2025-12-16_ai-app_standup_76810be0

Meeting Title: AI-App Standup Date: 2025-12-16 Meeting participants: Samuel Roberts, Amber Lin, Mustafa Raja, Uttam Kumaran, Pranav, Gabriel Lam, Surf’s iPhone

---

WEBVTT

1 00:00:17.880 ⇒ 00:00:19.160 Amber Lin: Hello.

2 00:00:20.320 ⇒ 00:00:21.310 Samuel Roberts: Ayy.

3 00:00:21.600 ⇒ 00:00:22.810 Samuel Roberts: Sorry about that.

4 00:00:22.810 ⇒ 00:00:26.540 Amber Lin: It always happens. I’ve done this so many times, it’s okay.

5 00:00:26.540 ⇒ 00:00:29.809 Samuel Roberts: But does it… so does it remove it when you transfer ownership? Is that what happened?

6 00:00:29.810 ⇒ 00:00:36.689 Amber Lin: I… I think he never added it, because when I add meetings too fast.

7 00:00:37.070 ⇒ 00:00:38.459 Samuel Roberts: Doesn’t get enough time.

8 00:00:38.460 ⇒ 00:00:41.789 Amber Lin: to add Zoom, so it always happens.

9 00:00:41.790 ⇒ 00:00:54.319 Samuel Roberts: Okay, yeah, see, I wasn’t even on the original invite, so I didn’t see that until he made me owner of it, and then I must have just disappeared. And then, of course, I’m sitting on my, like, personal calendar, which I’ve shared the work calendar.

10 00:00:54.320 ⇒ 00:00:58.780 Amber Lin: calendar, too, and it won’t let me add a Zoom meeting, and I’m just like, why won’t it let me add? I’m in calendar.

11 00:00:58.780 ⇒ 00:01:00.709 Samuel Roberts: I realized I’m in the wrong one, and… but yeah.

12 00:01:00.710 ⇒ 00:01:01.520 Amber Lin: Yes, please.

13 00:01:01.520 ⇒ 00:01:06.489 Samuel Roberts: So I wasted a few minutes doing that, but… Oh, okay, there we go.

14 00:01:08.170 ⇒ 00:01:11.309 Amber Lin: Have you guys figured out the permissions?

15 00:01:12.960 ⇒ 00:01:15.300 Samuel Roberts: No.

16 00:01:16.030 ⇒ 00:01:21.299 Amber Lin: I believe we own that doc, but we can discuss it later.

17 00:01:21.300 ⇒ 00:01:23.369 Samuel Roberts: Okay, yeah, I wasn’t sure, like…

18 00:01:23.980 ⇒ 00:01:27.560 Samuel Roberts: If it’s something that we would have changed, or that they might have changed, or if it’s just a…

19 00:01:27.730 ⇒ 00:01:33.340 Samuel Roberts: thing that expired, maybe? Mustopo.

20 00:01:33.340 ⇒ 00:01:37.409 Amber Lin: I transfer only… I’ll figure that out. Let’s talk about the other stuff.

21 00:01:37.630 ⇒ 00:01:38.250 Samuel Roberts: Okay.

22 00:01:38.840 ⇒ 00:01:49.050 Uttam Kumaran: Okay, so maybe let’s talk about, maybe, Sam, I can… so, maybe I’ll just, like, tee up the goal for this meeting. So, I would say very similar to…

23 00:01:49.200 ⇒ 00:02:01.910 Uttam Kumaran: the other meetings we’re running, so… so far now, we have one for sort of DEA stuff, we have one, sort of, strategy analysis, and this one I’m just considering, like, AI and application.

24 00:02:02.220 ⇒ 00:02:14.389 Uttam Kumaran: You can… I’ll let you think about a name and a mascot in your own time. But I… this is sort of heading kind of the direction that Clarence mentioned, is starting to build…

25 00:02:14.440 ⇒ 00:02:24.079 Uttam Kumaran: Like, ownership over service lines, and a service line is, like, this is the, this is, like, if we were to…

26 00:02:24.260 ⇒ 00:02:34.709 Uttam Kumaran: I don’t know what analogy I used, but let’s say we’re building planes for customers. This could be the cockpit, something else could be the engine, something else could be the wings.

27 00:02:34.950 ⇒ 00:02:43.710 Uttam Kumaran: For our customers, they may just be buying one part, they may be buying a whole plane, they don’t really care which service line it comes from.

28 00:02:43.810 ⇒ 00:03:00.150 Uttam Kumaran: Right? But internally, we’re gonna start to get organized sort of around these services. I think right now you’re seeing that Robert is leading strategy and analytics. I was leading more of the data stuff, now Wish is leading that daily, and I think, Sam, you’ll be able to start to lead this.

29 00:03:00.150 ⇒ 00:03:04.780 Samuel Roberts: I think we’re gonna get a little bit more formalized on, like, what leading this means.

30 00:03:04.780 ⇒ 00:03:08.989 Uttam Kumaran: And, like, what everybody who’s in this meeting, your role is, but…

31 00:03:09.520 ⇒ 00:03:18.909 Uttam Kumaran: just, like, to finish out the end of the year and, like, get through everything, I think we’ll just, like, keep this structure. It’s a bit of… it’s a… it’s… it kind of sucks, because we have three…

32 00:03:19.290 ⇒ 00:03:39.129 Uttam Kumaran: sort of stand-ups. I don’t think anyone should be in all of them, necessarily, anymore. And so it’s… I’ll sort of leave it to you to schedule this, Sam, however you want. I… I wanted to be here today, but I don’t necessarily need to be here every day, so I want to kind of leave that to you. I think for me, the most important thing

33 00:03:39.450 ⇒ 00:03:42.079 Uttam Kumaran: Is that you have ultimate clarity on

34 00:03:42.100 ⇒ 00:04:00.350 Uttam Kumaran: the progress on the clients that are purchasing this service from us. So, that’s ABC with Andy, that is Lilo, and that is, Remo. Of course, some clients we’re closer to further from, but ultimately.

35 00:04:00.390 ⇒ 00:04:05.799 Uttam Kumaran: your job is to have the confidence that you know the status of those projects. Some are bigger than others.

36 00:04:05.960 ⇒ 00:04:08.850 Uttam Kumaran: Of course, we previously had, like, stuff like,

37 00:04:09.060 ⇒ 00:04:24.910 Uttam Kumaran: Interlude, right, we’ve had other AI clients. I’m gonna go get us a bunch more clients, and so it’s all gonna kind of fit under here. The lovely thing is, is this is much more of a full-stack, like, app crew, right? So, our world has been…

38 00:04:24.910 ⇒ 00:04:33.910 Uttam Kumaran: the Brainforge world has started mostly in data, and then we started doing more analytics, and now we’re doing, sort of, we’re doing more, like, full-stack app development.

39 00:04:34.240 ⇒ 00:04:41.290 Uttam Kumaran: Which is, like, this is what I worked in, you know, before Brainforge and building applications, so,

40 00:04:41.570 ⇒ 00:05:00.469 Uttam Kumaran: I’m actually really excited for that. I also think this crew is going to be able to probably adopt AI a lot faster, given that most of the AI tools are being geared towards, like, software engineering. And so it’ll kind of disseminate… what we learn here to, like, how to develop faster will disseminate through the other crews.

41 00:05:00.820 ⇒ 00:05:11.129 Uttam Kumaran: So I don’t have, like, you know, I think over time, we will structure these meetings in a lot clearer way, and probably end up crushing through them in, like, 30 minutes, and then leaving a little bit of time.

42 00:05:11.130 ⇒ 00:05:26.210 Uttam Kumaran: But for now, we’ll just run them like normal stand-ups. You just have, like, 3 client… kind of clients to go through. So, maybe we… we start by talking about Lilo, and then we can end kind of talking about ABC, but I can sort of hand it to you, Sam, to…

43 00:05:27.350 ⇒ 00:05:33.820 Amber Lin: I need to hop at, in 5 minutes to the AV… their client… my client, the client’s call, so…

44 00:05:33.980 ⇒ 00:05:41.670 Amber Lin: If I need to be… yeah, they’re doing, like, a internal leadership meeting, like, of the trainers, so I want to be there.

45 00:05:41.670 ⇒ 00:05:47.490 Uttam Kumaran: Okay, okay. So on… so, I guess, yeah, Sam, if you want to give, Amber, like, the update on ABC, where we’re at.

46 00:05:48.080 ⇒ 00:05:56.850 Samuel Roberts: Yeah, so we have been… Working on moving the… Like, text-to-SQL database query.

47 00:05:57.160 ⇒ 00:06:08.689 Samuel Roberts: off of windmill, one, because it’s very slow, and we realized how slow it was when we saw all the times yesterday, and that’s all the cold startup, so that should be pretty good. We then also need to,

48 00:06:09.300 ⇒ 00:06:14.219 Samuel Roberts: Well, right now it’s on par with the current windmill tool, so we want to

49 00:06:14.460 ⇒ 00:06:20.490 Samuel Roberts: improve that prompt, for some of the issues that we saw yesterday, which I think we can do today now that we’ve got it stood up.

50 00:06:23.530 ⇒ 00:06:32.840 Samuel Roberts: The other side of it is the migration stuff, but we’re… that’s, like, we’re getting to that. But, fixing that, going through the…

51 00:06:34.260 ⇒ 00:06:38.290 Samuel Roberts: the errors… what was the status of that after I got off last night? Were there any.

52 00:06:38.290 ⇒ 00:06:41.969 Uttam Kumaran: I didn’t make much progress last night, so that’ll be today.

53 00:06:41.970 ⇒ 00:06:45.940 Samuel Roberts: Okay. So yeah, I think we can get a pretty good sense of those errors today, and then…

54 00:06:46.360 ⇒ 00:06:53.280 Samuel Roberts: See where other things besides the database query tool can use a little bit of…

55 00:06:53.620 ⇒ 00:06:59.709 Samuel Roberts: band-aids on N8N as we’re doing some of the migration stuff, but… And then we’ll also…

56 00:06:59.920 ⇒ 00:07:04.519 Samuel Roberts: I know, Utam, you said you were taking a little bit of the migration plan stuff. Did you want me to…

57 00:07:04.640 ⇒ 00:07:08.160 Samuel Roberts: work on that Gantt or anything today? Like, I… how did you want to split that.

58 00:07:08.160 ⇒ 00:07:16.379 Uttam Kumaran: Yeah, I… so, I’m happy to work on the Gantt. I… I would like the smarter people in the room to work on fixing.

59 00:07:16.380 ⇒ 00:07:16.730 Samuel Roberts: Okay.

60 00:07:16.730 ⇒ 00:07:19.730 Uttam Kumaran: So I can do the Gantt. Basically…

61 00:07:19.870 ⇒ 00:07:22.070 Uttam Kumaran: I just want to bias as much time as possible.

62 00:07:22.790 ⇒ 00:07:26.500 Uttam Kumaran: Additionally, I want to know… what’s…

63 00:07:26.660 ⇒ 00:07:30.730 Uttam Kumaran: what’s possible by next Tuesday, and then what’s

64 00:07:31.180 ⇒ 00:07:33.559 Uttam Kumaran: Well, by… end of the year.

65 00:07:33.680 ⇒ 00:07:48.389 Uttam Kumaran: And so that’s, like, what I’m gonna kind of figure out, and then I can… so that way, I can… when you guys ask for this versus that, I can give you a good understanding. I haven’t… I don’t… I don’t exactly know the whole scope. Also, I’m gonna… I’m gonna drive pushing Tim to get us access to stuff.

66 00:07:48.560 ⇒ 00:07:52.740 Samuel Roberts: Okay, cool. So, if you guys can focus, especially today and tomorrow.

67 00:07:52.930 ⇒ 00:07:57.910 Uttam Kumaran: Because Friday… yeah, as much as we can get done today and tomorrow, in terms of.

68 00:07:57.910 ⇒ 00:07:58.340 Samuel Roberts: Okay.

69 00:07:58.340 ⇒ 00:08:04.219 Uttam Kumaran: Getting the new system going, being able to, have great observability and edit prompts.

70 00:08:04.390 ⇒ 00:08:11.289 Uttam Kumaran: And test, like, that… that would be golden. Because I can… then I can start to go fix a lot of the issues, you know, so…

71 00:08:11.290 ⇒ 00:08:22.599 Samuel Roberts: Okay, yeah, I think we can do that. I know Mustafa and I were talking about that a little bit. He didn’t get a chance to tie the tool into the current Mastra flow yet, but we can probably do that and get, like, a…

72 00:08:22.900 ⇒ 00:08:26.820 Samuel Roberts: a baby version of Andy going that can at least do database lookups.

73 00:08:27.610 ⇒ 00:08:28.709 Uttam Kumaran: Okay.

74 00:08:28.910 ⇒ 00:08:32.150 Samuel Roberts: pretty, pretty quickly, because I think Casey has some of the stuff already set up for the.

75 00:08:32.159 ⇒ 00:08:34.219 Mustafa Raja: Yeah, I think that part is set up.

76 00:08:34.220 ⇒ 00:08:35.280 Samuel Roberts: Yeah, okay.

77 00:08:35.289 ⇒ 00:08:39.029 Mustafa Raja: It’s the new architecture that we, discussed.

78 00:08:39.260 ⇒ 00:08:40.840 Mustafa Raja: This one is pretty good.

79 00:08:40.990 ⇒ 00:08:46.280 Samuel Roberts: Yeah, so okay, so in that case, the monster is, like, already a baby Andy,

80 00:08:47.080 ⇒ 00:08:47.920 Samuel Roberts: it’s not…

81 00:08:48.020 ⇒ 00:08:53.630 Samuel Roberts: been tested and everything, that’s where we’re gonna add some… hopefully get it in Monster Cloud today, and maybe that gives us the observability right out of the box.

82 00:08:53.810 ⇒ 00:08:55.629 Samuel Roberts: If not, we’ll figure out a better way to do that.

83 00:08:55.780 ⇒ 00:09:01.339 Samuel Roberts: But yeah, we should… I think by next week, we should be in a pretty good position then.

84 00:09:01.340 ⇒ 00:09:05.250 Uttam Kumaran: Yeah, so I think, Amber, if you can translate anything, I put a lot of notes in the email.

85 00:09:05.250 ⇒ 00:09:08.000 Samuel Roberts: You can basically say that, like, we’ve been…

86 00:09:08.210 ⇒ 00:09:12.210 Uttam Kumaran: We’ve been planning on moving to this new system, and we’re just accelerating that.

87 00:09:13.730 ⇒ 00:09:17.550 Uttam Kumaran: And then, like, yeah, we’re just… we’re basically gonna meet daily until we’ve fixed it.

88 00:09:17.550 ⇒ 00:09:18.080 Samuel Roberts: Yeah.

89 00:09:18.080 ⇒ 00:09:18.590 Uttam Kumaran: So…

90 00:09:18.990 ⇒ 00:09:20.080 Amber Lin: Okay.

91 00:09:21.670 ⇒ 00:09:22.690 Amber Lin: Sounds good.

92 00:09:22.940 ⇒ 00:09:23.480 Uttam Kumaran: Okay.

93 00:09:24.120 ⇒ 00:09:28.430 Samuel Roberts: Yeah, and then Amber, if you can figure out if we own that dock or not, just let me know.

94 00:09:28.430 ⇒ 00:09:29.959 Amber Lin: I own that dog, because I created.

95 00:09:29.960 ⇒ 00:09:30.790 Samuel Roberts: You own the dog?

96 00:09:30.790 ⇒ 00:09:34.680 Amber Lin: And we can only transfer within our…

97 00:09:35.120 ⇒ 00:09:43.340 Amber Lin: our organization, so I can’t transfer to our, like, Rain Forge ABC account. But whoever needs that ownership, I’ll transfer to you.

98 00:09:43.340 ⇒ 00:09:46.410 Samuel Roberts: I don’t know if we need the ownership so much as we just need to make sure that the…

99 00:09:46.600 ⇒ 00:09:55.369 Samuel Roberts: the credentials that we have accessing it were fine. But we can sort that out later, we don’t have to worry about that now if you need to run. I just want to make sure we know if we had to get them to do something.

100 00:09:55.920 ⇒ 00:09:56.590 Amber Lin: Yep.

101 00:09:56.590 ⇒ 00:09:58.070 Samuel Roberts: Okay, I think we’re good then.

102 00:09:58.710 ⇒ 00:09:59.429 Amber Lin: Sounds good.

103 00:09:59.970 ⇒ 00:10:03.010 Samuel Roberts: Anything else that I didn’t get on ABC?

104 00:10:04.220 ⇒ 00:10:05.459 Samuel Roberts: From an AI side.

105 00:10:07.120 ⇒ 00:10:07.820 Samuel Roberts: Okay.

106 00:10:10.220 ⇒ 00:10:16.999 Samuel Roberts: Cool. Alright, so then, I guess we’ll jump to Lilo, and Amber, I guess you’re probably good then.

107 00:10:17.150 ⇒ 00:10:18.889 Amber Lin: Yeah, thanks everyone.

108 00:10:19.170 ⇒ 00:10:20.020 Uttam Kumaran: Thank you.

109 00:10:20.530 ⇒ 00:10:25.260 Samuel Roberts: So yeah, Lilo, looks like… Pranav, you got some…

110 00:10:25.710 ⇒ 00:10:28.659 Samuel Roberts: some of the front end at least moved over? Is that what…

111 00:10:28.950 ⇒ 00:10:29.490 Pranav: Yeah.

112 00:10:30.910 ⇒ 00:10:38.860 Pranav: From what I could get from the Stitch web app, I think I kind of translated most of, if not…

113 00:10:39.090 ⇒ 00:10:41.049 Pranav: all of that UI over.

114 00:10:41.050 ⇒ 00:10:41.540 Samuel Roberts: Okay.

115 00:10:41.540 ⇒ 00:10:54.049 Pranav: What I realized, though, is something has changed with the authentication system that was in, like, the old app. So before, we were able to, like, put in our email, get, like, a verification code, and then sign in.

116 00:10:54.100 ⇒ 00:11:02.960 Pranav: So, their authentication was working. Yesterday, when I was working on it, I wasn’t able to see that same behavior, and I wonder if, like, the old agency, like.

117 00:11:03.100 ⇒ 00:11:10.580 Pranav: like, they had some deployed app for authentication that they now removed. Not that that, like.

118 00:11:10.580 ⇒ 00:11:14.340 Samuel Roberts: This is for their app as it, like, the live one that they had.

119 00:11:14.900 ⇒ 00:11:15.250 Pranav: Yeah.

120 00:11:15.250 ⇒ 00:11:16.140 Samuel Roberts: Okay.

121 00:11:16.560 ⇒ 00:11:18.050 Pranav: Yeah, yup.

122 00:11:18.240 ⇒ 00:11:22.909 Pranav: So, because they had, like, a bunch of those legacy repos, I don’t know if that…

123 00:11:23.040 ⇒ 00:11:25.900 Pranav: They had that BNRZ platform?

124 00:11:25.900 ⇒ 00:11:31.880 Samuel Roberts: Yeah, that’s… that’s the… so the… that’s basically their… the previous agency’s, like.

125 00:11:32.780 ⇒ 00:11:42.490 Samuel Roberts: I don’t know if it’s internal to them, necessarily, but it’s got a lot of stuff in there. I think Surfside, too, there’s, like, crypto stuff in there. Like, I think it’s just a kind of catch-all thing that they’ve been building.

126 00:11:42.650 ⇒ 00:11:49.720 Samuel Roberts: And then when someone comes to them, they plug in and get a bunch of endpoints that they already have set up for things like chatbots and auth and…

127 00:11:49.860 ⇒ 00:11:57.179 Samuel Roberts: If you’re doing any crypto stuff, or… So, basically, I have some stuff that I’ve chatted with Cursor where I was trying to pull all the endpoints that go to that.

128 00:11:58.470 ⇒ 00:12:10.349 Samuel Roberts: And we can start moving that over. It looks like we’re not gonna get much from them, is that what I was seeing on the chat? So, I think we’ll just… we’ll just work from the code that we have.

129 00:12:10.500 ⇒ 00:12:23.070 Samuel Roberts: And so we’ll have to use that repo to figure out… I have a little… I’ll share some of the docs that I was getting in Cursor, like some Markdown that just, like, went through and found whichever endpoints go to that BRNZ

130 00:12:23.390 ⇒ 00:12:32.770 Samuel Roberts: live, and then I tried to map that into the app. So, we can… we can get together, and I can share that with you, and hopefully that’ll then show, like, the authentication.

131 00:12:32.880 ⇒ 00:12:37.850 Samuel Roberts: You know, we’re already gonna get rid of, like… because they had a whole other thing for…

132 00:12:38.820 ⇒ 00:12:40.290 Samuel Roberts: I forget what it was now.

133 00:12:40.970 ⇒ 00:12:46.880 Samuel Roberts: blanking on the word, but… yeah. No, no, they had a portal thing, I think.

134 00:12:46.880 ⇒ 00:12:47.540 Uttam Kumaran: Yeah, yeah, yeah.

135 00:12:47.540 ⇒ 00:12:50.869 Samuel Roberts: And so, like, they had different… because I think different people can off onto their same.

136 00:12:50.870 ⇒ 00:12:57.509 Uttam Kumaran: Yeah, I think just, like, fuck whatever these people… Yeah, like, you know, they’re just trying to copy-paste stuff from… I get what they’re trying to do, it’s like.

137 00:12:57.510 ⇒ 00:13:05.219 Samuel Roberts: Yeah, yeah. But we can just now, basically, all those endpoints, including off, we will now replace, and Better Off is already set up, so that should be good to…

138 00:13:05.470 ⇒ 00:13:08.050 Uttam Kumaran: Okay. To switch over pretty well, and then…

139 00:13:08.050 ⇒ 00:13:10.679 Samuel Roberts: all the other ones, it… from what I saw, there was, like.

140 00:13:10.840 ⇒ 00:13:17.030 Samuel Roberts: bots, which I think is just the agents, and then the, like, getters and setters for changing those prompts and things, but we can…

141 00:13:17.340 ⇒ 00:13:21.300 Samuel Roberts: we can figure out how to replace all that into the Next and Nest apps.

142 00:13:22.640 ⇒ 00:13:26.470 Pranav: So, like, that main, like, dashboard page, I kind of, like…

143 00:13:26.750 ⇒ 00:13:33.730 Pranav: followed, basically, the UI. I didn’t… I didn’t want to install, like, a bunch of, like, other, like, component libraries to make it, like, an exact kind of…

144 00:13:33.730 ⇒ 00:13:34.849 Samuel Roberts: Yeah, no.

145 00:13:34.850 ⇒ 00:13:36.019 Pranav: So, like, I’m just using the shadow.

146 00:13:36.020 ⇒ 00:13:45.070 Uttam Kumaran: I also… I also think, guys, like, there’s… they’re… they’re not married at all to the design or anything, so don’t, pull your hair out trying to match

147 00:13:45.200 ⇒ 00:13:49.479 Uttam Kumaran: these guys, like, what I’m learning, and this is where, like, we don’t know until we’re in the client, like.

148 00:13:49.480 ⇒ 00:13:50.100 Samuel Roberts: Yeah.

149 00:13:50.280 ⇒ 00:14:00.820 Uttam Kumaran: I didn’t know how married or not they are. We pick an assumption, and we go, and then we turn around. It’s clear that they just don’t care. So, fundamentally, like, I think…

150 00:14:01.250 ⇒ 00:14:09.260 Uttam Kumaran: what I want to prove to them in this first… in this first phase is that, like, we know what we’re doing, and, like, we’re… we’ve built an environment where we can develop.

151 00:14:09.310 ⇒ 00:14:22.380 Uttam Kumaran: Part of that is showing them that stuff is working and they can log in, but our alpha is gonna be on having a great chat experience, having a great, like, where we’re landing data somewhere to run models on top of.

152 00:14:23.660 ⇒ 00:14:39.040 Uttam Kumaran: that’s gonna be really, like, where the alpha is. So in this initial phase, whatever you can… whatever is helpful to speed up the process, do, but they don’t care at all about the designs or anything. They purely care about the functionality.

153 00:14:39.040 ⇒ 00:14:42.670 Samuel Roberts: I was gonna say, and the basic functionality was just they had chatbots over…

154 00:14:43.060 ⇒ 00:14:46.959 Samuel Roberts: different cli… different of their clients. And so you… you could have the different brands.

155 00:14:46.960 ⇒ 00:14:51.539 Uttam Kumaran: Yeah, and so, but also, you tell me, like, if you’re like, hey, I need, I need, like.

156 00:14:52.140 ⇒ 00:14:57.420 Uttam Kumaran: I need some type of front-end mock-up to be able to build towards, then I can go work on that.

157 00:14:57.540 ⇒ 00:15:04.069 Uttam Kumaran: You know… Ideally, just build… if we could just replicate whatever exists there, even if it’s ugly.

158 00:15:04.070 ⇒ 00:15:08.209 Samuel Roberts: Yeah. I’m gonna… we’re gonna build… we’re gonna rebuild a version of ours.

159 00:15:08.780 ⇒ 00:15:13.890 Uttam Kumaran: You know, and I’ll get design feedback and stuff from them. That’ll be, like, kind of part of that PRD, but…

160 00:15:14.420 ⇒ 00:15:15.569 Pranav: Yeah, I think…

161 00:15:15.670 ⇒ 00:15:35.599 Pranav: I don’t even need something that, like, detailed. I think what would be, like, super useful for me was just to see, like, what is their workflow for this app? So, like, when they click on, like, that… I think it’s, like, the new client button, and then they provide a URL, that then creates some card, and then it sounds like…

162 00:15:35.600 ⇒ 00:15:48.270 Pranav: what we’re talking about right now is, like, that… if you click on, like, that, I think… I don’t… I forget, like, the verbiage they use, but once you click on that card, it allows you to then, what, have a chatbot to…

163 00:15:48.380 ⇒ 00:15:50.359 Pranav: To ask specific questions on…

164 00:15:50.360 ⇒ 00:15:50.810 Uttam Kumaran: Yes.

165 00:15:50.810 ⇒ 00:15:54.170 Pranav: store, right? Okay. So, yeah, that already.

166 00:15:54.170 ⇒ 00:15:56.620 Uttam Kumaran: Tell me the format, like, would you prefer, like…

167 00:15:56.960 ⇒ 00:16:05.340 Uttam Kumaran: would you want that to be, like, in a… in, like, a fig jam, or can I put that directly in, like, the PRD, basically, for, like, here are the couple flows to enable?

168 00:16:05.880 ⇒ 00:16:11.429 Pranav: I think directly in the PRD, whatever’s case, because it seems, like, pretty, like, not complicated.

169 00:16:11.430 ⇒ 00:16:15.420 Uttam Kumaran: Okay, so then I’m just gonna make it really clear in there, like…

170 00:16:15.780 ⇒ 00:16:20.459 Uttam Kumaran: what the core flows are. So, one is it’s clear there’s also, like, a…

171 00:16:20.980 ⇒ 00:16:32.790 Uttam Kumaran: general platform admin PRD that I’m gonna work on, which is, like, creating a new client, like, initialization, logging in, and so, like, I’ll wrap all of that in there.

172 00:16:32.790 ⇒ 00:16:40.349 Samuel Roberts: User management and stuff like that. User management. They were using their admin panel right now for that thing, so we gotta pull that into here, too.

173 00:16:40.350 ⇒ 00:16:49.669 Uttam Kumaran: Okay, okay. So I’ll work on that, and then that way, it’s just super clear, the couple features to make sure, and then the second and third PRDs will build off of, like.

174 00:16:49.920 ⇒ 00:16:51.300 Uttam Kumaran: that functionality.

175 00:16:51.500 ⇒ 00:16:52.740 Uttam Kumaran: I can work on that.

176 00:16:53.060 ⇒ 00:16:53.710 Pranav: Perfect.

177 00:16:56.920 ⇒ 00:17:04.160 Samuel Roberts: Yeah, because we’re just kind of working with what we have from the code, but something’s in the code, and we’re not going to find out what was complete and what wasn’t, so we’ll have to…

178 00:17:04.670 ⇒ 00:17:09.340 Samuel Roberts: you know, if we get the PRD settled and everything, then I think we’ll be pretty good to just run with what we have.

179 00:17:09.660 ⇒ 00:17:17.720 Pranav: Yeah, and then also, just, on Railway, this morning, I was trying to just deploy the app on Railway,

180 00:17:18.349 ⇒ 00:17:27.849 Pranav: was having a little bit of trouble with that, Surf. I know, like, you were able to deploy the, like, the base repo on Railway, so I don’t see the…

181 00:17:28.440 ⇒ 00:17:39.300 Pranav: there to be that many different changes that I made that it would, like, create, like, a completely different way of deploying on Railway for, like, for Lilo. So, surf…

182 00:17:39.430 ⇒ 00:17:41.429 Pranav: If you could also, kind of.

183 00:17:41.810 ⇒ 00:17:50.279 Pranav: I forget, I know you demoed the… the app, like, using Railway, but I don’t remember if you created, like, a config file, like a Railway config.

184 00:17:51.700 ⇒ 00:17:56.099 Surf’s iPhone: Yeah, there’s no railway config, created.

185 00:17:56.230 ⇒ 00:17:59.330 Surf’s iPhone: Well, what’s the error that you’re having?

186 00:17:59.330 ⇒ 00:18:00.790 Samuel Roberts: Yeah, I was gonna ask something.

187 00:18:00.790 ⇒ 00:18:07.000 Pranav: Yeah, so I was able to get the build working, but for some reason it just crashes.

188 00:18:07.530 ⇒ 00:18:09.620 Surf’s iPhone: Well, which one? The front end or the back end?

189 00:18:09.620 ⇒ 00:18:10.140 Pranav: back.

190 00:18:10.710 ⇒ 00:18:13.029 Surf’s iPhone: The back end crashes, but the front end is good.

191 00:18:13.630 ⇒ 00:18:17.650 Pranav: I haven’t gotten to the front end yet, because the backend has just been crashing.

192 00:18:17.650 ⇒ 00:18:22.940 Surf’s iPhone: Alright, cool. Could you send me a picture of what the crash looks like? Because you should have, like, the report logs there, and I can help you with that.

193 00:18:23.370 ⇒ 00:18:24.100 Samuel Roberts: Yeah, yeah, yeah.

194 00:18:24.100 ⇒ 00:18:25.360 Surf’s iPhone: Not a problem.

195 00:18:26.010 ⇒ 00:18:26.540 Pranav: Cool.

196 00:18:28.190 ⇒ 00:18:28.840 Surf’s iPhone: Cool.

197 00:18:28.840 ⇒ 00:18:32.159 Uttam Kumaran: And I’m also basically free until 11.30, and…

198 00:18:32.580 ⇒ 00:18:38.219 Uttam Kumaran: Sam, we’ll probably just… we can just chill and work on ABC stuff, so… Perfect. I can just stay on this meeting, and we can work on stuff.

199 00:18:38.220 ⇒ 00:18:38.840 Samuel Roberts: Yeah.

200 00:18:38.990 ⇒ 00:18:48.100 Samuel Roberts: Yeah, we can… we can do a little co-working, that’s… that’s fine. Can you add… so you had the railway set up, did we set up, like, a Brainforge or a Lilo railway that we did?

201 00:18:49.010 ⇒ 00:18:50.830 Pranav: I don’t think we’ve done that yet.

202 00:18:50.830 ⇒ 00:18:53.360 Samuel Roberts: Okay, well, can you add me to the… is there a way to add me to that?

203 00:18:53.360 ⇒ 00:18:56.579 Surf’s iPhone: Oh, we… We do have a Lilo Railway, right?

204 00:18:56.580 ⇒ 00:18:57.460 Uttam Kumaran: Yeah, there is, yeah.

205 00:18:58.580 ⇒ 00:19:01.220 Uttam Kumaran: So, you can use the Stitch Gmail.

206 00:19:01.540 ⇒ 00:19:02.490 Surf’s iPhone: Sam?

207 00:19:02.960 ⇒ 00:19:05.929 Uttam Kumaran: Okay. So Stitch at Lilo Social.

208 00:19:06.530 ⇒ 00:19:08.739 Uttam Kumaran: And you can log into Railway. And so, yeah.

209 00:19:08.740 ⇒ 00:19:09.370 Samuel Roberts: I just want to keep it.

210 00:19:09.370 ⇒ 00:19:11.069 Uttam Kumaran: All on their stuff.

211 00:19:11.330 ⇒ 00:19:18.800 Samuel Roberts: Okay, okay, that’s, that’s the one. Stitch at Leo Social… Looks like… Oh, God, here.

212 00:19:22.800 ⇒ 00:19:31.850 Samuel Roberts: with GitHub? Is GitHub the only way to… off and railway? Oh, there’s his email, okay, Gmail, that’s it. Alright, I got it, okay, I’ll figure that out. I’ll get in there and take a look, too.

213 00:19:32.020 ⇒ 00:19:32.580 Uttam Kumaran: Okay.

214 00:19:32.850 ⇒ 00:19:34.250 Samuel Roberts: Cool.

215 00:19:34.550 ⇒ 00:19:36.169 Samuel Roberts: And then…

216 00:19:37.570 ⇒ 00:19:43.719 Samuel Roberts: Anything else to discuss legal-wise, or to start working on that? And then you said Remo’s the other thing to discuss here in this meeting, Tom?

217 00:19:43.720 ⇒ 00:19:44.300 Uttam Kumaran: Yeah.

218 00:19:45.680 ⇒ 00:19:49.039 Samuel Roberts: Okay, that I don’t have a ton of, context into.

219 00:19:49.250 ⇒ 00:19:54.939 Uttam Kumaran: So, sir, if I think kind of what I mentioned in the beginning of this call is, like, this is just gonna be the meeting. We’re just gonna break off.

220 00:19:55.240 ⇒ 00:20:01.159 Uttam Kumaran: this crew, like, which is, like, AI and, like, app development crew into this meeting, so I do want Sam to have a…

221 00:20:01.200 ⇒ 00:20:07.959 Uttam Kumaran: purview into Remo, and he’s gonna start to… he can start to become, like, a little bit of another layer

222 00:20:07.980 ⇒ 00:20:25.439 Uttam Kumaran: you know, in addition to Robert and I, I’m sort of furthest. Robert is closer. You’re the closest Sam. You’re the closest surf, so I do want to have Sam sort of overview over anything that’s… that we’re doing that is AI and app development. So, I think you may have to…

223 00:20:25.440 ⇒ 00:20:37.020 Uttam Kumaran: I think one is, like, Sam, I think we can brief you on, like, what the scope is for Remo, and, like, how Surf is executing. I don’t think there’s… I mean, I’ll let Surf

224 00:20:37.300 ⇒ 00:20:46.789 Uttam Kumaran: basically, in this meeting, we just want to start, like, are there resourcing needs? Yeah. And are there other blockers that you need to loop people in? And, like, that’s what we’re talking about today, so…

225 00:20:46.810 ⇒ 00:21:00.109 Uttam Kumaran: I’ll… I’ll end up adding… I’ll add Sam to the Eden OS Remo channel, and then Sam, you can take a look at, like, there’s… there’s docs there and things, stuff like that. Okay. But, yeah, maybe, Surf, if you want to give an update on how it’s going today.

226 00:21:00.110 ⇒ 00:21:00.710 Surf’s iPhone: Cool.

227 00:21:00.850 ⇒ 00:21:04.250 Surf’s iPhone: The high level is,

228 00:21:04.390 ⇒ 00:21:22.570 Surf’s iPhone: It’s going as well as it can be. More or less I’m handling it. We’re supposed to be getting some code from a developer, that they had contracted to do development. We’ve been at this with them for a while, so I’m at the point where I’m throwing my hands up of that.

229 00:21:22.570 ⇒ 00:21:33.380 Surf’s iPhone: So we’re gonna end up having to do, potentially, a rebuild of the entire infrastructure and architecture, thus using that same framework that we now have, which was, like, out of scope originally.

230 00:21:33.380 ⇒ 00:21:34.620 Samuel Roberts: No.

231 00:21:34.910 ⇒ 00:21:47.939 Surf’s iPhone: If it goes that path, which is a today call, they’re gonna take legal action against the other developer. So, ideally not the best start, but not Brainforge’s fault.

232 00:21:47.940 ⇒ 00:21:51.279 Samuel Roberts: Ideally.

233 00:21:51.990 ⇒ 00:22:09.140 Surf’s iPhone: We’re pretty sufficiently staffed. I have two engineers working under me. They brought on a product manager, another engineer starts next week, and then, they have a company out in India, that’s gonna be bringing on, like, a pod of, like, 10 engineers.

234 00:22:09.160 ⇒ 00:22:26.439 Surf’s iPhone: So it’s going basically from 0 to, like, 15 people in, like, the course of 2 weeks, so… there’s a lot. What I could do is potentially get you caught up on our next stand-up, Sam. That’ll probably be the easiest way to transfer, because there’s a lot of information overall.

235 00:22:26.440 ⇒ 00:22:30.820 Surf’s iPhone: We usually do them at, like, 10, AM Eastern.

236 00:22:30.870 ⇒ 00:22:42.469 Surf’s iPhone: Okay. And then you’ll get, like, a full brain dump from the whole team of everything, and then I think that’ll be the fastest way to get you running. It’s just a lot of context, I can manage, but it’s just really annoying, because…

237 00:22:42.470 ⇒ 00:22:43.020 Samuel Roberts: Yes.

238 00:22:43.260 ⇒ 00:22:54.039 Surf’s iPhone: The original scope was, we have code to run with, now the scope is looking like we don’t have code to run with, which messes up all of the timelines that were built originally. Right.

239 00:22:54.040 ⇒ 00:22:54.760 Samuel Roberts: Okay.

240 00:22:56.070 ⇒ 00:23:05.249 Surf’s iPhone: And then, me and Robert chatted a little bit earlier today, so Robert is going the path of telling them that the timeline might need to be adjusted, all that sort of stuff.

241 00:23:05.410 ⇒ 00:23:06.200 Samuel Roberts: Okay.

242 00:23:06.860 ⇒ 00:23:08.180 Surf’s iPhone: Cool.

243 00:23:08.180 ⇒ 00:23:10.780 Uttam Kumaran: I think this is an interesting one, Sam, in that, like.

244 00:23:11.300 ⇒ 00:23:15.089 Uttam Kumaran: Surf has the expertise to run, like, a project this size.

245 00:23:15.090 ⇒ 00:23:15.500 Samuel Roberts: Yeah.

246 00:23:15.500 ⇒ 00:23:18.869 Uttam Kumaran: But I do want us as, like, an AI app crew.

247 00:23:19.390 ⇒ 00:23:21.189 Uttam Kumaran: But know what everybody’s doing.

248 00:23:21.190 ⇒ 00:23:21.850 Samuel Roberts: Totally.

249 00:23:22.000 ⇒ 00:23:25.700 Uttam Kumaran: So, it’ll just be… I just want to make sure you have

250 00:23:26.010 ⇒ 00:23:38.089 Uttam Kumaran: One of the things that, you know, we’re listening to on the feedback side is just making sure everybody sort of, like, has… sees everything. Especially if you’re tasked with seeing everything, so that you do see everything. Totally. So, okay.

251 00:23:39.050 ⇒ 00:23:43.359 Samuel Roberts: Yeah, no, that sounds good. I mean, yeah, I’ll get some context if you add me… It’s also, like, look, if, like.

252 00:23:43.360 ⇒ 00:23:59.630 Uttam Kumaran: And on ABC, if we need service help, if we all, like… this is where, like, I want… part of the thing that Clarence mentioned is, like, we’re running a consultancy, it’s sort of like a product crew, and so that’s, like, the benefit of being, like, okay, we have, like, these… we have, like, these tenants, like, how are we supporting, you know?

253 00:23:59.630 ⇒ 00:24:00.190 Samuel Roberts: Yeah.

254 00:24:00.580 ⇒ 00:24:01.020 Uttam Kumaran: Cool.

255 00:24:01.020 ⇒ 00:24:02.120 Samuel Roberts: I think that makes sense.

256 00:24:02.230 ⇒ 00:24:06.530 Samuel Roberts: Okay, cool. Great.

257 00:24:06.870 ⇒ 00:24:10.289 Samuel Roberts: I’ll get caught up on that a little bit, at least some more context, but yeah, like I said, if there’s…

258 00:24:10.440 ⇒ 00:24:15.330 Samuel Roberts: blockers or things, we can surface them here, but, obviously this is out of our control, so…

259 00:24:15.330 ⇒ 00:24:22.730 Uttam Kumaran: Yeah, can we also maybe go through, Pranav, I don’t know if you had a chance to do the Gantt chart, but maybe we can, if not, we can also just, like.

260 00:24:23.440 ⇒ 00:24:29.319 Uttam Kumaran: maybe gen- generally chat about, like, at least, like, next 6 weeks for Lilo, and, like, talk through…

261 00:24:29.440 ⇒ 00:24:31.720 Uttam Kumaran: Like, how we feel about things.

262 00:24:32.120 ⇒ 00:24:38.430 Uttam Kumaran: You can even create the Gantt chart as we’re talking through it, but I just want to spend, like, 5-10 minutes there.

263 00:24:38.890 ⇒ 00:24:40.220 Pranav: Yeah, what do you think? That works.

264 00:24:41.260 ⇒ 00:24:43.380 Uttam Kumaran: Were you able to get into the Instagant?

265 00:24:44.110 ⇒ 00:24:46.179 Pranav: No, I was just,

266 00:24:46.530 ⇒ 00:24:55.319 Pranav: working on, kind of, the migration and then, the deployment. Okay. So, yeah, because I was having issues with that yesterday.

267 00:24:55.770 ⇒ 00:24:56.469 Uttam Kumaran: Try to log in.

268 00:24:56.470 ⇒ 00:24:56.970 Samuel Roberts: Yeah.

269 00:24:56.970 ⇒ 00:25:01.639 Uttam Kumaran: That way, we can just, like… I can even just… we can do a couple things live, and…

270 00:25:02.130 ⇒ 00:25:05.539 Uttam Kumaran: You should… I mean, there already is a Gantt chart.

271 00:25:05.800 ⇒ 00:25:08.390 Uttam Kumaran: So I, but…

272 00:25:08.580 ⇒ 00:25:14.220 Uttam Kumaran: I think this is just… we just copied over from, like, one of our existing ones, so we do have to, like, remove some of this.

273 00:25:14.960 ⇒ 00:25:18.620 Pranav: Gotcha, gotcha, okay. Let me find that thread real quick.

274 00:25:21.830 ⇒ 00:25:24.969 Uttam Kumaran: This is the link that I have, I sent it, I’ll send it in soon.

275 00:25:26.230 ⇒ 00:25:27.219 Pranav: Okay, perfect.

276 00:25:33.840 ⇒ 00:25:40.879 Pranav: Yeah, it’s the same issue, it’s weird, I just get, like, this… Okay, maybe it’s loading now.

277 00:25:51.290 ⇒ 00:25:56.760 Samuel Roberts: Yeah, I noticed that it was a little weird clicking around some of the projects, and the URL might not be changing and stuff, so I had some weird issues, but…

278 00:25:58.000 ⇒ 00:26:09.720 Uttam Kumaran: I’m gonna find a tool that has an API, so we do all this shit via AI. I don’t want fucking people sitting in Gantt charts, but it’s just… it’s super, super powerful to explain timeline-based.

279 00:26:09.720 ⇒ 00:26:11.110 Samuel Roberts: Oh, 100%, yeah.

280 00:26:11.500 ⇒ 00:26:13.000 Uttam Kumaran: So I just… whatever.

281 00:26:15.080 ⇒ 00:26:16.510 Pranav: Let me show you guys what I’m seeing.

282 00:26:16.510 ⇒ 00:26:17.310 Samuel Roberts: Yeah, yeah.

283 00:26:17.310 ⇒ 00:26:21.099 Uttam Kumaran: Okay, cool, and yeah, worst case, I’ll just… worst case, I can just kick you on, add you again.

284 00:26:21.540 ⇒ 00:26:26.350 Pranav: Yeah, that might be what we need to do at this point, because… It just gets.

285 00:26:26.350 ⇒ 00:26:26.680 Samuel Roberts: Oh.

286 00:26:28.630 ⇒ 00:26:30.770 Uttam Kumaran: Okay.

287 00:26:31.380 ⇒ 00:26:33.490 Samuel Roberts: This is just loading. Can you hit,

288 00:26:34.120 ⇒ 00:26:35.949 Samuel Roberts: You can see the dev… bring up the dev tools.

289 00:26:36.780 ⇒ 00:26:38.859 Samuel Roberts: Hit, is it F12?

290 00:26:42.720 ⇒ 00:26:45.080 Pranav: Yeah.

291 00:26:47.690 ⇒ 00:26:49.090 Pranav: I don’t know, everything’s just frozen.

292 00:26:49.090 ⇒ 00:26:52.440 Samuel Roberts: I was gonna say, Chrome might just be, yeah, who knows, or yeah, I’m assuming it’s Chrome, but…

293 00:26:52.680 ⇒ 00:26:54.010 Pranav: Yeah, yeah, it is Chrome.

294 00:26:54.010 ⇒ 00:26:56.620 Samuel Roberts: Okay, yeah, sometimes… Yeah.

295 00:26:58.460 ⇒ 00:26:59.460 Pranav: Let’s…

296 00:27:05.770 ⇒ 00:27:09.500 Uttam Kumaran: Okay, I’m gonna just… I removed you, and I’m gonna add you again.

297 00:27:13.370 ⇒ 00:27:16.289 Samuel Roberts: Go to, the console or network, maybe, real quick?

298 00:27:18.200 ⇒ 00:27:21.760 Samuel Roberts: Yeah, go back to the console, there was something there.

299 00:27:22.400 ⇒ 00:27:25.700 Samuel Roberts: Mmm… This is jQuery, okay.

300 00:27:26.320 ⇒ 00:27:30.399 Samuel Roberts: Now, did you, did you, like, log in and everything? Was it…

301 00:27:30.750 ⇒ 00:27:32.760 Samuel Roberts: Were you in Instagan at any point?

302 00:27:32.890 ⇒ 00:27:35.499 Pranav: Oh, no, no, I was never able to get in.

303 00:27:35.630 ⇒ 00:27:36.239 Pranav: Which is…

304 00:27:36.240 ⇒ 00:27:37.280 Samuel Roberts: Oh, interesting.

305 00:27:37.280 ⇒ 00:27:37.830 Pranav: go.

306 00:27:38.580 ⇒ 00:27:40.489 Uttam Kumaran: Okay, check now.

307 00:27:40.490 ⇒ 00:27:42.160 Samuel Roberts: Yeah, yeah.

308 00:27:42.160 ⇒ 00:27:44.229 Uttam Kumaran: guided you again. See if you got it in right.

309 00:27:47.640 ⇒ 00:27:49.599 Samuel Roberts: You’ve been signed off, yeah, it’s long, huh?

310 00:27:54.700 ⇒ 00:27:55.264 Samuel Roberts: So.

311 00:27:58.770 ⇒ 00:28:04.520 Samuel Roberts: Yeah, because even when I go, like, incognito, where I’m not logged in, it just says you’ve been logged out. It’s definitely working.

312 00:28:05.430 ⇒ 00:28:12.709 Uttam Kumaran: I invested a lot of time on the Gantt chart capabilities, but nothing on, like, user management, which I respect, so…

313 00:28:12.710 ⇒ 00:28:14.939 Samuel Roberts: Yeah, it was a little weird in there what the user stole, but…

314 00:28:15.860 ⇒ 00:28:18.289 Uttam Kumaran: They just went super ham on the Gantt chart.

315 00:28:18.290 ⇒ 00:28:21.840 Samuel Roberts: Well, they’re also tied to Asana, so they’re probably leaning on people that are just using Asana.

316 00:28:21.840 ⇒ 00:28:23.390 Uttam Kumaran: Oh, yeah, yeah, yeah.

317 00:28:27.260 ⇒ 00:28:28.580 Pranav: Yeah, same thing.

318 00:28:29.580 ⇒ 00:28:31.220 Pranav: Yeah, maybe I can…

319 00:28:33.080 ⇒ 00:28:36.010 Uttam Kumaran: This is definitely your fault, dude, it’s not my fault.

320 00:28:37.770 ⇒ 00:28:41.559 Uttam Kumaran: Hey, try, try, try Safari, or, or…

321 00:28:41.560 ⇒ 00:28:45.389 Samuel Roberts: Yeah, I was gonna say, is there… are there other, like, internet connectivity issues you’ve had at all?

322 00:28:45.900 ⇒ 00:28:49.039 Samuel Roberts: Yes, I was gonna say, maybe it could be that. Okay.

323 00:28:49.630 ⇒ 00:28:52.929 Pranav: But it’s weird that it’s happened… like, it’s usually not…

324 00:28:53.190 ⇒ 00:28:55.209 Samuel Roberts: Yeah, but who knows what, like…

325 00:28:55.210 ⇒ 00:28:56.180 Pranav: If it’s…

326 00:28:56.500 ⇒ 00:28:59.280 Uttam Kumaran: Try Safari, or if you’re using Adblock…

327 00:29:00.830 ⇒ 00:29:01.359 Pranav: Yeah, yeah.

328 00:29:01.360 ⇒ 00:29:02.500 Samuel Roberts: Oh, yeah.

329 00:29:02.910 ⇒ 00:29:05.070 Samuel Roberts: Yeah, oh my goodness, okay.

330 00:29:05.530 ⇒ 00:29:06.200 Pranav: Yeah.

331 00:29:10.060 ⇒ 00:29:13.659 Samuel Roberts: Can you just Google Instagant and, like, click a link and see if it takes you, like, the right…

332 00:29:14.040 ⇒ 00:29:16.009 Samuel Roberts: Into, like, the homepage, at least, maybe?

333 00:29:16.480 ⇒ 00:29:21.710 Samuel Roberts: Maybe their app is borked or something with the way it’s loading something, but…

334 00:29:30.360 ⇒ 00:29:35.200 Pranav: Yeah, so it tries to actually, like, auto-sign me in, so there’s some cash.

335 00:29:35.200 ⇒ 00:29:36.450 Samuel Roberts: Oh, okay.

336 00:29:36.600 ⇒ 00:29:37.680 Samuel Roberts: Okay.

337 00:29:38.050 ⇒ 00:29:43.279 Uttam Kumaran: Oh, then I… what I would do is, if you… you can go into DevTools and just delete all the cookies.

338 00:29:43.280 ⇒ 00:29:45.839 Samuel Roberts: Yeah, just clear everything for the site, totally.

339 00:29:46.110 ⇒ 00:29:49.110 Uttam Kumaran: Delete, go under local, and you can just delete the cookies, yeah.

340 00:29:49.330 ⇒ 00:29:53.489 Pranav: Right, right. I remember trying all these things. Let’s see if it works.

341 00:29:54.280 ⇒ 00:29:55.470 Uttam Kumaran: Try Safari.

342 00:29:55.870 ⇒ 00:29:56.550 Pranav: Yeah.

343 00:29:56.550 ⇒ 00:29:57.650 Uttam Kumaran: Clause her eyes, I don’t know.

344 00:29:58.050 ⇒ 00:29:59.870 Pranav: Let me try something else this time.

345 00:30:30.730 ⇒ 00:30:36.509 Samuel Roberts: Well, I guess, actually, now that I’m thinking about it, before we even get into Leela, was there internal AI stuff to discuss, Gabe?

346 00:30:37.120 ⇒ 00:30:40.299 Samuel Roberts: Hadn’t… hadn’t included that in the list of things.

347 00:30:40.300 ⇒ 00:30:46.239 Gabriel Lam: We’re good. I think… so I’m working on a couple things this week, so I…

348 00:30:46.510 ⇒ 00:30:51.370 Gabriel Lam: Me… there’s… okay, sorry, I’m gonna take a step back, because I was heads down earlier.

349 00:30:51.370 ⇒ 00:30:55.279 Samuel Roberts: Yeah, so no, I didn’t mean to just, like, catch you off guard either, I was just realizing as I was looking at you.

350 00:30:55.390 ⇒ 00:31:08.810 Gabriel Lam: There’s a couple things that I want to do a spike on or explore before we push on. First thing being, how do we manage linear tickets in… via cursor, and what that.

351 00:31:09.210 ⇒ 00:31:09.890 Uttam Kumaran: Oh, yeah.

352 00:31:09.890 ⇒ 00:31:10.320 Gabriel Lam: like.

353 00:31:11.240 ⇒ 00:31:17.349 Gabriel Lam: So hopefully that gets us… takes, like, our load off of doing it in the platform.

354 00:31:17.350 ⇒ 00:31:17.780 Samuel Roberts: Okay.

355 00:31:17.780 ⇒ 00:31:25.739 Gabriel Lam: So I want to get that out of the way, and then I’m running into blockers, on this contextual AI partnership,

356 00:31:25.980 ⇒ 00:31:30.810 Gabriel Lam: There’s two main blockers, the first one being… having…

357 00:31:31.450 ⇒ 00:31:39.500 Gabriel Lam: adequate documentation that we can reference, because it’s a rag, and when I’m testing it, it’s like, hey, I don’t have the information for that, so that’s something that I’m…

358 00:31:39.840 ⇒ 00:31:41.260 Gabriel Lam: I reached out…

359 00:31:41.260 ⇒ 00:31:44.069 Uttam Kumaran: Did you end up calling, yeah, did you end up calling any of their team?

360 00:31:44.510 ⇒ 00:31:45.940 Gabriel Lam: I’m happy because they’re…

361 00:31:45.970 ⇒ 00:31:47.280 Uttam Kumaran: Yeah, go ahead.

362 00:31:47.280 ⇒ 00:31:51.090 Gabriel Lam: I wanted to get all the questions and details

363 00:31:51.390 ⇒ 00:31:57.219 Gabriel Lam: that I needed from them out, instead of going back and forth and be like, hey, I need this, actually, no, I need that.

364 00:31:57.550 ⇒ 00:31:58.400 Gabriel Lam: The product?

365 00:31:58.400 ⇒ 00:32:04.070 Uttam Kumaran: Static is complicated, dude, so that’s why, like, I feel like you may just want to call them.

366 00:32:04.070 ⇒ 00:32:04.810 Gabriel Lam: Yeah, I think.

367 00:32:04.810 ⇒ 00:32:05.330 Uttam Kumaran: the main thing?

368 00:32:05.330 ⇒ 00:32:10.930 Gabriel Lam: that I will be calling them for is, like, model parameters and model configuration.

369 00:32:10.930 ⇒ 00:32:11.320 Uttam Kumaran: Yeah.

370 00:32:11.320 ⇒ 00:32:14.719 Gabriel Lam: Because they’re doing, like, retrieval, re-ranking, parsing, that part.

371 00:32:15.540 ⇒ 00:32:22.309 Uttam Kumaran: I just don’t want you to, like, think that part of some of it is possible, and then they’re like, oh, actually, you can’t do this in the tool.

372 00:32:22.310 ⇒ 00:32:22.890 Gabriel Lam: Yep.

373 00:32:24.350 ⇒ 00:32:32.809 Uttam Kumaran: So I don’t know, like, think of this, like, our relationship with Contextual is not… is gonna be much bigger beyond just this, like, thing we’re building.

374 00:32:33.250 ⇒ 00:32:40.610 Uttam Kumaran: So, like, I would… I would just, like… and they’re, like, they’re… it’s, like, a very, like, fast-growing software startup, so…

375 00:32:40.630 ⇒ 00:32:44.119 Gabriel Lam: They’re gonna… they’re super down to talk to, like, folks like us.

376 00:32:44.120 ⇒ 00:32:51.059 Uttam Kumaran: Like, and so on the relationship side, like, if I’m putting my sales hat on, I would be like, just call them, just build a relationship, yeah.

377 00:32:54.400 ⇒ 00:32:56.500 Gabriel Lam: I’ll go ahead and set that up.

378 00:32:56.500 ⇒ 00:33:09.370 Uttam Kumaran: And then, yeah, like, it would be very helpful if we could figure out, how to use linear, via cursor. In particular, though, Gabe, it would be great if, in the playbooks, you put an agent’s… put, like, a cursor.

379 00:33:09.830 ⇒ 00:33:19.320 Uttam Kumaran: like, how to use cursor, or how to use linear from cursor MD file, so that I can just point my cursor to that and say, like, look into this.

380 00:33:20.190 ⇒ 00:33:22.260 Gabriel Lam: Yeah, as, like, others walk through.

381 00:33:22.430 ⇒ 00:33:23.230 Uttam Kumaran: See you, sir.

382 00:33:23.860 ⇒ 00:33:27.270 Uttam Kumaran: Yeah, look into this as, like, how to…

383 00:33:27.920 ⇒ 00:33:31.559 Uttam Kumaran: rip linear tickets, and then what you can also do is you can upload… update, like.

384 00:33:31.910 ⇒ 00:33:33.889 Uttam Kumaran: My meeting workflows, Doc.

385 00:33:34.200 ⇒ 00:33:41.479 Uttam Kumaran: So, as part of the meeting workflows, you can tell the AI to take the meeting transcript, and then go make updates into

386 00:33:41.820 ⇒ 00:33:43.190 Uttam Kumaran: linear directly.

387 00:33:43.410 ⇒ 00:33:48.910 Gabriel Lam: Yeah, I have a question about meeting transcripts, because we are already saving all of them. Is there a way to…

388 00:33:49.960 ⇒ 00:33:51.970 Samuel Roberts: There is certainly a way.

389 00:33:51.970 ⇒ 00:33:53.730 Gabriel Lam: Like, I’m thinking about standouts.

390 00:33:53.730 ⇒ 00:33:55.160 Uttam Kumaran: There’s definitely a way.

391 00:33:55.160 ⇒ 00:33:57.819 Gabriel Lam: Do we have to parse it again, or just…

392 00:33:57.820 ⇒ 00:33:58.590 Uttam Kumaran: No, so, so…

393 00:33:58.590 ⇒ 00:33:59.080 Gabriel Lam: Find that out.

394 00:33:59.080 ⇒ 00:34:01.950 Uttam Kumaran: We’ll work on something longer term where…

395 00:34:02.070 ⇒ 00:34:13.130 Uttam Kumaran: transcripts maybe get committed to relevant repos. Problem is, we don’t own every repo, and so, like, I’m not as… I’m not as concerned now, because…

396 00:34:13.790 ⇒ 00:34:23.259 Uttam Kumaran: Yeah, I think it’s, like, easy… it’s, like, easier to do, but it’s, like, I don’t think it’s the biggest problem. The… more of the biggest problem right now is… is scaling, like.

397 00:34:23.510 ⇒ 00:34:28.399 Uttam Kumaran: Getting everybody just to write in cursor, and do more and more in cursor.

398 00:34:28.929 ⇒ 00:34:33.330 Uttam Kumaran: And then naturally, we’ll see, like, what other pain points. I think that getting the transcripts in there.

399 00:34:33.909 ⇒ 00:34:38.990 Uttam Kumaran: Like, is one thing, and we can make easier in a few ways on the platform,

400 00:34:40.090 ⇒ 00:34:44.759 Uttam Kumaran: like, one thing we can do is we can just have an internal GitHub where all the transcripts get committed.

401 00:34:46.550 ⇒ 00:34:47.989 Uttam Kumaran: So, but it’s like…

402 00:34:48.170 ⇒ 00:34:54.670 Uttam Kumaran: weekend project, or, like, when… probably, like, next week, or the week after, we can try to take it.

403 00:34:54.679 ⇒ 00:34:57.409 Samuel Roberts: So some of the… some of the repos are owned by other orgs, right?

404 00:34:57.700 ⇒ 00:35:00.620 Uttam Kumaran: Yeah, so I can’t… so we can’t, like… yeah, we can’t… and…

405 00:35:00.830 ⇒ 00:35:12.559 Uttam Kumaran: I just really want to avoid the mistake of moving other transcripts to places that other clients are. Yeah. So what would be easiest is, like, we can create one repo.

406 00:35:12.730 ⇒ 00:35:15.150 Uttam Kumaran: With potentially every transcript.

407 00:35:15.460 ⇒ 00:35:19.060 Uttam Kumaran: And that just gets… that’s just part of the flow that gets committed.

408 00:35:19.100 ⇒ 00:35:23.660 Samuel Roberts: Yeah. And… You’re already associating it with clients, so just…

409 00:35:24.320 ⇒ 00:35:26.210 Uttam Kumaran: Make a folder for every client.

410 00:35:26.420 ⇒ 00:35:31.690 Uttam Kumaran: And then that way, as a user, I can have that transcripts folder, And…

411 00:35:32.860 ⇒ 00:35:41.949 Uttam Kumaran: like, I don’t necessarily need to put the transcripts in the client GitHub. I just need the transcripts to write the meeting notes that needs to get in the client GitHub.

412 00:35:42.180 ⇒ 00:35:50.790 Uttam Kumaran: You know, so I could end up removing the transcripts from each thing, because nobody ever uses… they’re looking at the derivative documents that I’m producing.

413 00:35:51.250 ⇒ 00:35:52.120 Samuel Roberts: Right, right.

414 00:35:52.120 ⇒ 00:35:52.650 Uttam Kumaran: Yeah.

415 00:35:55.170 ⇒ 00:35:57.189 Samuel Roberts: And that step for the meeting notes.

416 00:35:57.440 ⇒ 00:36:02.770 Samuel Roberts: you’re taking the transcript, passing that into cursor, With some other contacts?

417 00:36:03.910 ⇒ 00:36:10.020 Uttam Kumaran: I’m putting the meeting notes and… I’m putting the transcripts into a folder, and then there’s a work… there’s a meeting workflow markdown

418 00:36:10.130 ⇒ 00:36:14.879 Uttam Kumaran: That explains to the AI what to do, which is, like, look through the transcript.

419 00:36:14.910 ⇒ 00:36:17.579 Samuel Roberts: If there, if there is a relevant meeting notes.

420 00:36:17.580 ⇒ 00:36:19.370 Uttam Kumaran: Go update it with, like.

421 00:36:19.740 ⇒ 00:36:20.480 Samuel Roberts: Hmm. Like, okay.

422 00:36:20.480 ⇒ 00:36:25.230 Uttam Kumaran: with, like, what happened in that meeting. Like, create a new section with, like, here’s how we answered it.

423 00:36:25.400 ⇒ 00:36:31.760 Uttam Kumaran: That way, meeting notes always get created, but no one ever updates them, right? So this is, like, solving that problem.

424 00:36:31.880 ⇒ 00:36:37.530 Uttam Kumaran: Right. And so we have not only the notes we took, but the outcomes of the meeting, and then…

425 00:36:37.730 ⇒ 00:36:41.249 Uttam Kumaran: there’s, like, that’s the flow. So as I add a new transcript, it’ll do that, so…

426 00:36:41.580 ⇒ 00:36:48.509 Uttam Kumaran: I’m sure we can get… I have to just spend more time in Kirscher thinking about the ergonomics, but yeah, that’s basically what it is.

427 00:36:48.790 ⇒ 00:36:50.080 Samuel Roberts: Okay, that makes sense.

428 00:36:50.900 ⇒ 00:36:55.939 Gabriel Lam: Yeah, it’s like either… either you just have a sort of time… Timestamped record, or…

429 00:36:56.120 ⇒ 00:37:00.640 Gabriel Lam: you have it update. You have, like, a master that’s live.

430 00:37:01.140 ⇒ 00:37:04.380 Gabriel Lam: And, you know, every day it gets saved.

431 00:37:05.310 ⇒ 00:37:05.890 Uttam Kumaran: Yeah.

432 00:37:06.680 ⇒ 00:37:09.970 Uttam Kumaran: No, I think if… I think if you just shove that into GitHub.

433 00:37:10.290 ⇒ 00:37:18.050 Uttam Kumaran: it’ll live in people’s… it could live in people’s desktop where Cursor lives, right? Because Cursor is not… if you put it in a data lake or something, it’s gonna be hard to pull down.

434 00:37:22.030 ⇒ 00:37:25.960 Samuel Roberts: Yeah, I’m wondering, I mean, so is that something you run manually, then, right now?

435 00:37:26.070 ⇒ 00:37:27.809 Samuel Roberts: When the new meeting comes in? Okay.

436 00:37:28.000 ⇒ 00:37:28.960 Uttam Kumaran: I’m confident.

437 00:37:29.520 ⇒ 00:37:33.990 Uttam Kumaran: Into the folder, and then I basically am like, we just finished this meeting that I…

438 00:37:34.270 ⇒ 00:37:37.850 Uttam Kumaran: that I produce notes for, go and update that. And then usually I’m like.

439 00:37:38.150 ⇒ 00:37:42.280 Uttam Kumaran: I… I mean, like, as a client progresses.

440 00:37:42.460 ⇒ 00:37:46.550 Uttam Kumaran: it becomes so easy to use AI to, like, write the next document.

441 00:37:46.550 ⇒ 00:37:46.980 Samuel Roberts: Yeah.

442 00:37:46.980 ⇒ 00:37:49.859 Uttam Kumaran: Versus, like, a client like Default, where I don’t have anything saved.

443 00:37:50.000 ⇒ 00:37:52.979 Samuel Roberts: It’s really, like, it’s not getting it right.

444 00:37:53.160 ⇒ 00:37:54.560 Uttam Kumaran: You know…

445 00:37:54.560 ⇒ 00:37:55.050 Samuel Roberts: Sure.

446 00:37:55.050 ⇒ 00:38:00.879 Uttam Kumaran: like, the reason the CTA thing got it right is because I have tons of contacts in that repo, you know?

447 00:38:02.290 ⇒ 00:38:02.950 Samuel Roberts: Okay.

448 00:38:04.670 ⇒ 00:38:06.199 Uttam Kumaran: For now, did this thing work, or no?

449 00:38:06.900 ⇒ 00:38:08.000 Uttam Kumaran: No. Okay.

450 00:38:08.000 ⇒ 00:38:08.860 Samuel Roberts: Not, alright.

451 00:38:08.860 ⇒ 00:38:12.400 Pranav: Okay. I don’t know why this is the most complex thing.

452 00:38:12.840 ⇒ 00:38:16.529 Samuel Roberts: So weird. So this is in Safari, too, it’s just… oh.

453 00:38:16.530 ⇒ 00:38:18.230 Pranav: Same thing in Safari, yeah, yeah.

454 00:38:19.070 ⇒ 00:38:20.680 Pranav: Oh my gosh, I don’t know.

455 00:38:20.680 ⇒ 00:38:22.629 Uttam Kumaran: I’m sharing. I thought, I was like, what the hell, it looks like it’s.

456 00:38:22.630 ⇒ 00:38:23.230 Pranav: No.

457 00:38:23.230 ⇒ 00:38:24.420 Samuel Roberts: I got it now.

458 00:38:24.420 ⇒ 00:38:33.890 Uttam Kumaran: So I guess let’s talk about, like, where we are now. So, so, we have,

459 00:38:34.690 ⇒ 00:38:42.640 Uttam Kumaran: I guess, explain to me, like, so this is just gonna be, like, discovery… Plus…

460 00:38:43.730 ⇒ 00:38:46.669 Uttam Kumaran: Like, Phase 1, and then let’s just call this, like…

461 00:38:46.890 ⇒ 00:38:55.270 Uttam Kumaran: Phase two. So the first thing, we had kickoff, we basically had, like, what are some of the other, like, discovery-related things that we did?

462 00:38:56.210 ⇒ 00:39:00.190 Samuel Roberts: We did, like, an initial investigation into the code, like…

463 00:39:00.190 ⇒ 00:39:01.519 Pranav: Never see results.

464 00:39:01.520 ⇒ 00:39:02.310 Samuel Roberts: Yeah.

465 00:39:02.420 ⇒ 00:39:04.350 Samuel Roberts: Found that other branch that was…

466 00:39:05.640 ⇒ 00:39:06.930 Uttam Kumaran: Boat acquisition…

467 00:39:06.930 ⇒ 00:39:07.750 Samuel Roberts: Yeah.

468 00:39:07.750 ⇒ 00:39:18.020 Uttam Kumaran: like… Code… Legacy Code. Discovery… Okay…

469 00:39:18.060 ⇒ 00:39:18.720 Samuel Roberts: Hmm.

470 00:39:22.750 ⇒ 00:39:25.330 Samuel Roberts: Don’t worry about the timelines for now. Yeah.

471 00:39:26.570 ⇒ 00:39:30.130 Uttam Kumaran: The legacy code discovery, we had,

472 00:39:31.060 ⇒ 00:39:34.220 Uttam Kumaran: Railway, setup, what else do we do?

473 00:39:35.690 ⇒ 00:39:42.610 Pranav: I don’t know if we want to include this, but the work that Surf did for setting up, like, the base repo?

474 00:39:42.810 ⇒ 00:39:47.160 Uttam Kumaran: Oh, yeah, yeah, yeah, okay. So, like, initial… Yeah.

475 00:39:47.680 ⇒ 00:39:48.909 Pranav: Like, software design.

476 00:39:49.120 ⇒ 00:39:49.920 Uttam Kumaran: Yeah.

477 00:39:52.620 ⇒ 00:39:53.700 Uttam Kumaran: Yeah.

478 00:39:55.220 ⇒ 00:39:56.149 Samuel Roberts: Yeah, that’s fine.

479 00:40:02.010 ⇒ 00:40:13.739 Uttam Kumaran: Let’s say, like, post, like, Simple hosting plus off.

480 00:40:15.460 ⇒ 00:40:16.080 Pranav: Yeah.

481 00:40:17.570 ⇒ 00:40:19.179 Uttam Kumaran: What else?

482 00:40:22.970 ⇒ 00:40:28.090 Samuel Roberts: I mean, so now we’ve moved some of the front end, right, into Next.js, right? So…

483 00:40:28.320 ⇒ 00:40:29.050 Pranav: Yeah.

484 00:40:30.370 ⇒ 00:40:38.230 Samuel Roberts: Is that… is that… where are we in phases here? I’m not sure, like… We’re in phase one. You want to keep going phase… okay. Yeah, so then Phase 1’s still the…

485 00:40:38.340 ⇒ 00:40:47.570 Samuel Roberts: like, initial migration of, like, the features they had. So that would be the auth front end. Well, we have the auth, but the simple hosting, yeah, the…

486 00:40:48.000 ⇒ 00:40:48.960 Samuel Roberts: bleh.

487 00:40:50.140 ⇒ 00:40:53.149 Samuel Roberts: moving the front end to React,

488 00:40:53.360 ⇒ 00:40:56.299 Samuel Roberts: Or, however you want to do it. Yeah.

489 00:40:58.510 ⇒ 00:41:01.570 Samuel Roberts: And then, yeah, exactly.

490 00:41:12.020 ⇒ 00:41:18.800 Samuel Roberts: I mean, there’s the… we can break that down a little bit, because some of the BE stuff is going to be pulling from their old repo, but we can…

491 00:41:18.800 ⇒ 00:41:23.239 Uttam Kumaran: Don’t worry about being verbose, let’s just get everything here, and I’ll consolidate.

492 00:41:23.240 ⇒ 00:41:24.309 Samuel Roberts: Got it, okay.

493 00:41:25.250 ⇒ 00:41:31.889 Samuel Roberts: Yeah, I mean, those two things, ideally, will make a complete, like, version of what they had.

494 00:41:31.890 ⇒ 00:41:32.790 Uttam Kumaran: Okay.

495 00:41:33.440 ⇒ 00:41:37.670 Samuel Roberts: I’m trying to think of anything that’s not covered by those two.

496 00:41:37.670 ⇒ 00:41:40.860 Uttam Kumaran: Complete with previous version.

497 00:41:41.080 ⇒ 00:41:41.760 Samuel Roberts: Yeah.

498 00:41:41.890 ⇒ 00:41:45.140 Uttam Kumaran: And this is more of, like, a… milestone.

499 00:41:45.140 ⇒ 00:41:45.800 Samuel Roberts: Yes.

500 00:41:47.140 ⇒ 00:41:47.710 Uttam Kumaran: Great.

501 00:41:49.560 ⇒ 00:41:50.420 Uttam Kumaran: Okay.

502 00:41:50.530 ⇒ 00:41:57.380 Uttam Kumaran: And then, as part of Phase 1, we want to do, like, create client…

503 00:41:57.380 ⇒ 00:41:57.960 Samuel Roberts: Excuse me.

504 00:41:58.820 ⇒ 00:42:00.040 Uttam Kumaran: Functionality.

505 00:42:00.640 ⇒ 00:42:01.420 Uttam Kumaran: Right?

506 00:42:02.770 ⇒ 00:42:09.410 Uttam Kumaran: We want to do, user… or, like, what is it, like, Basic admin panel.

507 00:42:09.910 ⇒ 00:42:11.389 Uttam Kumaran: Yeah, I’ll do admin.

508 00:42:11.390 ⇒ 00:42:12.530 Samuel Roberts: Yeah, that’s fine.

509 00:42:15.260 ⇒ 00:42:17.100 Samuel Roberts: Yeah, because I would be, like, off, and…

510 00:42:17.280 ⇒ 00:42:19.310 Samuel Roberts: Oh, the… not just the auth, but the.

511 00:42:19.680 ⇒ 00:42:20.070 Uttam Kumaran: Yeah.

512 00:42:20.070 ⇒ 00:42:26.349 Samuel Roberts: Access, user management, Client management, brand management, all that stuff’s covered in there, okay.

513 00:42:31.090 ⇒ 00:42:32.240 Samuel Roberts: Excuse me.

514 00:42:34.460 ⇒ 00:42:35.190 Samuel Roberts: Yeah.

515 00:42:36.440 ⇒ 00:42:39.449 Uttam Kumaran: Are we gonna do anything on security? Like, anything else?

516 00:42:42.620 ⇒ 00:42:47.160 Uttam Kumaran: Or on caching, or… I guess it’s all internal, doesn’t really matter.

517 00:42:47.620 ⇒ 00:42:56.019 Samuel Roberts: Yeah, that’s the thing, when they were saying, like, it’s scalable and fast, and I was like, there’s not a ton to worry about there yet, or even based on their numbers.

518 00:42:56.150 ⇒ 00:43:00.120 Samuel Roberts: I wouldn’t worry too much about that. The only thing I do want to worry about security-wise is…

519 00:43:00.240 ⇒ 00:43:04.310 Samuel Roberts: Like, best practices around storing the keys that they’re gonna need to store?

520 00:43:05.050 ⇒ 00:43:10.379 Samuel Roberts: So, like, you know, we’ll have environment variables for the app, but when they’re adding brands and stuff.

521 00:43:12.940 ⇒ 00:43:21.990 Samuel Roberts: Okay. We gotta make sure to keep those keys secure. But that’s just a… that’s, like, the one thing that I’m, you know, if we’re doing Google Auth, I’m not worried about the rest of it. BetterOff’s gonna handle all that.

522 00:43:21.990 ⇒ 00:43:22.560 Uttam Kumaran: Okay.

523 00:43:22.840 ⇒ 00:43:23.950 Samuel Roberts: But making sure that

524 00:43:24.920 ⇒ 00:43:29.939 Samuel Roberts: Yeah, because our app will have its own invars and file, but as they’re adding other…

525 00:43:30.390 ⇒ 00:43:34.540 Samuel Roberts: things. We need to make sure to… those are… that’s the most sensitive stuff, I think, so…

526 00:43:36.150 ⇒ 00:43:37.279 Samuel Roberts: Excuse me, sorry.

527 00:43:39.260 ⇒ 00:43:41.540 Uttam Kumaran: Told this to turn off, like…

528 00:43:41.670 ⇒ 00:43:44.979 Uttam Kumaran: the pop-ups, and now they’re… I’m not able to delete shit.

529 00:43:45.970 ⇒ 00:43:47.019 Samuel Roberts: Turn off the pop-up?

530 00:43:47.470 ⇒ 00:43:49.129 Samuel Roberts: Oh, like, action from the page?

531 00:43:49.470 ⇒ 00:43:52.169 Uttam Kumaran: Yeah, I said don’t allow pop-ups anymore.

532 00:43:52.170 ⇒ 00:43:53.810 Samuel Roberts: Oh, my God.

533 00:43:53.810 ⇒ 00:43:56.629 Uttam Kumaran: God, this fucking sucks. Fuck.

534 00:43:56.980 ⇒ 00:44:02.040 Samuel Roberts: Is this, is this, like… the best Gantt thing we’ve found so far? Yeah. It’s just… okay.

535 00:44:02.040 ⇒ 00:44:02.929 Uttam Kumaran: Believe me.

536 00:44:03.430 ⇒ 00:44:06.449 Samuel Roberts: I was gonna say, I haven’t touched, like, besides just a simple, like.

537 00:44:06.450 ⇒ 00:44:07.710 Uttam Kumaran: Spent 2 hours.

538 00:44:07.710 ⇒ 00:44:08.730 Samuel Roberts: Oh, man.

539 00:44:09.020 ⇒ 00:44:13.569 Uttam Kumaran: And I feel like I’m getting… I’m really pretty good at finding software.

540 00:44:13.570 ⇒ 00:44:14.120 Samuel Roberts: Yeah.

541 00:44:14.120 ⇒ 00:44:20.360 Uttam Kumaran: your job, so this is the best. There’s a couple of others, but none were as feature-rich as this.

542 00:44:20.360 ⇒ 00:44:26.300 Samuel Roberts: Yeah, the only other thing I was thinking AI-wise is, like, and I’ve said this before, but, like, Mermaid documents can do GANTs.

543 00:44:27.350 ⇒ 00:44:29.560 Samuel Roberts: But… then it’s all text-based.

544 00:44:30.030 ⇒ 00:44:33.609 Uttam Kumaran: Yeah, like, again, a lot of our job is just so performative.

545 00:44:33.610 ⇒ 00:44:35.669 Samuel Roberts: Yeah, that’s what I feel… yeah.

546 00:44:35.900 ⇒ 00:44:38.489 Uttam Kumaran: It’s hard to be, like, pure play.

547 00:44:38.590 ⇒ 00:44:42.129 Uttam Kumaran: Yeah, I don’t… I’m trying to delete this thing, I don’t know what just happened.

548 00:44:42.950 ⇒ 00:44:44.400 Samuel Roberts: Maybe they’re just having issues today.

549 00:44:45.080 ⇒ 00:44:52.509 Uttam Kumaran: No, I literally said… Don’t send me this notification, but it wasn’t… it’s like, don’t allow the inbuilt…

550 00:44:52.980 ⇒ 00:44:54.799 Uttam Kumaran: Thang to send me a toast.

551 00:44:55.380 ⇒ 00:44:57.530 Samuel Roberts: Oh, so it’s something in their app, potentially, then?

552 00:44:57.530 ⇒ 00:44:59.520 Uttam Kumaran: Fucking toast, yeah.

553 00:44:59.520 ⇒ 00:45:01.319 Samuel Roberts: Oh, okay, yeah.

554 00:45:01.320 ⇒ 00:45:04.900 Uttam Kumaran: Well, it’s sending me… it’s sending me a native toast, like, from the browser.

555 00:45:04.900 ⇒ 00:45:05.730 Samuel Roberts: Oh.

556 00:45:05.730 ⇒ 00:45:11.509 Uttam Kumaran: So then I hit the button that said, don’t allow… I was like, I’m just deleting the things, like, don’t tell me every time.

557 00:45:11.680 ⇒ 00:45:12.570 Samuel Roberts: Yeah, that’s…

558 00:45:12.570 ⇒ 00:45:14.679 Uttam Kumaran: See, I don’t know what else I blocked here.

559 00:45:18.770 ⇒ 00:45:20.420 Uttam Kumaran: Okay, I’m just gonna delete this.

560 00:45:20.420 ⇒ 00:45:21.040 Samuel Roberts: Yeah.

561 00:45:21.530 ⇒ 00:45:22.540 Uttam Kumaran: Start over.

562 00:45:29.510 ⇒ 00:45:30.380 Samuel Roberts: Excuse me.

563 00:45:31.450 ⇒ 00:45:33.409 Pranav: This screen is gonna be in my nightmares.

564 00:45:35.200 ⇒ 00:45:36.819 Uttam Kumaran: Wait, am I getting it too?

565 00:45:36.820 ⇒ 00:45:44.119 Samuel Roberts: That’s what I’m wondering if they’re just having… like, if you’re not authenticated or something to begin, or, you know, since you cleared the cache or something, maybe it…

566 00:45:47.500 ⇒ 00:45:48.469 Pranav: Oh, jeez.

567 00:45:49.380 ⇒ 00:45:50.310 Uttam Kumaran: Brother.

568 00:45:51.080 ⇒ 00:45:51.779 Uttam Kumaran: What?

569 00:46:00.600 ⇒ 00:46:05.199 Uttam Kumaran: Where’s all the…

570 00:46:05.200 ⇒ 00:46:07.510 Samuel Roberts: I think it’s under Application?

571 00:46:08.710 ⇒ 00:46:10.210 Samuel Roberts: And then, yeah…

572 00:46:12.880 ⇒ 00:46:16.970 Pranav: There’s also an option to just, like, don’t store cookies.

573 00:46:21.200 ⇒ 00:46:23.870 Samuel Roberts: There you go, that’s the screen you wanted to see, okay.

574 00:46:30.180 ⇒ 00:46:31.400 Uttam Kumaran: Where is that option?

575 00:46:31.720 ⇒ 00:46:33.590 Uttam Kumaran: It’s on individual site settings?

576 00:46:34.760 ⇒ 00:46:36.289 Samuel Roberts: No, no, it’s not.

577 00:46:36.290 ⇒ 00:46:37.740 Pranav: It’s also in the dev tools.

578 00:46:37.890 ⇒ 00:46:41.539 Pranav: It’s, what did I do?

579 00:46:41.810 ⇒ 00:46:43.190 Pranav: Was it a network?

580 00:46:44.650 ⇒ 00:46:46.560 Samuel Roberts: Oh, is that just for when the DevTools are open, though?

581 00:46:48.180 ⇒ 00:46:49.449 Uttam Kumaran: Oh, disable cache.

582 00:46:49.630 ⇒ 00:46:50.370 Pranav: Yeah.

583 00:46:50.590 ⇒ 00:46:52.580 Samuel Roberts: I think this is when the tools are open for that.

584 00:46:59.470 ⇒ 00:47:03.520 Samuel Roberts: Yeah, so, Pranab, there’s probably something still stored somewhere that’s…

585 00:47:03.720 ⇒ 00:47:06.179 Samuel Roberts: You gotta discover fully, yeah, okay.

586 00:47:06.710 ⇒ 00:47:08.469 Pranav: It’s clear, everything like that.

587 00:47:09.530 ⇒ 00:47:15.880 Samuel Roberts: Interesting, okay. Either their app is looking for something that’s not there, or there’s something telling it there’s still something there, maybe? That’s what you just…

588 00:47:15.880 ⇒ 00:47:16.450 Pranav: Yeah.

589 00:47:17.130 ⇒ 00:47:21.129 Pranav: Interesting. Even though I went through Safari, but maybe they just don’t support Safari.

590 00:47:21.680 ⇒ 00:47:24.560 Samuel Roberts: Yeah, that’s… I mean… That’s fair. Yeah.

591 00:47:24.760 ⇒ 00:47:28.650 Uttam Kumaran: To give you guys the other… there’s a Scant software…

592 00:47:30.120 ⇒ 00:47:38.300 Uttam Kumaran: That’s very, very famous, too. That’s the other big one. Yeah, Gantt project.

593 00:47:38.920 ⇒ 00:47:42.390 Uttam Kumaran: This is, like, the other one, but it’s… this is, like, a desktop app that, like…

594 00:47:42.390 ⇒ 00:47:42.900 Samuel Roberts: Of course.

595 00:47:42.900 ⇒ 00:47:45.059 Uttam Kumaran: It’s all these, like, it’s like a big boomer app.

596 00:47:45.390 ⇒ 00:47:46.130 Pranav: Mmm.

597 00:47:46.370 ⇒ 00:47:48.210 Uttam Kumaran: So, I decided not to… Oh, gotcha.

598 00:47:48.880 ⇒ 00:47:57.939 Uttam Kumaran: There’s a couple of others, but, like, again, like, there wasn’t anything as sophisticated as the one that I saw. Like, a lot of these, a landing page will be really good, and then the thing sucks.

599 00:47:58.380 ⇒ 00:48:01.729 Uttam Kumaran: So, I’ll find it. I mean, for me, the… I… the…

600 00:48:03.060 ⇒ 00:48:08.899 Uttam Kumaran: if we switch, it has to have, like, a sick API, because I want to be able to do this.

601 00:48:09.400 ⇒ 00:48:12.420 Uttam Kumaran: from… Like, a chat agent.

602 00:48:12.430 ⇒ 00:48:14.530 Samuel Roberts: That’s… yeah, that’s… well, that’s the only reason I was thinking, like…

603 00:48:14.530 ⇒ 00:48:17.940 Uttam Kumaran: able to do anything. Like, I’m so anti-UI.

604 00:48:18.180 ⇒ 00:48:19.259 Samuel Roberts: Yeah, no, I hear you.

605 00:48:19.260 ⇒ 00:48:32.529 Uttam Kumaran: I want us to be… I want us to… at Brainforge, like, we’re… people are talking about how, like, things are gonna… not gonna be UI, because you’ll just do everything chat. We have to do… we have to pull that future forward.

606 00:48:32.870 ⇒ 00:48:43.190 Uttam Kumaran: Because otherwise, we are going to be paying… like, the thing is, here, we’ve been able to do a good job not hiring project managers whose sole job is to do this type of busy work.

607 00:48:43.350 ⇒ 00:48:53.179 Uttam Kumaran: And they’re pure cost centers. Like, they don’t do engineering work. They just, like, move shit around. So my… what I’m trying to think about is, like, how do we…

608 00:48:53.760 ⇒ 00:49:06.089 Uttam Kumaran: how do I give everybody, like, 5% of that person’s job, or 10%, and so we don’t have to hire that person? And don’t feel bad, that person is, like, a fictional person, so we’re not letting people go, but, like,

609 00:49:06.090 ⇒ 00:49:14.770 Uttam Kumaran: I want everybody on engineering to do, like, 10-15% PM, and it all adds up to, like, 60% of a PM to do just the stuff that matters, which is, like.

610 00:49:14.780 ⇒ 00:49:20.109 Uttam Kumaran: get a Gantt chart going, get tickets, get the docs, things like that. And we’re showing that it’s possible.

611 00:49:20.170 ⇒ 00:49:26.610 Uttam Kumaran: Because, like, I don’t think you need a… I don’t want… well, I have a full-time person whose job is to do this, but then they don’t know any of the words.

612 00:49:26.720 ⇒ 00:49:30.189 Uttam Kumaran: Like, they don’t know what they’re really doing, you know? And getting a.

613 00:49:30.190 ⇒ 00:49:30.880 Samuel Roberts: Great.

614 00:49:30.880 ⇒ 00:49:32.480 Uttam Kumaran: Like, technical project manager?

615 00:49:32.480 ⇒ 00:49:33.250 Samuel Roberts: Yeah, I was…

616 00:49:33.250 ⇒ 00:49:34.990 Uttam Kumaran: Really hard, I was one.

617 00:49:34.990 ⇒ 00:49:35.930 Samuel Roberts: And…

618 00:49:36.120 ⇒ 00:49:39.469 Uttam Kumaran: It’s a… it’s a stupid job if you’re an engineer to do.

619 00:49:39.470 ⇒ 00:49:42.760 Samuel Roberts: And so that’s the thing, like, part of me that wants to be like, oh, I could… but I could know this.

620 00:49:42.760 ⇒ 00:49:44.129 Uttam Kumaran: No, no, no, you don’t want to do that.

621 00:49:44.130 ⇒ 00:49:44.609 Samuel Roberts: What did you do?

622 00:49:44.610 ⇒ 00:49:48.419 Uttam Kumaran: I did it because I wanted to learn product management.

623 00:49:48.420 ⇒ 00:49:49.110 Samuel Roberts: Yep.

624 00:49:49.110 ⇒ 00:49:55.310 Uttam Kumaran: But… And… Product management is not that hard, because it’s mostly writing.

625 00:49:55.310 ⇒ 00:49:56.530 Samuel Roberts: And, like.

626 00:49:56.830 ⇒ 00:50:03.410 Uttam Kumaran: And, like, selling to the business. But it’s a thankless job, and you just get beat up, like, it’s horrible.

627 00:50:05.140 ⇒ 00:50:07.920 Uttam Kumaran: Okay, sorry, that’s… that’s just my…

628 00:50:07.920 ⇒ 00:50:09.160 Samuel Roberts: You’re good, you’re good.

629 00:50:09.160 ⇒ 00:50:09.830 Uttam Kumaran: Okay.

630 00:50:10.400 ⇒ 00:50:11.829 Samuel Roberts: Did you look at Team Gantt?

631 00:50:12.910 ⇒ 00:50:15.510 Uttam Kumaran: No, I thought that was this one. Okay, this is not.

632 00:50:16.300 ⇒ 00:50:20.450 Samuel Roberts: Well, I was just searching Gantt API, and I found one that’s, like, a…

633 00:50:20.840 ⇒ 00:50:25.850 Samuel Roberts: like, a component library for building GANs, but that’s… If we wanted to, like…

634 00:50:26.240 ⇒ 00:50:30.400 Samuel Roberts: have Gantt in the platform or something, but then I just searched Team Gantt APIs here.

635 00:50:30.640 ⇒ 00:50:34.840 Samuel Roberts: Yes, you can complete CRUD actions, so I wonder, but anyway.

636 00:50:34.840 ⇒ 00:50:37.600 Uttam Kumaran: I don’t know, my delete’s not working, so I’m just gonna…

637 00:50:38.000 ⇒ 00:50:39.139 Samuel Roberts: Jesus Christ.

638 00:50:39.140 ⇒ 00:50:40.099 Uttam Kumaran: I saw Andrew’s gonna.

639 00:50:40.100 ⇒ 00:50:42.080 Samuel Roberts: Which one are you trying to delete? I can get in there, maybe, and try it.

640 00:50:42.280 ⇒ 00:50:43.970 Uttam Kumaran: Phase 2, everything in Phase 2.

641 00:50:43.970 ⇒ 00:50:44.900 Samuel Roberts: Everything in Phase 2.

642 00:50:44.900 ⇒ 00:50:47.720 Uttam Kumaran: So just, like, uninstall Arc and reinstall or something.

643 00:50:47.990 ⇒ 00:50:49.030 Samuel Roberts: Oh, jeez.

644 00:50:51.220 ⇒ 00:50:59.520 Uttam Kumaran: Phase 2, we have, PRD… creation, we have PRD, Sign off.

645 00:51:02.180 ⇒ 00:51:05.159 Samuel Roberts: Alright, so the one, like, the Amazon wholesale one, that’s what we’re looking for.

646 00:51:05.160 ⇒ 00:51:08.590 Uttam Kumaran: Charity, sign off… Tech.

647 00:51:09.240 ⇒ 00:51:15.230 Uttam Kumaran: And then we’ll sort of, like… Placeholder…

648 00:51:18.290 ⇒ 00:51:21.870 Uttam Kumaran: roadmap.

649 00:51:22.370 ⇒ 00:51:26.660 Uttam Kumaran: And then we also have a… Phase 3…

650 00:51:27.880 ⇒ 00:51:31.230 Uttam Kumaran: Where we have very, very similar ones with these.

651 00:51:32.060 ⇒ 00:51:36.460 Uttam Kumaran: How do I just include duplicate?

652 00:51:36.460 ⇒ 00:51:38.730 Samuel Roberts: This should be gone now. Okay, cool.

653 00:51:39.560 ⇒ 00:51:41.880 Samuel Roberts: But yeah, I see what you’re saying, it was using the default alert.

654 00:51:43.670 ⇒ 00:51:47.880 Uttam Kumaran: Yeah, see, look, they have native toast, but they didn’t use it before.

655 00:51:49.640 ⇒ 00:51:52.069 Uttam Kumaran: Okay. Yeah, if you’re gonna eat, pick one, like…

656 00:51:52.430 ⇒ 00:51:52.940 Samuel Roberts: Yeah.

657 00:51:54.360 ⇒ 00:51:56.200 Samuel Roberts: I know the word is, dude, I’m done with it.

658 00:51:56.200 ⇒ 00:51:59.090 Uttam Kumaran: modals, dude, I’m done. Look, no more.

659 00:51:59.510 ⇒ 00:52:00.440 Samuel Roberts: I hear ya.

660 00:52:00.440 ⇒ 00:52:03.949 Uttam Kumaran: No more. I used to really ship a lot of pop-ups.

661 00:52:03.950 ⇒ 00:52:04.590 Samuel Roberts: Oh, yeah.

662 00:52:04.590 ⇒ 00:52:06.560 Uttam Kumaran: company I was working for, and…

663 00:52:07.020 ⇒ 00:52:09.159 Uttam Kumaran: Yeah, I don’t know. Okay, so…

664 00:52:09.160 ⇒ 00:52:16.550 Samuel Roberts: That’s always the sign when there’s a bunch of new, like, product managers on an app that’s starting to take off, and everything’s a little pop-up for their feature, and it’s like, oh no.

665 00:52:16.550 ⇒ 00:52:17.570 Uttam Kumaran: It’s lazy, it’s lazy.

666 00:52:17.570 ⇒ 00:52:18.200 Samuel Roberts: Yeah.

667 00:52:20.110 ⇒ 00:52:25.810 Uttam Kumaran: Okay, purity sign-off, purity creation copy… Okay.

668 00:52:26.690 ⇒ 00:52:31.740 Uttam Kumaran: So Zach and Bobby onboarded…

669 00:52:33.680 ⇒ 00:52:38.129 Uttam Kumaran: Couple of things we want to also enable as part of this Phase 1 is, like,

670 00:52:38.250 ⇒ 00:52:44.860 Uttam Kumaran: We want to talk through… Well…

671 00:52:45.460 ⇒ 00:52:48.669 Uttam Kumaran: I guess, like, the chat functionality…

672 00:52:55.180 ⇒ 00:53:00.519 Uttam Kumaran: I guess maybe a question for you, Sam, like, is our go-to chat just gonna be, like, Copilot kit?

673 00:53:07.460 ⇒ 00:53:10.360 Uttam Kumaran: Or do you want to do another, like, spike around that?

674 00:53:10.580 ⇒ 00:53:15.839 Samuel Roberts: Well, so, here’s the thing. The Copilot kit was great for what we needed, because it was, like, a drop-in UI and everything.

675 00:53:18.430 ⇒ 00:53:24.309 Samuel Roberts: And there are still some nice things that we could get out of it, and they just released a new update with the.

676 00:53:25.020 ⇒ 00:53:25.699 Uttam Kumaran: And so, like…

677 00:53:25.980 ⇒ 00:53:27.320 Samuel Roberts: I… I… it…

678 00:53:27.640 ⇒ 00:53:36.270 Samuel Roberts: might be right still. It’s just… there… it’s… there’s a lot that we might not need from it, so I’m wondering if we need… if we want to just go, like, the AI SDK.

679 00:53:37.040 ⇒ 00:53:38.760 Uttam Kumaran: Yeah, I would prefer that.

680 00:53:38.800 ⇒ 00:53:46.909 Samuel Roberts: I… I… yeah, especially because if we’re already doing… gonna be doing UI stuff, like, it was nice to drop into the platform. It was nice to…

681 00:53:47.090 ⇒ 00:53:49.730 Uttam Kumaran: I would prefer to just do whatever is, like.

682 00:53:50.060 ⇒ 00:53:54.049 Uttam Kumaran: basically state-of-the-art in terms of the chat. I think that’s AI SDK, basically.

683 00:53:54.050 ⇒ 00:54:06.539 Samuel Roberts: The only other thing is that the… what CopilotKit does well is it integrates with apps that have, like, stayed on the front end and back end, and you can easily just be like, here’s the shared state.

684 00:54:06.890 ⇒ 00:54:08.619 Uttam Kumaran: Yeah, that’s fine, dude.

685 00:54:08.620 ⇒ 00:54:13.340 Samuel Roberts: But I don’t think… if we’re just… if it’s just chat, and it’s gonna be… I mean, eventually when it’s the…

686 00:54:13.720 ⇒ 00:54:24.499 Samuel Roberts: the Forecaster and stuff, I don’t know if the co-pilot kit’s gonna be… it might be the better option for that, where there’s an actual app, and then there’s chat next to it, but for just, like, a master chat like they had.

687 00:54:27.250 ⇒ 00:54:27.890 Uttam Kumaran: Yeah.

688 00:54:27.900 ⇒ 00:54:32.129 Samuel Roberts: I mean, I still don’t like reinventing the wheel, building a whole new, like, all the streaming and all the… all the.

689 00:54:32.130 ⇒ 00:54:35.389 Uttam Kumaran: I thought the AI SDK, like, supports that.

690 00:54:35.540 ⇒ 00:54:37.240 Samuel Roberts: Does it do all that? Okay.

691 00:54:37.240 ⇒ 00:54:38.349 Uttam Kumaran: I thought so.

692 00:54:38.350 ⇒ 00:54:45.059 Samuel Roberts: Yeah, it didn’t have the UI, is that what it didn’t have? No, I guess it does. Okay. I thought it didn’t have, like, pre-built components and stuff.

693 00:54:46.050 ⇒ 00:54:47.550 Samuel Roberts: It has all the primitives.

694 00:54:47.550 ⇒ 00:54:49.319 Uttam Kumaran: Mustafa, what do you think?

695 00:54:53.820 ⇒ 00:54:55.040 Mustafa Raja: Holly, can you say that again?

696 00:54:56.120 ⇒ 00:54:58.459 Uttam Kumaran: I said, what do you think about the AI SDK?

697 00:54:59.830 ⇒ 00:55:07.080 Mustafa Raja: Oh, for the UI, hmm. I think the AI SDK, it’s built on ShadCN.

698 00:55:07.600 ⇒ 00:55:10.249 Mustafa Raja: That we are already using.

699 00:55:11.380 ⇒ 00:55:15.240 Samuel Roberts: Yeah, that’s… so there might actually be just things out there already.

700 00:55:15.240 ⇒ 00:55:16.460 Mustafa Raja: I think…

701 00:55:16.460 ⇒ 00:55:19.050 Samuel Roberts: when I’m looking at the AI SDK, I don’t see…

702 00:55:19.190 ⇒ 00:55:22.170 Samuel Roberts: actual, like, UI components. I see…

703 00:55:22.670 ⇒ 00:55:24.880 Samuel Roberts: Which is, honestly, this is what we need, and I think…

704 00:55:24.880 ⇒ 00:55:27.370 Uttam Kumaran: There is. There is AI SDK UI.

705 00:55:28.170 ⇒ 00:55:30.159 Uttam Kumaran: Here, I’ll send it to you.

706 00:55:34.360 ⇒ 00:55:35.359 Uttam Kumaran: I sent it in Zoom.

707 00:55:35.360 ⇒ 00:55:40.479 Samuel Roberts: Is this what I’m looking at? Yeah, I think this is what I’m… oh, chat bot. Okay, that’s all I’m looking at. I was looking at the chat, okay.

708 00:55:42.860 ⇒ 00:55:45.380 Samuel Roberts: What was I? Wait, what is this? Hold on.

709 00:55:50.210 ⇒ 00:55:55.619 Samuel Roberts: Oh, chatbot… what was I at? Oh, I was at the docked reference AI, okay. Of course.

710 00:55:55.620 ⇒ 00:55:58.519 Uttam Kumaran: There’s the AI SDK core, and then the SDK UI.

711 00:55:58.640 ⇒ 00:56:03.590 Samuel Roberts: Yeah, I was at the documentation for the UI, so I was seeing all the, like, the…

712 00:56:03.590 ⇒ 00:56:14.069 Mustafa Raja: Yeah, I think just by looking at the documentation page, it looks pretty… pretty much that they are using, chat scene over here too, you know?

713 00:56:14.300 ⇒ 00:56:14.790 Uttam Kumaran: Yeah.

714 00:56:14.790 ⇒ 00:56:18.080 Mustafa Raja: So they’re pretty married to Shanxian, which is good, because I love it.

715 00:56:18.080 ⇒ 00:56:24.320 Samuel Roberts: Yeah, no, I agree with that. I just didn’t know if this had, like, the… Because what I’m looking.

716 00:56:24.320 ⇒ 00:56:35.519 Mustafa Raja: Champion is crazy good because, you know, it does not download everything in one go, but we can decide which thing we want, which reduces the…

717 00:56:35.810 ⇒ 00:56:40.889 Mustafa Raja: you know, build. And you own it. And you own it, which is good, too. Yeah, also, yeah.

718 00:56:41.680 ⇒ 00:56:43.940 Samuel Roberts: Yeah, does chatbot exist… is this an egg?

719 00:56:44.500 ⇒ 00:56:52.769 Samuel Roberts: Yeah, the use chat hook is… is what I’m looking at here. So the difference here is that, like, when we did Copilot Kit, we just dropped in a React component.

720 00:56:52.930 ⇒ 00:56:57.430 Samuel Roberts: And then had a few other hooks that we could use to interact with it.

721 00:56:58.080 ⇒ 00:57:00.010 Samuel Roberts: This is just giving us those hooks.

722 00:57:03.100 ⇒ 00:57:08.070 Samuel Roberts: So it is still, like, a UI thing, but they’re still putting separate inputs, buttons, like, it’s all…

723 00:57:08.570 ⇒ 00:57:13.570 Samuel Roberts: Which, again, might be what we need for this project more, because if it’s just a master chat thing.

724 00:57:14.650 ⇒ 00:57:19.989 Mustafa Raja: Yeah, and I think, Versa owns both of these, ChadCN and AI SDK2.

725 00:57:20.700 ⇒ 00:57:21.860 Samuel Roberts: Yeah, they’re definitely stuck on top.

726 00:57:21.860 ⇒ 00:57:24.020 Mustafa Raja: So they are pretty integrated with each other.

727 00:57:24.780 ⇒ 00:57:28.279 Samuel Roberts: I don’t think… I don’t know what co-pilot kids, but they are…

728 00:57:31.550 ⇒ 00:57:35.800 Samuel Roberts: I don’t think it’s… yeah, but Monster’s on top of AI SDK as well, I think, so…

729 00:57:37.820 ⇒ 00:57:39.160 Mustafa Raja: Hmm, yeah.

730 00:57:44.790 ⇒ 00:57:48.930 Mustafa Raja: And the new app that we are doing, is it in Next or something?

731 00:57:50.150 ⇒ 00:58:02.019 Samuel Roberts: Yeah, so there’s a… there’s a new, like, base repo that’s sort of put together that has, like, auth hooked up with Next, with, Postgres.

732 00:58:02.350 ⇒ 00:58:04.160 Samuel Roberts: And.

733 00:58:04.730 ⇒ 00:58:09.320 Samuel Roberts: Prisma… and Nest.js on the back end.

734 00:58:10.920 ⇒ 00:58:17.159 Samuel Roberts: Yeah, so I think it… Yeah, we’re gonna be using Prisma for this. Have you used Prisma before?

735 00:58:17.530 ⇒ 00:58:18.290 Mustafa Raja: Yep.

736 00:58:18.290 ⇒ 00:58:25.600 Samuel Roberts: Yeah, okay. Yeah, so I mean, it’s gonna give us a lot of the stuff that we’re getting from Superbase, but this is all gonna be hosted right on Railway, and not have to worry about.

737 00:58:26.530 ⇒ 00:58:27.240 Mustafa Raja: Yeah.

738 00:58:27.240 ⇒ 00:58:30.269 Samuel Roberts: Tying it to another thing, another separate app.

739 00:58:32.120 ⇒ 00:58:35.510 Samuel Roberts: Because, yeah, we’re getting off and better off now, so it’s gonna work.

740 00:58:36.730 ⇒ 00:58:45.200 Samuel Roberts: Pretty well. Where was it going? Okay, I want to see the GitHub.

741 00:59:22.470 ⇒ 00:59:37.130 Uttam Kumaran: Okay, I feel pretty good about this. Is there anything else on the front end? Like, we have the chat framework, I want to figure out external, like, how do we call MCPs, schedule a Slack message, I think other things for us to figure out, too, is, like, Replit integration with GitHub.

742 00:59:37.130 ⇒ 00:59:38.499 Samuel Roberts: Oh, that’s right, yep.

743 00:59:38.500 ⇒ 00:59:40.550 Uttam Kumaran: What else?

744 00:59:40.550 ⇒ 00:59:42.220 Samuel Roberts: Grab our replett.

745 00:59:46.260 ⇒ 00:59:46.930 Samuel Roberts: MCP.

746 00:59:46.930 ⇒ 00:59:49.160 Uttam Kumaran: I think we should enable voice.

747 00:59:50.180 ⇒ 00:59:51.280 Uttam Kumaran: as input.

748 00:59:51.690 ⇒ 00:59:52.510 Samuel Roberts: Oh, sure.

749 00:59:53.200 ⇒ 00:59:57.130 Uttam Kumaran: That’s just a nice… and then the other thing, guys, like, as we do these.

750 00:59:57.330 ⇒ 01:00:00.700 Uttam Kumaran: Try to keep writing shit back to the base repo.

751 01:00:01.190 ⇒ 01:00:07.550 Uttam Kumaran: Because, we’re gonna… we’re gonna do chat-type stuff like this for a lot of folks.

752 01:00:08.000 ⇒ 01:00:14.130 Uttam Kumaran: on AI SDK and things like that, so as much as you can push back into the base repo that is reusable.

753 01:00:14.280 ⇒ 01:00:17.500 Uttam Kumaran: would be great. So, like, for example, if we decide on using a voice

754 01:00:18.220 ⇒ 01:00:24.699 Uttam Kumaran: provider, and, like, we have the framework for voice to share… put that component back. I’m just using… probably not using.

755 01:00:24.700 ⇒ 01:00:27.619 Pranav: What is the best way for us to, like, maintain…

756 01:00:27.680 ⇒ 01:00:28.800 Samuel Roberts: That’s what I… that’s what…

757 01:00:28.800 ⇒ 01:00:36.439 Uttam Kumaran: I would rather push it to Lilo first, and then every month or so, we can come back and rewrite it to the base.

758 01:00:36.570 ⇒ 01:00:38.759 Samuel Roberts: Yeah, I don’t think it… yeah, exactly.

759 01:00:38.760 ⇒ 01:00:39.140 Uttam Kumaran: Yeah.

760 01:00:39.140 ⇒ 01:00:44.979 Samuel Roberts: like, pull anything from Lilo down, we’ll just be like, this is how we implemented it, we can see the code, and then we’ll add it to the template.

761 01:00:45.040 ⇒ 01:00:51.070 Uttam Kumaran: Yeah, and honestly, what you can do is, like, once a month or so, just have AI, like, do that.

762 01:00:51.480 ⇒ 01:00:51.910 Samuel Roberts: We’ll go 3.

763 01:00:51.910 ⇒ 01:00:53.490 Uttam Kumaran: It goes up in cursor and have it.

764 01:00:53.840 ⇒ 01:00:56.130 Uttam Kumaran: Have it just move reusable stuff back?

765 01:00:56.300 ⇒ 01:00:59.620 Uttam Kumaran: Like, I’m starting to maintain base repos for data and stuff, too.

766 01:00:59.620 ⇒ 01:01:00.420 Samuel Roberts: Yeah, yeah.

767 01:01:00.420 ⇒ 01:01:04.990 Uttam Kumaran: So voice is input, and then we want to also do, like, link fuse, right?

768 01:01:06.580 ⇒ 01:01:07.230 Samuel Roberts: Yes.

769 01:01:08.460 ⇒ 01:01:12.409 Uttam Kumaran: you know what you wanna do? You’re gonna host Mastra on…

770 01:01:14.730 ⇒ 01:01:15.560 Samuel Roberts: For this.

771 01:01:16.940 ⇒ 01:01:20.909 Samuel Roberts: It would probably be all part of the same thing, because it can be part of the framework, so…

772 01:01:22.540 ⇒ 01:01:23.280 Uttam Kumaran: So…

773 01:01:23.590 ⇒ 01:01:28.230 Samuel Roberts: Well, okay, so we may not need Monster yet if we’re gonna just use the AI SDK for chat.

774 01:01:28.370 ⇒ 01:01:29.340 Uttam Kumaran: Oh, okay, okay, yeah.

775 01:01:29.340 ⇒ 01:01:32.169 Samuel Roberts: The question becomes, like, as it does more agentic things.

776 01:01:32.590 ⇒ 01:01:36.839 Uttam Kumaran: If it’s… for just the chat element, it’s a little overkill. Okay.

777 01:01:36.880 ⇒ 01:01:40.019 Samuel Roberts: But it might… well, I’m sure AISDK has pretty good length of these integrations online.

778 01:01:40.020 ⇒ 01:01:42.620 Uttam Kumaran: Yeah, yeah, yeah, okay, so I would just do LangFuse…

779 01:01:42.760 ⇒ 01:01:43.410 Samuel Roberts: Yeah.

780 01:01:46.770 ⇒ 01:01:50.210 Samuel Roberts: Yeah, where we… I mean, phase… like…

781 01:01:50.690 ⇒ 01:01:55.270 Samuel Roberts: 2 and 3, when we start to implement those other…

782 01:01:55.720 ⇒ 01:02:04.150 Samuel Roberts: forecasting and then interacting with the forecasting in the chat and stuff, that’s where some of this more agentic, maybe even Copilot kit stuff would be relevant.

783 01:02:04.510 ⇒ 01:02:07.609 Samuel Roberts: But that’s a…

784 01:02:08.140 ⇒ 01:02:13.120 Uttam Kumaran: Are you… are you, sam, are you out? You’re out Monday and Tuesday, too, or no?

785 01:02:14.640 ⇒ 01:02:15.370 Samuel Roberts: Next week?

786 01:02:15.370 ⇒ 01:02:16.030 Uttam Kumaran: Yeah.

787 01:02:17.250 ⇒ 01:02:23.239 Samuel Roberts: I was not planning to be, but it’s… I’m gonna have family in town and stuff, so it’s gonna be a little, like…

788 01:02:23.640 ⇒ 01:02:24.660 Samuel Roberts: Yeah, I think it’ll just…

789 01:02:24.660 ⇒ 01:02:31.229 Uttam Kumaran: be… I think it’ll be a chill week. I’m just wondering if we can basically sign off on the PRD before we go on break.

790 01:02:31.230 ⇒ 01:02:33.139 Samuel Roberts: That, yeah, okay, that’s fine, yeah, totally.

791 01:02:33.660 ⇒ 01:02:36.439 Uttam Kumaran: Like, I kinda wanna sign on… I just kind of want to s-

792 01:02:36.610 ⇒ 01:02:39.509 Uttam Kumaran: Try to sign off on it,

793 01:02:40.310 ⇒ 01:02:50.950 Uttam Kumaran: from the client side on Friday, and then on our side, at least by, Whatever the…

794 01:02:51.340 ⇒ 01:02:52.609 Uttam Kumaran: I don’t move this thing?

795 01:02:52.610 ⇒ 01:02:58.039 Samuel Roberts: Yeah, Monday, Tuesday, probably not Wednesday’s Christmas Eve, and it’s my wife’s birthday, so…

796 01:02:58.040 ⇒ 01:03:03.120 Uttam Kumaran: No, no, no, yeah, definitely not, definitely… Yeah. No, no, after, after, after Tuesday, it’s… we’re done.

797 01:03:03.780 ⇒ 01:03:04.750 Samuel Roberts: Okay, okay.

798 01:03:05.050 ⇒ 01:03:13.090 Uttam Kumaran: Yeah, so PRD sign-off tech, PRD sign-off client by Friday, and then… There’s a placeholder roadmap.

799 01:03:13.470 ⇒ 01:03:13.800 Samuel Roberts: Okay.

800 01:03:17.930 ⇒ 01:03:22.190 Uttam Kumaran: And ideally, what I’m gonna… we can work to see, like, who wants to take what.

801 01:03:22.540 ⇒ 01:03:23.589 Samuel Roberts: Yeah. For this? Yeah.

802 01:03:24.210 ⇒ 01:03:28.630 Uttam Kumaran: Cool. Okay.

803 01:03:28.960 ⇒ 01:03:37.779 Uttam Kumaran: So let me, let me just… I’ll wrap this up today, and then I’m gonna send this for… also for their review, at least on Phase 1. Gonna work on the Phase 1 PRD.

804 01:03:38.030 ⇒ 01:03:45.989 Uttam Kumaran: It’ll be, like, it’ll be a lot more direct than, like, it’s just gonna be, like, making sure we have a couple of these. Yeah, yeah.

805 01:03:46.590 ⇒ 01:03:55.730 Uttam Kumaran: And then, if you guys could also take another look at the Phase 2, and just, like, leave comments in line with, like, this is confusing, this doesn’t make sense.

806 01:03:56.600 ⇒ 01:04:06.729 Uttam Kumaran: it’s… it’s a… it’s a… it’s gonna be interesting. There have to… we have to make some technical design decisions on, like, how we land data, how we’re running a process over it.

807 01:04:07.050 ⇒ 01:04:15.590 Uttam Kumaran: And how we’re segmenting by client. So, we will have to… like, it’s important for everybody to have a good understanding of that scope.

808 01:04:16.240 ⇒ 01:04:17.340 Samuel Roberts: That’s an exciting one.

809 01:04:17.340 ⇒ 01:04:18.800 Uttam Kumaran: So, yeah, okay.

810 01:04:20.400 ⇒ 01:04:22.539 Samuel Roberts: And that’s in… that’s in Docs right now?

811 01:04:22.540 ⇒ 01:04:24.019 Uttam Kumaran: Yeah, that’s in Docs, yeah.

812 01:04:24.020 ⇒ 01:04:24.770 Samuel Roberts: Okay.

813 01:04:24.770 ⇒ 01:04:28.170 Uttam Kumaran: It’s in the repo, too, but I would just, like, edit the doc, because…

814 01:04:28.170 ⇒ 01:04:29.530 Samuel Roberts: Yeah, that’s what I’m making sure, okay.

815 01:04:30.270 ⇒ 01:04:32.370 Uttam Kumaran: And then, I guess…

816 01:04:32.780 ⇒ 01:04:41.499 Uttam Kumaran: Sam, do you guys want to keep working on ABC stuff, and just let me know when things get to a point where I can start to test, or what do you think is best there?

817 01:04:42.490 ⇒ 01:04:48.610 Samuel Roberts: Yeah, I mean, I think we gotta figure out the logging right now.

818 01:04:48.610 ⇒ 01:04:54.880 Uttam Kumaran: Do you need… I guess, like, do you need… I’m trying to get Tim to give us another chat instance.

819 01:04:55.190 ⇒ 01:04:58.149 Samuel Roberts: So you can basically just completely rebuild.

820 01:04:58.440 ⇒ 01:05:03.559 Uttam Kumaran: a staging version, and a dev… if you’re like, that’s the blocker, then I’ll try to move that forward today.

821 01:05:03.560 ⇒ 01:05:09.640 Samuel Roberts: I mean, it’s not… we can still be testing locally, that’s the nice thing about the master dev environment, is, like, we don’t need a chat interface, it has one there.

822 01:05:09.640 ⇒ 01:05:10.030 Uttam Kumaran: Okay.

823 01:05:10.030 ⇒ 01:05:13.589 Samuel Roberts: Like, we can be developing that, and then hook that in once…

824 01:05:13.590 ⇒ 01:05:18.480 Uttam Kumaran: So I guess just let me know what the… yeah, so I just want you guys to be able to do the swap.

825 01:05:19.240 ⇒ 01:05:23.790 Uttam Kumaran: I’m maintaining the same… functionality today.

826 01:05:23.790 ⇒ 01:05:24.340 Samuel Roberts: Yeah.

827 01:05:24.540 ⇒ 01:05:26.429 Uttam Kumaran: And then we can keep editing the prompt.

828 01:05:27.290 ⇒ 01:05:28.639 Samuel Roberts: Okay, the swap…

829 01:05:28.930 ⇒ 01:05:30.360 Uttam Kumaran: Or you don’t want to do a swap.

830 01:05:30.810 ⇒ 01:05:40.269 Samuel Roberts: No, no, no, I mean, the swap right now is just gonna give us the time benefit of… the swap for the DB tool, is the piece that we have, yeah. It’s just gonna give us the time benefit.

831 01:05:40.270 ⇒ 01:05:40.620 Uttam Kumaran: What is that?

832 01:05:40.620 ⇒ 01:05:41.190 Samuel Roberts: Okay.

833 01:05:41.440 ⇒ 01:05:43.289 Samuel Roberts: Remember, it was taking 10 seconds and stuff?

834 01:05:44.510 ⇒ 01:05:46.249 Uttam Kumaran: Yeah, dude, then I would just ship that, yeah.

835 01:05:46.250 ⇒ 01:05:48.559 Samuel Roberts: I was gonna say, we can do that right now, it’s, it’s, it’s…

836 01:05:48.880 ⇒ 01:06:00.780 Samuel Roberts: Mustafa’s got it, so it’s given the same outputs, which is still, like, broken with that mosquito misting thing. So that’s where the prompting can start to… we can start to iterate on that. Do you want to iterate on that before we plug it in, or do you want to plug it in and just…

837 01:06:01.060 ⇒ 01:06:04.400 Uttam Kumaran: I would plug it in. Okay. Like, knock out one thing at a time.

838 01:06:04.590 ⇒ 01:06:05.300 Samuel Roberts: Okay.

839 01:06:05.300 ⇒ 01:06:09.510 Uttam Kumaran: plug it in, and then if you can give… if… I would say…

840 01:06:09.780 ⇒ 01:06:11.890 Uttam Kumaran: Do what you need to make sure that

841 01:06:12.260 ⇒ 01:06:26.589 Uttam Kumaran: it’s working. So whatever logging that doesn’t exist, try to enable that. Ultimately, I’m gonna ask both of you, is this working in production? And I don’t want to hear, like, we tested in dev, like, show me… show something so I can then show the client.

842 01:06:26.730 ⇒ 01:06:28.859 Uttam Kumaran: We’ve migrated this one piece.

843 01:06:29.070 ⇒ 01:06:40.570 Uttam Kumaran: And now we’re working on changing the prompt, and we have logging. So if you guys can do that today, that would be… that would be helpful. I’m… I’m gonna… and then I think maybe, Sam, I’m gonna… you have, like, an hour…

844 01:06:41.380 ⇒ 01:06:51.670 Uttam Kumaran: In the hour before CTA, or the 30 minutes before CTA, if you have time to work on some of my comments, that would be great. But I wouldn’t… I would just try to budget that time for comment review.

845 01:06:51.670 ⇒ 01:06:52.449 Samuel Roberts: 30 minutes? Okay.

846 01:06:52.450 ⇒ 01:06:53.050 Uttam Kumaran: Yeah.

847 01:06:53.260 ⇒ 01:06:54.030 Uttam Kumaran: Okay.

848 01:06:54.560 ⇒ 01:06:55.180 Samuel Roberts: Sounds good.

849 01:06:55.180 ⇒ 01:06:57.130 Uttam Kumaran: Alright, guys. Thank you.

850 01:06:57.430 ⇒ 01:06:58.790 Samuel Roberts: Alright, yeah, okay.

851 01:07:00.340 ⇒ 01:07:00.920 Samuel Roberts: Bye.

852 01:07:02.650 ⇒ 01:07:03.570 Samuel Roberts: Alright.

853 01:07:04.670 ⇒ 01:07:05.570 Mustafa Raja: Hey.

854 01:07:06.640 ⇒ 01:07:12.540 Samuel Roberts: Should we hop on a different Zoom link for our keeping track purposes? Yeah. Okay.

855 01:07:12.540 ⇒ 01:07:16.600 Mustafa Raja: Yeah, this is going to be counted as, stand-up, right?

856 01:07:16.910 ⇒ 01:07:18.569 Samuel Roberts: Yeah, so I’ll send you a Zoom link.

857 01:07:19.930 ⇒ 01:07:20.460 Samuel Roberts: Okay.

858 01:07:20.460 ⇒ 01:07:23.959 Mustafa Raja: Okay, I’ll join in 5 minutes. I’m going to grab some tea.

859 01:07:24.460 ⇒ 01:07:29.069 Samuel Roberts: Okay, I’ll just send it, and when you hop on, you’re good. I might make more coffee. Yeah, okay. Sounds good.

860 01:07:29.070 ⇒ 01:07:29.820 Mustafa Raja: Thank you.

861 01:07:29.820 ⇒ 01:07:30.380 Samuel Roberts: Yep.