.100 00:28:53.020 Uttam Kumaran: And is there anything else? I’ll add the ability to scrape, scrape databases.

279 00:28:53.250 00:28:54.730 Mustafa Raja: Database. Okay.

280 00:28:55.380 00:28:57.690 Uttam Kumaran: And was there anything else, Miguel.

281 00:28:57.910 00:29:00.519 Miguel de Veyra: Sub pages, databases.

282 00:29:01.186 00:29:03.399 Mustafa Raja: And do we have an assignee for this.

283 00:29:04.680 00:29:08.779 Miguel de Veyra: No, I all right, that’s pretty much it. I’ll add what I what I can think of later.

284 00:29:08.780 00:29:11.550 Uttam Kumaran: Yeah, I think you can put you can put a wish on this one.

285 00:29:11.760 00:29:15.470 Mustafa Raja: But we can leave this here. It doesn’t. I don’t think it needs to be immediately done.

286 00:29:15.580 00:29:18.420 Uttam Kumaran: But, for example, we’re gonna wanna bring in

287 00:29:18.740 00:29:25.059 Uttam Kumaran: like I wish in in notion we have all of our all bunch more client documents that we want to bring in.

288 00:29:26.990 00:29:33.059 Uttam Kumaran: you know. And so I want basically, I think, yeah, if you have a yaml structure for this script, then we can add this, there.

289 00:29:34.310 00:29:37.730 Awaish Kumar: Yeah, like we right now, it just takes from Yaml. We can say, like.

290 00:29:38.110 00:29:44.280 Awaish Kumar: client. Didn’t these 10 pages like.

291 00:29:44.700 00:29:45.190 Uttam Kumaran: Perfect.

292 00:29:49.530 00:29:51.930 Mustafa Raja: Yeah, should we move on.

293 00:29:52.760 00:29:53.510 Uttam Kumaran: Yes.

294 00:29:53.510 00:29:55.729 Mustafa Raja: Yeah, so this this ticket.

295 00:29:58.426 00:30:04.370 Casie Aviles: So for my case, I’ve attached the zoom embeddings already and also tested them.

296 00:30:05.310 00:30:10.719 Casie Aviles: So I guess it’s just part of this ticket. But there’s also the slack messages.

297 00:30:13.010 00:30:14.780 Casie Aviles: Yeah, I think I I need.

298 00:30:14.780 00:30:15.360 Mustafa Raja: Oh, yeah.

299 00:30:15.360 00:30:16.979 Casie Aviles: Miguel’s script for that.

300 00:30:20.330 00:30:23.959 Casie Aviles: So that’s the next thing to work on which is to get the slop

301 00:30:24.350 00:30:27.909 Casie Aviles: here. I think it’s related to one of Miguel’s tickets.

302 00:30:27.910 00:30:28.320 Mustafa Raja: Oops!

303 00:30:28.320 00:30:28.810 Casie Aviles: 1st Level.

304 00:30:29.350 00:30:30.699 Mustafa Raja: Put the videos in there.

305 00:30:30.700 00:30:36.640 Uttam Kumaran: So for the for the zoom and embeddings like, are we? We’re not doing the free retrieval

306 00:30:36.850 00:30:38.270 Uttam Kumaran: stuff for for zoom. Yeah.

307 00:30:40.636 00:30:45.589 Casie Aviles: What I did for zoom is to just add some metadata.

308 00:30:47.100 00:30:48.940 Casie Aviles: I just characterized him.

309 00:30:49.625 00:30:51.449 Casie Aviles: Okay, that’s the pre retrieve item.

310 00:30:56.514 00:30:58.359 Uttam Kumaran: Is there anything else we should do for that.

311 00:30:58.360 00:31:07.709 Miguel de Veyra: No, no cause. The different. The pre retrieval is basically the reason we need a pre retrieval for slack is because we needed to group the messages to make sure they’re, you know, the context is the

312 00:31:08.620 00:31:12.569 Miguel de Veyra: yeah for this. It’s 1 meeting, right? So it doesn’t. Really. There’s no.

313 00:31:12.570 00:31:12.970 Casie Aviles: Yeah.

314 00:31:12.970 00:31:17.720 Miguel de Veyra: As long as you add the metadata properly, and we can filter that later on for retrieval phase.

315 00:31:19.500 00:31:20.080 Casie Aviles: Yes.

316 00:31:20.080 00:31:20.710 Uttam Kumaran: Okay.

317 00:31:22.290 00:31:23.860 Mustafa Raja: Okay, should we move on.

318 00:31:23.860 00:31:24.490 Miguel de Veyra: Yeah.

319 00:31:26.770 00:31:27.260 Mustafa Raja: And.

320 00:31:27.260 00:31:32.289 Uttam Kumaran: Yeah. So the so the only thing remaining for this one is the slack work which I know is, I think.

321 00:31:32.650 00:31:33.929 Uttam Kumaran: for tomorrow.

322 00:31:33.930 00:31:34.770 Miguel de Veyra: Yeah, yeah.

323 00:31:35.980 00:31:36.370 Mustafa Raja: Okay.

324 00:31:36.370 00:31:41.190 Uttam Kumaran: Yeah, Miguel, honestly like, if the pool parts image stuff is like too much, I would just work the slack work as

325 00:31:42.170 00:31:44.500 Uttam Kumaran: we didn’t promise them that anyways so.

326 00:31:45.430 00:31:47.199 Miguel de Veyra: Oh, okay. Yeah.

327 00:31:47.200 00:31:48.320 Uttam Kumaran: 2 week, the only

328 00:31:48.440 00:31:53.840 Uttam Kumaran: they only gave us 2 weeks. And they wanted this thing. I’m like, we’re not gonna do all that. If we if we get to it, we’ll get to it.

329 00:31:54.100 00:31:58.570 Uttam Kumaran: And so today, we’re basically having conversation. We’re trying to sign them onto a longer deal.

330 00:31:59.000 00:32:03.090 Uttam Kumaran: So I I was like. Look, if you have it already. There, I was like, let’s ship it.

331 00:32:03.090 00:32:04.900 Miguel de Veyra: Okay. Okay, yeah. I’ll I’ll work.

332 00:32:04.900 00:32:07.469 Uttam Kumaran: Whatever, it’s not like, yeah, yeah.

333 00:32:08.410 00:32:17.479 Mustafa Raja: Yeah. So for this one I’ll be working within Clay’s integration with Hubspot today.

334 00:32:19.770 00:32:20.350 Miguel de Veyra: Okay.

335 00:32:20.350 00:32:22.510 Mustafa Raja: This, this is my right. So.

336 00:32:23.080 00:32:27.270 Miguel de Veyra: Will you? Have you talked to Ericsson about clay and stuff, or not yet?

337 00:32:27.440 00:32:29.339 Miguel de Veyra: Or have we had that meeting.

338 00:32:30.132 00:32:40.459 Uttam Kumaran: We like. I asked him, because he said he was gonna run a training. I added him to the Articles Channel. So in case you have any questions, Mustafa. Just ping Erickson.

339 00:32:40.460 00:32:45.530 Uttam Kumaran: It’s just a friend friend of the company. He’s like. He runs a clay related agency.

340 00:32:45.530 00:32:46.420 Mustafa Raja: But I.

341 00:32:46.420 00:32:53.539 Uttam Kumaran: I think we you basically figured it out yesterday. I think the one thing that’s helpful is your problem. Are you looking at the other clay workbooks that we have.

342 00:32:53.900 00:32:56.300 Mustafa Raja: Yeah, I did. I did look at some of those.

343 00:32:56.300 00:32:58.869 Uttam Kumaran: Okay, cool as long as you’re looking at those. And that has like

344 00:32:59.340 00:33:01.265 Uttam Kumaran: that has like a bunch of stuff.

345 00:33:01.540 00:33:01.880 Mustafa Raja: Yeah.

346 00:33:01.880 00:33:05.490 Uttam Kumaran: The only other thing I’ll probably send you is I will send you

347 00:33:06.850 00:33:13.966 Uttam Kumaran: Here, let me. I’ll send. I’ll send it to you. I’ll send it in slack is, I have another sort of examples, clay workbook, that you should

348 00:33:14.820 00:33:18.860 Uttam Kumaran: that you should check out just in case you need inspiration on how to build any of this.

349 00:33:19.140 00:33:20.970 Mustafa Raja: Yeah, yeah, thank you.

350 00:33:21.590 00:33:28.049 Uttam Kumaran: I just wish Clay had an Api it it only has. That’s right.

351 00:33:28.050 00:33:28.650 Mustafa Raja: Yeah.

352 00:33:28.650 00:33:29.620 Uttam Kumaran: Yeah.

353 00:33:30.428 00:33:33.710 Uttam Kumaran: I mean, once we do it. If we do it in clay first, st

354 00:33:33.880 00:33:35.970 Uttam Kumaran: then we can decide to move it.

355 00:33:35.970 00:33:36.510 Mustafa Raja: Okay.

356 00:33:37.190 00:33:38.859 Uttam Kumaran: Somewhere else later. Right? So that.

357 00:33:38.860 00:33:41.700 Mustafa Raja: Yeah, yeah, yeah, we can always move it.

358 00:33:42.100 00:33:42.800 Awaish Kumar: No.

359 00:33:43.070 00:33:46.480 Mustafa Raja: Okay. So next, should we move to next ticket.

360 00:33:47.300 00:33:48.150 Uttam Kumaran: Yes.

361 00:33:48.350 00:33:49.810 Mustafa Raja: Yeah, so this one.

362 00:33:52.450 00:33:54.419 Casie Aviles: I think this is related to that.

363 00:33:55.370 00:34:09.149 Miguel de Veyra: Yeah, this is related to that thing where I only have Eden and Yavi’s done the Eden. I’m gonna run. Now. I’m probably yeah. I’m running now and then. I wish I need the rest in super base. I think you’ve already noted on that.

364 00:34:10.538 00:34:16.800 Awaish Kumar: Yeah. So I actually have asked the related team members to send me the client ids

365 00:34:17.020 00:34:20.250 Awaish Kumar: the slack Channel Id is basically for every client.

366 00:34:20.250 00:34:21.080 Miguel de Veyra: I think we can.

367 00:34:21.080 00:34:23.269 Awaish Kumar: Add it into my configuration setup.

368 00:34:23.270 00:34:25.729 Miguel de Veyra: Okay, I I think we can send that to you. Right?

369 00:34:25.730 00:34:27.820 Awaish Kumar: If you yeah, if you know

370 00:34:28.395 00:34:33.270 Awaish Kumar: the for our clients, like the ids of all external and internal channels.

371 00:34:33.920 00:34:39.299 Awaish Kumar: just share it with me. I can just upgrade the config, and it should be fair after that.

372 00:34:39.760 00:34:41.489 Miguel de Veyra: Okay, yeah, sure. We’ll send it to you.

373 00:34:42.810 00:34:44.830 Mustafa Raja: Okay, should we move on.

374 00:34:44.830 00:34:47.109 Uttam Kumaran: Can we? Can you, can you? Actually.

375 00:34:47.230 00:34:50.169 Uttam Kumaran: I think you guys should work on this together because.

376 00:34:50.179 00:34:50.499 Mustafa Raja: Okay.

377 00:34:50.922 00:34:54.729 Uttam Kumaran: Miguel, the Dagster is just like a python executor.

378 00:34:54.969 00:35:08.259 Uttam Kumaran: So you can schedule any of your jobs in there. It’s basically gonna be our replacement for Windmill. But anything we want to script and like Trigger, we’re gonna use Dagster. So I think it’d be helpful for you to meet

379 00:35:08.640 00:35:11.119 Uttam Kumaran: with a wish and just do this together.

380 00:35:12.620 00:35:21.509 Uttam Kumaran: because I have experience in Dexter and airflow, like, I’m okay with it. And like, I don’t know, maybe you guys can meet today or like Casey, you can join, or Mustafa. You can join.

381 00:35:21.510 00:35:22.110 Mustafa Raja: Yeah.

382 00:35:22.110 00:35:46.729 Uttam Kumaran: And just like how to use Daxter, because that’s gonna be our orchestration workflow. Anything that we need to trigger on a schedule or trigger on any sort of external trigger. We can use Daxter to do so. So we’re gonna move stuff. We’re gonna try to end up moving everything from windmill and stuff into one place there. That way. When any jobs you need to trigger, you can do it there. So if you guys can

383 00:35:47.260 00:35:52.329 Uttam Kumaran: do that today, or or meet whenever like. I think it would be really helpful for everybody to know.

384 00:35:53.658 00:35:59.409 Miguel de Veyra: Cause I’m gonna do the backfill first.st Do. Are you also doing that on Daxter? A wish.

385 00:36:01.070 00:36:02.180 Awaish Kumar: Sorry, what.

386 00:36:02.180 00:36:08.490 Miguel de Veyra: Cause. Basically, the thing the 1st thing we need to do before scheduling is we need to backfill everything right.

387 00:36:08.820 00:36:13.720 Miguel de Veyra: I need to run that in there. Is that also something that you run on that.

388 00:36:13.720 00:36:15.840 Uttam Kumaran: You can do that in Daxer, too.

389 00:36:16.472 00:36:27.600 Uttam Kumaran: You can create a job for backfill and just not put it on a schedule so you can just run it manually, and then you can leave it there, meaning like anytime we need to run a backfill.

390 00:36:28.050 00:36:31.020 Uttam Kumaran: You have that script, but then it’s not on a schedule.

391 00:36:31.180 00:36:32.280 Miguel de Veyra: Okay. Okay. Sure.

392 00:36:34.090 00:36:39.319 Awaish Kumar: Yes, right. It’s not like airflow. Catch up thing. You have to write your script to do that.

393 00:36:39.945 00:36:40.570 Miguel de Veyra: Yeah.

394 00:36:40.570 00:36:41.799 Uttam Kumaran: Yeah, that’s fine. Yeah.

395 00:36:43.644 00:36:44.819 Mustafa Raja: Should you move on.

396 00:36:44.820 00:36:45.470 Miguel de Veyra: Yep.

397 00:36:47.290 00:36:48.140 Mustafa Raja: Yeah, sure.

398 00:36:48.650 00:36:51.849 Uttam Kumaran: Yeah, Miguel, do you wanna do you wanna book this today? Or.

399 00:36:53.206 00:36:54.140 Miguel de Veyra: Yeah, sure.

400 00:36:55.340 00:36:57.170 Uttam Kumaran: Okay, I just wanna make sure it happens.

401 00:36:59.700 00:37:00.140 Mustafa Raja: Yeah.

402 00:37:00.580 00:37:01.600 Awaish Kumar: And this one.

403 00:37:09.165 00:37:11.120 Miguel de Veyra: Yes, right.

404 00:37:11.120 00:37:14.850 Uttam Kumaran: There’s no assignee but this one. I sent a note about this.

405 00:37:17.490 00:37:21.630 Uttam Kumaran: Does anyone wanna try to take this one like

406 00:37:22.350 00:37:25.860 Uttam Kumaran: I think I know, Mustafa, you’re busy on Hubspot stuff.

407 00:37:26.420 00:37:28.130 Mustafa Raja: And the.

408 00:37:28.130 00:37:31.360 Uttam Kumaran: Like, just a simple, Yeah, go ahead.

409 00:37:31.360 00:37:35.109 Mustafa Raja: Yeah, yeah, yeah. I guess I’ll be done with it today.

410 00:37:37.000 00:37:37.500 Uttam Kumaran: Yeah.

411 00:37:37.500 00:37:45.789 Uttam Kumaran: I think I think it either. You I mean anyone. I think this one’s not like so bad. But I think it’d be helpful for anyone who wants experience in clay.

412 00:37:46.040 00:37:59.260 Uttam Kumaran: Basically, the goal here is we’re we are trying to search for a particular type of company which are called implementation partners. And we want to basically find them sorted by

413 00:37:59.500 00:38:02.290 Uttam Kumaran: the types of software. They implement

414 00:38:03.038 00:38:06.539 Uttam Kumaran: because we’re gonna reach out to them and propose a partnership.

415 00:38:07.294 00:38:11.560 Uttam Kumaran: So I I did a couple. I did like a a few manual searches.

416 00:38:11.940 00:38:17.870 Uttam Kumaran: And then I put the example there. And we basically want to use clay to fill this out Clay and Apollo to fill this out.

417 00:38:18.610 00:38:19.560 Uttam Kumaran: So

418 00:38:24.820 00:38:26.520 Uttam Kumaran: I don’t know, like, I guess

419 00:38:26.640 00:38:29.290 Uttam Kumaran: I know, Mustafa, you’re doing a lot of sales stuff like

420 00:38:29.760 00:38:32.479 Uttam Kumaran: Miguel, or I know, Casey, you’ve done some of this.

421 00:38:32.610 00:38:38.049 Uttam Kumaran: but I know it’s probably not that interesting Miguel, is it? Is this is any of this, that interesting.

422 00:38:40.690 00:38:41.960 Miguel de Veyra: Transaction.

423 00:38:41.960 00:38:43.790 Uttam Kumaran: It could be like. It’s pretty chill work.

424 00:38:44.100 00:38:46.180 Casie Aviles: It’s just enrichment, right?

425 00:38:46.180 00:38:47.400 Mustafa Raja: Yeah, yeah.

426 00:38:47.698 00:38:49.490 Miguel de Veyra: Yeah, sure. I’ll look into it.

427 00:38:51.100 00:38:52.229 Mustafa Raja: Should I assign?

428 00:38:52.470 00:38:56.499 Miguel de Veyra: Yeah, yeah, yeah, yeah. Wait. Who can I reach out to about this? Is it? Aish?

429 00:38:56.790 00:38:58.990 Uttam Kumaran: I would reach out to Mustafa about this.

430 00:38:58.990 00:39:00.970 Miguel de Veyra: Oh, Mustafa. Sorry. Sorry. Yeah. Yeah. Okay.

431 00:39:00.970 00:39:11.059 Uttam Kumaran: But like I mean, of course, like I, I put the ticket up, but in terms of like how to do the clay work, I would I would reach out to Mustafa, and then, worst case, if there’s more stuff reach out to Erickson.

432 00:39:12.900 00:39:17.460 Uttam Kumaran: But I think this would be pretty chill like, I just want everybody on the team to know how to use clay. Basically.

433 00:39:17.940 00:39:21.709 Mustafa Raja: Yeah, yeah, should we move on.

434 00:39:22.320 00:39:22.970 Uttam Kumaran: Yes.

435 00:39:25.760 00:39:26.790 Mustafa Raja: And then this.

436 00:39:26.790 00:39:27.330 Casie Aviles: Yeah.

437 00:39:30.170 00:39:34.665 Casie Aviles: So I added some comments at the bottom just before the meeting.

438 00:39:35.150 00:39:41.219 Casie Aviles: I basically just added this zoom search card to the existing ui.

439 00:39:41.900 00:39:49.930 Casie Aviles: And then there’s a page that when you click, it’s gonna so these are just dummy data. But it’s gonna look like that.

440 00:39:50.270 00:39:56.290 Casie Aviles: And you could search the Zoom Meeting. So I started with just the search ability first.st

441 00:39:59.180 00:40:00.130 Casie Aviles: Yeah. So.

442 00:40:00.130 00:40:00.560 Uttam Kumaran: Nice.

443 00:40:00.560 00:40:02.540 Casie Aviles: Search in the meetings, and

444 00:40:02.960 00:40:08.939 Casie Aviles: I guess the next step here is to connect the super base records that we have. So it appears here.

445 00:40:08.940 00:40:09.400 Mustafa Raja: No.

446 00:40:09.400 00:40:11.480 Miguel de Veyra: Do you? Do? You have the

447 00:40:11.770 00:40:16.929 Miguel de Veyra: the you cause? Ideally, we wanna we want the users to be able to watch the videos here right.

448 00:40:18.170 00:40:18.980 Casie Aviles: Yeah.

449 00:40:18.980 00:40:20.459 Miguel de Veyra: Do we have the?

450 00:40:21.010 00:40:22.609 Miguel de Veyra: Oh, we don’t have it yet. Okay.

451 00:40:22.940 00:40:24.640 Casie Aviles: Just the transcripts.

452 00:40:24.640 00:40:26.380 Miguel de Veyra: Yeah, because.

453 00:40:27.420 00:40:33.010 Casie Aviles: Yeah, I guess I’m thinking how we how we should do it. Because

454 00:40:33.801 00:40:42.229 Casie Aviles: when it’s on Google drive, it’s easily shareable. I mean you could share it, and then you could view it with the link. But from S. 3 I don’t think.

455 00:40:42.570 00:40:44.862 Uttam Kumaran: I’m not sure if that’s possible. But

456 00:40:45.190 00:40:47.860 Uttam Kumaran: build a little like video viewer.

457 00:40:47.860 00:40:55.589 Miguel de Veyra: Yeah. And then basically, what you have to do is, do you need like a public thing? Yeah, public. URL, in.

458 00:40:55.590 00:41:04.169 Uttam Kumaran: It. It needs to be public, URL, but like it basically needs to be hidden behind like the the site will be locked. Basically.

459 00:41:06.140 00:41:06.920 Casie Aviles: Okay.

460 00:41:08.350 00:41:11.110 Miguel de Veyra: And then use that as like. Source. Yada Yada.

461 00:41:13.020 00:41:13.770 Mustafa Raja: Yep.

462 00:41:13.770 00:41:14.110 Casie Aviles: Thanks.

463 00:41:18.640 00:41:23.899 Mustafa Raja: Yes, that’s all from, and that is.

464 00:41:23.900 00:41:26.419 Casie Aviles: It’s the same. Yeah, it’s the same, because I think.

465 00:41:26.810 00:41:30.520 Mustafa Raja: Okay, and then we have these.

466 00:41:30.520 00:41:37.599 Uttam Kumaran: So I guess, Casey, can you send a loom on the zoom transcript thing like whenever you make progress?

467 00:41:37.810 00:41:48.079 Uttam Kumaran: Because I think amber can give feedback. I can get some feedback. There’s a good couple of big things to keep in mind. One is like, we yeah, we want users to be able to search

468 00:41:48.270 00:42:02.289 Uttam Kumaran: Zoom Meetings by the title, by attendees by the team. Right? All that metadata. The second thing we wanna also do is people can watch the video right there, they can. Also. The 3rd thing is, they should be able to download it

469 00:42:02.290 00:42:07.579 Uttam Kumaran: right? So these are just added functionalities to think about, not in any priority, but like

470 00:42:07.700 00:42:19.389 Uttam Kumaran: just things to think about when you’re sort of vibe coding it. The other piece is, we want the video and a chat window to appear next to each other, because you want to be able to chat with the meeting

471 00:42:19.770 00:42:22.530 Uttam Kumaran: like while you’re watching it. Does that make sense.

472 00:42:22.530 00:42:27.420 Casie Aviles: And so it’s like side by side. The video in the chat interface.

473 00:42:27.950 00:42:28.860 Uttam Kumaran: Yes.

474 00:42:29.320 00:42:30.370 Mustafa Raja: Okay. Yeah.

475 00:42:30.370 00:42:36.460 Uttam Kumaran: So the the other piece is like we want to be able to download the transcript right.

476 00:42:37.832 00:42:41.758 Uttam Kumaran: And then the last piece we wanna make sure is that we can.

477 00:42:42.859 00:42:46.689 Uttam Kumaran: eventually. What we’ll do is we’ll have some helpful buttons, for example.

478 00:42:46.830 00:42:49.190 Uttam Kumaran: create linear tickets from this meeting.

479 00:42:49.660 00:42:53.800 Uttam Kumaran: or like draft linear tickets from this meeting. Send me the transcript in slack.

480 00:42:53.800 00:42:54.340 Mustafa Raja: Calendar.

481 00:42:54.610 00:42:58.389 Uttam Kumaran: Like email me the transfer. Right? We’ll we’ll we’ll add these helpful features.

482 00:42:59.652 00:43:04.030 Uttam Kumaran: But I I definitely don’t. Don’t worry about how dense the screen is like

483 00:43:04.260 00:43:09.320 Uttam Kumaran: we want to have probably, like the search we want to have probably the list of meetings on the left.

484 00:43:09.530 00:43:13.379 Uttam Kumaran: the video and then the chat somewhere. So like, probably those 3 elements.

485 00:43:13.990 00:43:17.089 Uttam Kumaran: and then extra room for for buttons, probably.

486 00:43:17.550 00:43:18.300 Uttam Kumaran: Okay.

487 00:43:18.530 00:43:18.860 Casie Aviles: Okay.

488 00:43:18.860 00:43:25.579 Mustafa Raja: Yeah, yeah, so that’s pretty much it.

489 00:43:25.580 00:43:34.060 Uttam Kumaran: Once. Once, yeah, okay, okay, how does everyone feel about this week? So I know, tomorrow we have.

490 00:43:34.060 00:43:34.670 Casie Aviles: World.

491 00:43:34.850 00:43:37.919 Uttam Kumaran: The slack. We want to get slack and zoom done by tomorrow.

492 00:43:37.920 00:43:38.610 Casie Aviles: Well.

493 00:43:39.001 00:43:41.108 Uttam Kumaran: Is there anything else that’s like?

494 00:43:41.460 00:43:42.140 Mustafa Raja: I’m not.

495 00:43:42.580 00:43:45.459 Uttam Kumaran: We’re sort of like, yeah, we’re not sure if it’s gonna get done.

496 00:43:46.660 00:43:47.569 Mustafa Raja: Hi! Everyone.

497 00:43:47.858 00:43:50.169 Casie Aviles: I think so far none. The only thing.

498 00:43:50.610 00:43:53.730 Miguel de Veyra: Just to clarify pool parts. I stopped first, st right.

499 00:43:54.980 00:43:57.420 Uttam Kumaran: I would I would send the note to Amber

500 00:43:58.700 00:44:01.479 Uttam Kumaran: and tell her. Hey? I did some research.

501 00:44:02.660 00:44:04.260 Uttam Kumaran: It’s gonna take longer.

502 00:44:04.890 00:44:09.469 Uttam Kumaran: Do you need me to work on it. Right? That’s I would just I would let her make the decision.

503 00:44:10.690 00:44:12.610 Uttam Kumaran: but share that with her.

504 00:44:13.620 00:44:24.260 Uttam Kumaran: I, from my perspective, is it’s not required for today, because they only gave us 2 weeks to work on this. And we clearly put in the list

505 00:44:24.670 00:44:26.820 Uttam Kumaran: that we’re not gonna have this done by then.

506 00:44:27.160 00:44:36.080 Uttam Kumaran: So I’m I don’t mind this happening. I just I want to set her up for success. So I would just let her know what you, what you decide.

507 00:44:36.380 00:44:37.130 Miguel de Veyra: Okay.

508 00:44:41.950 00:44:48.240 Uttam Kumaran: okay? And then, yeah, I think for the rest of the week, I really feel like it’s mostly gonna be getting slack and zoom done.

509 00:44:49.690 00:44:58.719 Uttam Kumaran: We also have. I think, Miguel, we have 2 things, I think, after Wednesday, if you, if you have more time, is one, I think, helping Casey get the zoom

510 00:44:59.418 00:45:02.909 Uttam Kumaran: thing out and sort of recording that whole process

511 00:45:04.530 00:45:10.700 Uttam Kumaran: team can sort of vibe code apps. And so we can do that, I think, getting getting yourself familiar with Dagster.

512 00:45:12.100 00:45:23.519 Uttam Kumaran: And then I think hopefully, by the end of the week everyone will have done some work on play right? So we’re gonna be shifting a lot of our time to helping Robert and myself on the sales side.

513 00:45:23.520 00:45:24.190 Mustafa Raja: Otherwise.

514 00:45:24.190 00:45:27.820 Uttam Kumaran: The zoom thing helps the sales work because we have a lot of sales meetings where we need to generate.

515 00:45:27.820 00:45:29.370 Mustafa Raja: Like.

516 00:45:29.370 00:45:35.850 Uttam Kumaran: Follow ups from but you’ll see. We’ll sort of expand the amount of features.

517 00:45:38.330 00:45:40.730 Uttam Kumaran: And then we’ll keep filling out the

518 00:45:41.430 00:45:44.769 Uttam Kumaran: about the ready for development. So if everyone finishes work.

519 00:45:45.660 00:45:48.699 Uttam Kumaran: we can have a quick conversation and pull stuff in the backlog.

520 00:45:51.280 00:45:51.820 Miguel de Veyra: Okay.

521 00:45:54.360 00:45:54.970 Uttam Kumaran: Okay.

522 00:45:55.250 00:45:55.850 Mustafa Raja: Okay. Cool.

523 00:45:55.870 00:46:02.059 Uttam Kumaran: So yeah, I guess just I’ll just hammer, hammer it again. Send looms, send slacks

524 00:46:02.433 00:46:17.739 Uttam Kumaran: give me some info. Give me something to run. Give me something to give you feedback on quickly. That’s what I’m probably most helpful with. Is you send me the information or send me a question, I could help answer it. If not, we’re just gonna come back to this meeting and have the conversation. So

525 00:46:17.930 00:46:20.700 Uttam Kumaran: if I have one, ask that that would be it so.

526 00:46:21.200 00:46:21.850 Miguel de Veyra: Okay.

527 00:46:22.150 00:46:26.989 Uttam Kumaran: The last piece is, I’m hearing a lot of really good stuff about opus

528 00:46:29.920 00:46:32.899 Uttam Kumaran: I think at some point like we should try to

529 00:46:33.240 00:46:36.300 Uttam Kumaran: maybe use it in cursor to Vibe Code. Some of this.

530 00:46:37.115 00:46:41.669 Uttam Kumaran: I was talking to Mustafa about it. The Gemini model is also pretty good.

531 00:46:41.870 00:46:42.730 Miguel de Veyra: Yeah, it’s super expensive.

532 00:46:42.730 00:46:47.480 Uttam Kumaran: We have any. I don’t think we. Yeah, I don’t think we have any direct use case right now.

533 00:46:50.600 00:46:52.840 Uttam Kumaran: But maybe.

534 00:46:52.840 00:47:02.519 Mustafa Raja: Gemini wouldn’t be Gemini wouldn’t be good in cursor. Gemini. Gemini is good for tool usage in general agents.

535 00:47:03.485 00:47:13.709 Mustafa Raja: But it has some issues with anything. If we work with lama index or line chain it, it is pretty good, I’d say better than Gbt.

536 00:47:16.660 00:47:19.060 Uttam Kumaran: Yeah. So I think we maybe can consider.

537 00:47:19.180 00:47:23.370 Uttam Kumaran: I mean, maybe we should consider it for the slack agents later.

538 00:47:23.820 00:47:25.910 Uttam Kumaran: But otherwise, like, I feel like

539 00:47:26.400 00:47:30.009 Uttam Kumaran: the only other thing that I want. We still haven’t finished up.

540 00:47:30.140 00:47:32.089 Uttam Kumaran: Miguel is like this client hub

541 00:47:32.230 00:47:35.200 Uttam Kumaran: work that we were we originally were doing where it’s like

542 00:47:35.360 00:47:39.299 Uttam Kumaran: once a client. So the thesis is like once a client signs with us

543 00:47:39.951 00:47:47.110 Uttam Kumaran: all of the documents, which is like the notion docs, but also Pdfs and things like that. All get put into one agent.

544 00:47:47.500 00:47:53.020 Uttam Kumaran: which is basically like what we’re doing with. It’s like the slack agent work is that work, but, like

545 00:47:54.890 00:47:59.569 Uttam Kumaran: we sort of want it want it to. Probably we. One thing we’ll end up doing is probably have that

546 00:47:59.990 00:48:05.324 Uttam Kumaran: sit outside of slack, too, because the thing I’m realizing with slack is like handling

547 00:48:06.330 00:48:12.390 Uttam Kumaran: files is hard handling it. Sending back stuff is hard, like you’re kind of constrained.

548 00:48:12.390 00:48:12.780 Mustafa Raja: Okay.

549 00:48:12.780 00:48:16.780 Uttam Kumaran: So we may end up building like just chat interfaces for every client

550 00:48:17.430 00:48:20.649 Uttam Kumaran: like in our demo site, like or somewhere else.

551 00:48:20.780 00:48:21.290 Miguel de Veyra: Yeah.

552 00:48:22.560 00:48:25.339 Uttam Kumaran: And for that maybe Gemini is the best.

553 00:48:26.070 00:48:26.970 Uttam Kumaran: Raoul.

554 00:48:30.827 00:48:34.799 Uttam Kumaran: The last piece is, I’m signing the partnership with contextual.

555 00:48:35.120 00:48:38.690 Uttam Kumaran: We’re gonna they’re gonna put us all in through training and certification.

556 00:48:38.840 00:48:41.429 Uttam Kumaran: They’re also gonna give us a thousand dollars free

557 00:48:41.710 00:48:44.430 Uttam Kumaran: to use with for contextual credits.

558 00:48:46.270 00:48:48.649 Uttam Kumaran: So I think once we get

559 00:48:48.920 00:48:52.890 Uttam Kumaran: this week done, we can talk about that hopefully. That gets spun up soon.

560 00:48:53.380 00:48:54.030 Miguel de Veyra: Okay.

561 00:48:57.090 00:49:02.440 Uttam Kumaran: Okay. Cool, alright. Great. Anything else.

562 00:49:08.340 00:49:10.830 Miguel de Veyra: No, I think that’s it, and we’re a bit over time.

563 00:49:11.730 00:49:16.980 Uttam Kumaran: Okay, thank you. Guys, thank you for hosting.

564 00:49:16.980 00:49:19.180 Mustafa Raja: Yeah. Thank you. Bye-bye.