Meeting Title: AI-App Standup Date: 2025-12-22 Meeting participants: Uttam Kumaran, Samuel Roberts, Mustafa Raja, Casie Aviles, Gabriel Lam
WEBVTT
1 00:00:13.020 ⇒ 00:00:14.040 Samuel Roberts: Hello.
2 00:00:15.330 ⇒ 00:00:16.190 Uttam Kumaran: Hi!
3 00:00:16.910 ⇒ 00:00:17.960 Samuel Roberts: How’s it going?
4 00:00:19.440 ⇒ 00:00:29.500 Uttam Kumaran: I’m good, dude, I feel a sense of, relaxation and euphoria going into… The holiday?
5 00:00:29.500 ⇒ 00:00:29.960 Samuel Roberts: Nice.
6 00:00:29.960 ⇒ 00:00:32.849 Uttam Kumaran: I think things are going very, very well.
7 00:00:33.110 ⇒ 00:00:35.289 Uttam Kumaran: Which, even just to say that.
8 00:00:35.830 ⇒ 00:00:36.310 Samuel Roberts: Yeah.
9 00:00:36.310 ⇒ 00:00:37.560 Uttam Kumaran: Like, chills?
10 00:00:38.020 ⇒ 00:00:42.950 Uttam Kumaran: But, like, I feel like we’re… we’re really on the… on the… on the right path, I think.
11 00:00:43.300 ⇒ 00:00:45.999 Uttam Kumaran: You know, the company, I feel like.
12 00:00:46.210 ⇒ 00:00:48.979 Uttam Kumaran: Is going to be excited to see a lot of the…
13 00:00:49.410 ⇒ 00:00:52.849 Uttam Kumaran: the… I hate when people are like, oh, there’s changes coming. It’s not like.
14 00:00:52.850 ⇒ 00:00:53.709 Samuel Roberts: You know what you mean.
15 00:00:53.710 ⇒ 00:01:02.179 Uttam Kumaran: I don’t know, it’s so hard in this company, like, I’m really trying to pick words properly, and also not use, like, corporate jargon, because…
16 00:01:03.230 ⇒ 00:01:05.750 Uttam Kumaran: I don’t know, you guys know me, I’m not, like, that type of person.
17 00:01:05.750 ⇒ 00:01:06.749 Samuel Roberts: I’m glad. Yeah.
18 00:01:06.750 ⇒ 00:01:23.319 Uttam Kumaran: We are tweaking and making things more well-defined in terms of, like, day-to-day project responsibilities, and also going to make more money available for folks that are taking those additional project responsibilities, which… which…
19 00:01:23.340 ⇒ 00:01:26.919 Uttam Kumaran: you know, for me and for everybody, I think I always tried to make that
20 00:01:27.020 ⇒ 00:01:41.309 Uttam Kumaran: super clear, which is, like, our job is going to be to have less people, and… and pay those people more. And I think we’re finally defining that path, and so I’m really, really happy, but…
21 00:01:41.720 ⇒ 00:01:54.480 Uttam Kumaran: to see stuff like Leela going well, to see us, like, getting, you know, additional scope on ABC, you know, I don’t know, just, it’s nice, so I feel pretty good.
22 00:01:54.760 ⇒ 00:02:14.859 Samuel Roberts: That’s awesome. Yeah, no, I can feel that a little bit. I would also say, like, having been on the side where you find yourself using that jargon, it made me realize, like, oh, this comes from a place that I actually used to, like… the jargon meant something at one point, and it’s become jargon. But it’s the same thing with, like, even, like, adding layers of bureaucracy or pieces of bureaucracy. I’m like, oh, there is a reason
23 00:02:15.280 ⇒ 00:02:28.939 Samuel Roberts: you have to do some of that stuff at a certain time, you know. Yes. And so, like, it’s one of those things where I always have an aversion to that, too, but, you know, you need to use that, you know, changes are… you know what I mean? There’s certain things that are like, well, yeah, it sounds cliche.
24 00:02:28.940 ⇒ 00:02:36.019 Uttam Kumaran: Yeah, I just know that… but, like, this is where I… I tend to try to think about what, from the industry.
25 00:02:36.270 ⇒ 00:02:45.350 Uttam Kumaran: are things that I should push on, because they were developed in a haphazard way, without care, and then what are things that…
26 00:02:45.680 ⇒ 00:02:50.110 Uttam Kumaran: I shouldn’t… we’re not, like, we’re not gonna innovate in every area.
27 00:02:50.530 ⇒ 00:02:53.830 Uttam Kumaran: But I am… I am constantly frustrated with
28 00:02:54.090 ⇒ 00:02:58.759 Uttam Kumaran: What the industry has set up in terms of expectations between
29 00:02:59.020 ⇒ 00:03:02.999 Uttam Kumaran: Someone in my position, and someone in the team’s position.
30 00:03:03.000 ⇒ 00:03:06.270 Samuel Roberts: In terms of the way we communicate, the way we, like.
31 00:03:06.270 ⇒ 00:03:10.650 Uttam Kumaran: set goals together. I think it’s… it’s, like, a horrible, like.
32 00:03:10.880 ⇒ 00:03:15.830 Uttam Kumaran: way that things have been done. Additionally, because of that, people come into our company with a lot of baggage.
33 00:03:15.970 ⇒ 00:03:35.649 Uttam Kumaran: And I hate to say… and I don’t want… and it is their baggage, but for me, my job is to show… is to remove that weight, right? To show that it’s a place where you can trust people, and that we are growing. And, like, it is hard. As someone who cares a lot, I have to deal with that baggage.
34 00:03:35.770 ⇒ 00:03:39.850 Uttam Kumaran: When people bring it from, you know, their previous places, but…
35 00:03:40.310 ⇒ 00:03:43.949 Uttam Kumaran: that’s okay, you know, I’ll continue to try to just, like, show through, like.
36 00:03:43.950 ⇒ 00:03:45.690 Samuel Roberts: Yeah, no, exactly, exactly.
37 00:03:45.690 ⇒ 00:03:46.420 Uttam Kumaran: You know?
38 00:03:46.700 ⇒ 00:03:47.670 Uttam Kumaran: So…
39 00:03:50.160 ⇒ 00:03:54.180 Samuel Roberts: I think that’s a good attitude to have about it, because it’s totally true.
40 00:03:54.540 ⇒ 00:03:55.380 Uttam Kumaran: Yeah.
41 00:03:55.380 ⇒ 00:04:02.610 Samuel Roberts: It’s not their fault that things have gone that way in other engagements with, you know, other places they’ve worked, but… yeah.
42 00:04:02.880 ⇒ 00:04:03.760 Samuel Roberts: It’s like…
43 00:04:03.970 ⇒ 00:04:11.499 Samuel Roberts: Like, a relationship, where you come in and you’ve been, like, hurt before, kind of thing, and you, like, don’t know how to trust, and it’s like, you know, you gotta establish that and make sure it’s, you know.
44 00:04:11.860 ⇒ 00:04:14.409 Samuel Roberts: Clear that it’s different, but yeah, that’s cool.
45 00:04:15.060 ⇒ 00:04:16.610 Samuel Roberts: Hey, Casey, Mustafa.
46 00:04:18.370 ⇒ 00:04:19.200 Mustafa Raja: A…
47 00:04:21.810 ⇒ 00:04:22.700 Casie Aviles: Hey, guys.
48 00:04:23.390 ⇒ 00:04:24.870 Samuel Roberts: How are things going today so far?
49 00:04:25.940 ⇒ 00:04:26.770 Mustafa Raja: Good…
50 00:04:35.980 ⇒ 00:04:39.579 Samuel Roberts: Who are we missing today? So, Panava’s not here,
51 00:04:39.760 ⇒ 00:04:43.100 Samuel Roberts: Today, I think he’ll be back tomorrow, and Gabe, where’s Gabe?
52 00:04:44.240 ⇒ 00:04:45.170 Samuel Roberts: Same.
53 00:04:51.140 ⇒ 00:04:52.709 Samuel Roberts: Yeah, where’s…
54 00:05:12.230 ⇒ 00:05:13.809 Samuel Roberts: Follow me, it’s messaging outline.
55 00:05:14.290 ⇒ 00:05:22.549 Samuel Roberts: Alright. Anyway, yeah, let’s talk, ABC, I guess, when we wait for Gabe, then.
56 00:05:23.020 ⇒ 00:05:25.570 Samuel Roberts: Updates from Friday…
57 00:05:27.480 ⇒ 00:05:35.380 Samuel Roberts: I saw… I saw the… the post with a bunch of things crossed out, which looked good, I guess. Can you give me, like, the TLDR on what happened?
58 00:05:36.390 ⇒ 00:05:37.870 Mustafa Raja: Yeah, so.
59 00:05:37.870 ⇒ 00:05:38.409 Casie Aviles: Oh, I’m.
60 00:05:38.410 ⇒ 00:05:53.770 Mustafa Raja: So, we discussed two things. One was, we should add, some more metadata where, Andy would be aware of the hierarchy of how the document is structured in terms of…
61 00:05:54.080 ⇒ 00:06:04.549 Mustafa Raja: Yeah, so I added that metadata, and then rerun all of the embeddings, so all prior embeddings would also have, that metadata.
62 00:06:05.140 ⇒ 00:06:05.740 Samuel Roberts: Cool.
63 00:06:06.160 ⇒ 00:06:11.100 Mustafa Raja: And then I, created this sheet, where I ran both…
64 00:06:11.270 ⇒ 00:06:21.880 Mustafa Raja: both the, Mastra agent that… that… that is exactly the edit and clone, and then the multi-agent version one. I’ll link…
65 00:06:21.980 ⇒ 00:06:24.050 Mustafa Raja: I’ll link the sheet in the Zoom also.
66 00:06:24.240 ⇒ 00:06:27.980 Samuel Roberts: Perfect, yeah, I had it open this morning, I was… I don’t know if I still have it open or not.
67 00:06:29.750 ⇒ 00:06:31.129 Samuel Roberts: Hmm, thank you.
68 00:06:32.220 ⇒ 00:06:44.549 Mustafa Raja: Yeah, so the, so I believe they, for the most part, it’s pretty good. The multi-agent one, the, the router prompt might need some adjustments.
69 00:06:44.830 ⇒ 00:06:49.010 Mustafa Raja: But for the most part, it looks… it looks good to me.
70 00:06:49.250 ⇒ 00:06:54.989 Mustafa Raja: So, I guess, we are at a point where we would want to commit to any one of those, right?
71 00:06:54.990 ⇒ 00:06:57.669 Samuel Roberts: Yep, hold on one second, it comes out…
72 00:06:57.980 ⇒ 00:07:02.070 Samuel Roberts: So, this is thumbs up and thumbs down stuff. How many are here? Sorry, I’m just trying to…
73 00:07:02.110 ⇒ 00:07:05.079 Mustafa Raja: Yes, 100-100. Both of them are 100.
74 00:07:05.080 ⇒ 00:07:06.389 Samuel Roberts: My 201, okay, cool.
75 00:07:07.750 ⇒ 00:07:12.619 Samuel Roberts: Great. Yeah, I mean… Did we do any…
76 00:07:13.900 ⇒ 00:07:16.449 Samuel Roberts: Overall check, or is it… are you just talking, like.
77 00:07:16.580 ⇒ 00:07:19.480 Samuel Roberts: Your… your gut based on how things looked so far.
78 00:07:19.480 ⇒ 00:07:26.229 Mustafa Raja: Yeah, so for now, it’s based on my gut. I did not do any sort of analysis on both of these.
79 00:07:26.230 ⇒ 00:07:26.760 Samuel Roberts: Camp.
80 00:07:26.760 ⇒ 00:07:29.660 Mustafa Raja: With the editing output. I could do that today.
81 00:07:31.320 ⇒ 00:07:40.489 Samuel Roberts: Yeah, let’s keep talking about what happened, we can plan ahead afterwards, but I think that might be a good idea, because especially for the ones that were thumbs up, we don’t want to progress.
82 00:07:41.100 ⇒ 00:07:41.480 Mustafa Raja: Oh, yeah.
83 00:07:42.020 ⇒ 00:07:47.809 Samuel Roberts: So that would be, hopefully, an easy one to check, at least. And then the other ones, might be a little more.
84 00:07:49.900 ⇒ 00:07:54.200 Samuel Roberts: I think that’s fine to say, okay. Yeah, I could… I could add an AI step.
85 00:07:54.200 ⇒ 00:08:03.730 Mustafa Raja: Here, where, we judge, both of these with the… for thumbs up, we could judge both of these based on the output.
86 00:08:06.400 ⇒ 00:08:10.019 Samuel Roberts: Yeah, I think that at least makes sense as a first pass for the good ones.
87 00:08:10.470 ⇒ 00:08:11.410 Samuel Roberts: Yeah.
88 00:08:12.780 ⇒ 00:08:18.670 Samuel Roberts: The bad ones, we have to do a little more digging ourselves to figure out what the proper response is, and if the new ones match it, but…
89 00:08:19.740 ⇒ 00:08:21.530 Samuel Roberts: You can do that, too. Okay.
90 00:08:21.940 ⇒ 00:08:27.780 Samuel Roberts: So that’s… so then we have those two to compare, the N8N clone and the multi-agent.
91 00:08:29.380 ⇒ 00:08:33.640 Samuel Roberts: And you said there’s a little routing stuff you wanted to do in the multi-agent, is that the other thing you said?
92 00:08:34.320 ⇒ 00:08:35.520 Mustafa Raja: Sorry, could you say that again?
93 00:08:35.520 ⇒ 00:08:38.139 Samuel Roberts: The routing step for the multi-agent, you mentioned something about.
94 00:08:38.140 ⇒ 00:08:46.580 Mustafa Raja: Oh, yeah, so, so, I saw that, multi-agents sometimes, you know, would route, a question that would
95 00:08:46.780 ⇒ 00:08:54.770 Mustafa Raja: clearly be for, DB database agent, but it will route it to the vector agent, or the general agent.
96 00:08:55.080 ⇒ 00:09:05.639 Mustafa Raja: Most of the times, this mistake is happening when the question is for the vector agent, but it goes to the router… sorry, goes to the general information agent.
97 00:09:06.030 ⇒ 00:09:08.010 Samuel Roberts: Hmm, okay.
98 00:09:13.250 ⇒ 00:09:14.040 Samuel Roberts: Okay.
99 00:09:14.840 ⇒ 00:09:19.499 Samuel Roberts: Yeah, the differences between different agents, I guess?
100 00:09:20.130 ⇒ 00:09:20.830 Mustafa Raja: Sorry?
101 00:09:24.050 ⇒ 00:09:28.179 Samuel Roberts: Nothing, nothing, I was just, I was just thinking through, okay. Yeah, a little more context, and then…
102 00:09:29.400 ⇒ 00:09:34.249 Samuel Roberts: Is this still set up the way that it had been, or is it set up with the branching logic?
103 00:09:36.150 ⇒ 00:09:38.729 Mustafa Raja: Where it’s like the agent makes the.
104 00:09:38.730 ⇒ 00:09:43.240 Samuel Roberts: decision, and then the next agent, the next agent? Okay. Okay, that’s fine. I don’t think it’s invalid.
105 00:09:43.240 ⇒ 00:09:50.010 Mustafa Raja: the workflow, but yeah, if I’m going to, you know, work on the prompt, I’ll just, adjust that too, because it’s a small change.
106 00:09:50.430 ⇒ 00:09:56.600 Samuel Roberts: Yeah, I think it’s functionally the same thing, I just think it might display better.
107 00:09:56.870 ⇒ 00:09:57.740 Samuel Roberts: And… and…
108 00:09:57.740 ⇒ 00:10:00.250 Mustafa Raja: Yeah, it looks visually better the other way.
109 00:10:02.820 ⇒ 00:10:03.380 Samuel Roberts: Cool.
110 00:10:04.680 ⇒ 00:10:05.970 Samuel Roberts: Okay, great.
111 00:10:06.970 ⇒ 00:10:11.669 Samuel Roberts: Awesome. What else?
112 00:10:14.230 ⇒ 00:10:17.360 Samuel Roberts: Casey, I saw your update at the end of the day Friday.
113 00:10:17.360 ⇒ 00:10:21.270 Casie Aviles: Oh, yeah, yeah, so for ABC, Whoa.
114 00:10:21.270 ⇒ 00:10:22.000 Samuel Roberts: Sorry, go ahead.
115 00:10:22.230 ⇒ 00:10:25.140 Casie Aviles: Yeah, it’s just about adding, like, the…
116 00:10:25.430 ⇒ 00:10:29.390 Casie Aviles: Missing departments that… for the technicians that they…
117 00:10:29.820 ⇒ 00:10:32.409 Casie Aviles: That they gave us, so…
118 00:10:32.780 ⇒ 00:10:36.070 Casie Aviles: We were just learning that there were other technicians
119 00:10:36.950 ⇒ 00:10:43.860 Casie Aviles: that they wanted to add, so I just mainly handled that. The only one that I need to add
120 00:10:44.420 ⇒ 00:10:51.059 Casie Aviles: I believe the only one remaining is the home improvement, so I can work on that today.
121 00:10:52.030 ⇒ 00:10:55.850 Casie Aviles: Yeah, that’s pretty much, like, the updates from last week.
122 00:10:56.040 ⇒ 00:10:57.259 Casie Aviles: For this week.
123 00:10:58.040 ⇒ 00:11:02.160 Casie Aviles: We have a couple planned in the Gantt chart.
124 00:11:03.150 ⇒ 00:11:09.780 Casie Aviles: I think the main ones for me are… Let’s see,
125 00:11:11.400 ⇒ 00:11:18.539 Casie Aviles: Yeah, there’s one here where we have to, like, implement the Mastra…
126 00:11:18.720 ⇒ 00:11:21.860 Casie Aviles: endpoint to… in the Google environment, so…
127 00:11:23.060 ⇒ 00:11:29.650 Casie Aviles: Yeah, I think that’s something I can look into, so… The gifted.
128 00:11:30.280 ⇒ 00:11:38.310 Casie Aviles: if we can, like, you know, access the… or we can hit the endpoint in Google Chat, so we’re… yeah, I mean, in Google Cloud, so…
129 00:11:39.200 ⇒ 00:11:42.830 Casie Aviles: We could probably use, Google Cloud Run for that.
130 00:11:43.790 ⇒ 00:11:45.929 Casie Aviles: That’s something I can take a look at.
131 00:11:46.900 ⇒ 00:11:52.920 Casie Aviles: Yeah, I think… I think that’s… that’s for the week. There’s also, like, the concern of, like, the forms.
132 00:11:54.270 ⇒ 00:12:00.590 Casie Aviles: I don’t… I’m not sure if this is something we can finish within the week, but…
133 00:12:00.760 ⇒ 00:12:02.460 Uttam Kumaran: Basically, the…
134 00:12:02.760 ⇒ 00:12:08.940 Casie Aviles: Basically, for the forms, we’re seeing that there is a lot of friction, Since, you know,
135 00:12:09.230 ⇒ 00:12:16.549 Casie Aviles: Janice wasn’t able to use it as effectively, and even myself, right, last week, when I was adding these new
136 00:12:17.320 ⇒ 00:12:19.130 Casie Aviles: Technicians.
137 00:12:20.260 ⇒ 00:12:25.739 Casie Aviles: You know, it’s starting to… you know, show the limits, the innate N forms, so…
138 00:12:26.250 ⇒ 00:12:29.170 Casie Aviles: I think an idea that was raised was
139 00:12:30.750 ⇒ 00:12:35.590 Casie Aviles: Creating, like, a more custom form for them?
140 00:12:36.770 ⇒ 00:12:37.630 Casie Aviles: So…
141 00:12:38.160 ⇒ 00:12:43.109 Uttam Kumaran: Yeah, I mean, Mustafa, can you mind pulling up the Gantt? I just want to take one at a time.
142 00:12:43.290 ⇒ 00:12:51.109 Uttam Kumaran: Like, I feel like forms is a really good thing for us to focus on.
143 00:12:51.920 ⇒ 00:12:55.730 Uttam Kumaran: Versus others, because it directly affects the client.
144 00:12:56.120 ⇒ 00:13:00.769 Uttam Kumaran: Like, what do you think? I don’t want to take on 5 things and slowly inch.
145 00:13:00.900 ⇒ 00:13:09.989 Uttam Kumaran: like, if we have a through line to moving forms to something new, then maybe we should take that. I don’t know. Or we should just, again, continue to go…
146 00:13:10.760 ⇒ 00:13:11.570 Uttam Kumaran: like…
147 00:13:12.550 ⇒ 00:13:18.290 Uttam Kumaran: I guess what I’m… what I want to know is, like, okay, are we able to work on all these in parallel, and we think we can
148 00:13:18.420 ⇒ 00:13:20.799 Uttam Kumaran: Complete each of these, or should we just take
149 00:13:21.910 ⇒ 00:13:24.730 Uttam Kumaran: Forms and… and close that out.
150 00:13:25.970 ⇒ 00:13:26.880 Uttam Kumaran: Burst.
151 00:13:28.700 ⇒ 00:13:32.999 Casie Aviles: Yeah, I think… Probably the forms would be…
152 00:13:33.480 ⇒ 00:13:36.539 Casie Aviles: the most valuable, I guess, or most impactful.
153 00:13:39.680 ⇒ 00:13:40.440 Uttam Kumaran: Okay.
154 00:13:40.980 ⇒ 00:13:45.109 Casie Aviles: So, for the week, if I can focus on that, I guess, then…
155 00:13:46.240 ⇒ 00:13:53.930 Casie Aviles: Yeah, I think I just need to think about, like, how this form… how these new forms would look like, you know?
156 00:13:54.670 ⇒ 00:13:59.830 Casie Aviles: And… Yeah, between Lash.
157 00:14:09.920 ⇒ 00:14:13.739 Uttam Kumaran: But, like, I guess, like, what do you guys… what do you guys think? Should we just move
158 00:14:14.110 ⇒ 00:14:19.809 Uttam Kumaran: a couple of other things off to next week, and just focus on forums this week. Like, for me.
159 00:14:21.080 ⇒ 00:14:28.309 Uttam Kumaran: I want to just know, like, okay, can we… if I’m like, let’s go after forms, we really feel confident we can get to somewhere by Friday.
160 00:14:28.730 ⇒ 00:14:30.429 Uttam Kumaran: Or, like, at least by Monday.
161 00:14:33.800 ⇒ 00:14:34.950 Samuel Roberts: Oh, I’m…
162 00:14:40.540 ⇒ 00:14:41.280 Mustafa Raja: I think we…
163 00:14:45.250 ⇒ 00:14:54.940 Casie Aviles: Yeah, I think I need to… I probably would need to go over with the team and probably talk about the specifics of how we would implement this.
164 00:14:55.480 ⇒ 00:14:59.499 Casie Aviles: Because I don’t… I don’t want to, like, create new forms and then…
165 00:15:00.550 ⇒ 00:15:06.030 Casie Aviles: And then it doesn’t end up, you know, delivering… Any… any impact?
166 00:15:09.200 ⇒ 00:15:11.050 Samuel Roberts: Hey, sorry y’all, I think I’m back now.
167 00:15:11.050 ⇒ 00:15:11.470 Uttam Kumaran: Okay.
168 00:15:11.480 ⇒ 00:15:17.099 Samuel Roberts: I don’t know if my internet, it just keeps… Zoom just keeps disappearing and reappearing, so I think there’s got to be some connection issue.
169 00:15:17.270 ⇒ 00:15:22.589 Samuel Roberts: But… We were talking the forms, is that…
170 00:15:22.590 ⇒ 00:15:23.300 Uttam Kumaran: Yes.
171 00:15:23.300 ⇒ 00:15:24.040 Mustafa Raja: Junior.
172 00:15:24.640 ⇒ 00:15:39.039 Samuel Roberts: Yeah, my thought is to, like, make some custom forms, that can pull from Supabase, so that they have access to everyone who’s in there, all the jobs that are in there, if they need to create a new one. Basically, like, making a new little CRUD interface for them.
173 00:15:39.040 ⇒ 00:15:39.640 Casie Aviles: Yeah.
174 00:15:40.540 ⇒ 00:15:44.980 Samuel Roberts: And so… Well, I missed what you were saying about… worried about something.
175 00:15:45.500 ⇒ 00:15:50.650 Uttam Kumaran: Well, I guess, like, should we just take forms and close it? Because right now, if you look at the Gantt.
176 00:15:50.760 ⇒ 00:15:54.190 Uttam Kumaran: we have, like, if you scroll up, Mustafa.
177 00:15:54.530 ⇒ 00:15:58.460 Uttam Kumaran: We have, like, what, 4 or 5 items that we’re working on in parallel?
178 00:15:58.460 ⇒ 00:15:58.890 Samuel Roberts: Yeah.
179 00:15:58.890 ⇒ 00:16:07.490 Uttam Kumaran: I guess what I was… I’m not hearing confidence from Casey that, like, we can… that we understand how long Quietum 46 is gonna take.
180 00:16:07.490 ⇒ 00:16:08.290 Samuel Roberts: Sure.
181 00:16:08.610 ⇒ 00:16:11.500 Uttam Kumaran: And, like, what complete looks like.
182 00:16:12.540 ⇒ 00:16:15.880 Uttam Kumaran: You know, so…
183 00:16:16.100 ⇒ 00:16:34.600 Uttam Kumaran: My question is, like, if… if we don’t know what the acceptance criteria is, and we don’t know what the path is, then I want this to reflect, like, an accurate thing of, like, okay, we are working on this today, and we are gonna… I mean, right now, it’s just booked for two days. Yeah. One, it doesn’t seem like that’s accurate, right?
184 00:16:35.650 ⇒ 00:16:36.100 Samuel Roberts: Right.
185 00:16:36.100 ⇒ 00:16:39.889 Uttam Kumaran: We’re probably gonna need more than just these two days to finish this out.
186 00:16:42.760 ⇒ 00:16:43.760 Samuel Roberts: Yes, I think so.
187 00:16:43.760 ⇒ 00:16:48.929 Uttam Kumaran: So, give me, like, the conservative estimate to say if we were to complete 46,
188 00:16:49.570 ⇒ 00:16:52.299 Uttam Kumaran: When would that be? And then…
189 00:16:52.420 ⇒ 00:16:55.099 Uttam Kumaran: I would propose just working on that.
190 00:16:56.010 ⇒ 00:17:06.559 Uttam Kumaran: Because that seems… for me, that seems to have… unless someone else has a… has another priority that’s more important, that I’m happy to debate. But I would like this to be a little bit of, like, a debate on, like, what we’re gonna work on.
191 00:17:07.859 ⇒ 00:17:08.900 Uttam Kumaran: Deep in air.
192 00:17:08.900 ⇒ 00:17:09.960 Samuel Roberts: size 8.
193 00:17:10.480 ⇒ 00:17:15.229 Samuel Roberts: Yeah, I mean, I think the forms… so, it’s forms, it’s links service areas…
194 00:17:15.490 ⇒ 00:17:22.579 Samuel Roberts: And then, improved metadata that’s done. Okay, and then the top was the…
195 00:17:23.480 ⇒ 00:17:35.409 Samuel Roberts: Rebuild, routing logic, which… okay, and we’re finishing up the test different architectures. Okay. Yeah, I mean, forms is definitely something I want to focus on. I’m not sure what… the other one here is link service areas. I don’t think that’s a huge…
196 00:17:37.130 ⇒ 00:17:44.840 Casie Aviles: Yeah, that’s smartph, just… a quick, you know, change in the superbase tables, I believe.
197 00:17:47.650 ⇒ 00:17:51.830 Samuel Roberts: Yeah, I mean, I think forms is the big one here, and it probably is more than two days, you’re absolutely right.
198 00:17:51.830 ⇒ 00:17:52.450 Casie Aviles: Yeah.
199 00:17:52.900 ⇒ 00:17:56.810 Samuel Roberts: But I think that is definitely the priority here, based on those ones.
200 00:17:56.990 ⇒ 00:17:58.190 Samuel Roberts: Because that affects…
201 00:17:58.470 ⇒ 00:18:06.579 Samuel Roberts: pre- and post-migration stuff right now, anyway, if that, you know… like, it’s gonna affect them being able to update things currently in Andy, and in…
202 00:18:06.580 ⇒ 00:18:07.790 Casie Aviles: migrated Andy.
203 00:18:09.420 ⇒ 00:18:17.120 Samuel Roberts: So I don’t… I don’t really have an argument against doing that, because that’s what I probably would have said. The link… the service areas and service names, I think, can happen in parallel, because I don’t think that’s…
204 00:18:17.650 ⇒ 00:18:23.690 Uttam Kumaran: My question is, tell me your conservative estimate on, like, how much time is needed to close at 46.
205 00:18:25.370 ⇒ 00:18:28.530 Uttam Kumaran: Conservative, meaning the, like, longest.
206 00:18:29.270 ⇒ 00:18:37.799 Uttam Kumaran: Because again, I’d rather you guys say a time and then do it faster, but what I don’t want… what I would… what I don’t want is we say a time that takes longer.
207 00:18:37.800 ⇒ 00:18:42.020 Samuel Roberts: Yeah, I mean, in a non-holiday week, I would say end of the week, but… so that would be, what, then…
208 00:18:42.020 ⇒ 00:18:42.660 Casie Aviles: Yeah.
209 00:18:43.010 ⇒ 00:18:44.620 Uttam Kumaran: So then just drag it to the 30th.
210 00:18:44.820 ⇒ 00:18:46.320 Samuel Roberts: Drag it to… yeah, that’s fine.
211 00:18:47.030 ⇒ 00:18:50.539 Uttam Kumaran: But jackets, but this is, like, what I want you guys, like… Yeah.
212 00:18:50.840 ⇒ 00:18:56.259 Uttam Kumaran: I want you guys to think this way. Like, don’t over-promise, under-deliver. Opposite, you know?
213 00:18:57.380 ⇒ 00:19:09.290 Uttam Kumaran: Because today, I want to go to the client and say, what are we working on, and what can they expect, given the holiday week? And if I go to them, and I’m like, hey, we’re gonna come up with the new forms by the end of the year.
214 00:19:10.020 ⇒ 00:19:16.919 Uttam Kumaran: I do want to hit that. Like, I do want… I do want to… I want to have confidence that that is the latest.
215 00:19:17.110 ⇒ 00:19:22.200 Uttam Kumaran: Because I’d rather say it’s gonna be out by mid-Jan, And then hit jam.
216 00:19:22.480 ⇒ 00:19:25.199 Uttam Kumaran: I don’t want to say Jan, and then hit mid-Jan.
217 00:19:25.200 ⇒ 00:19:32.410 Samuel Roberts: Yeah, no, that’s fair. I mean, and with the… with… I mean, we have two days blocked off on each of the holidays there, but I will probably not be…
218 00:19:32.570 ⇒ 00:19:35.120 Uttam Kumaran: No, I think a lot of… no, a lot of the folks are off.
219 00:19:35.120 ⇒ 00:19:37.300 Samuel Roberts: That’s why I wanted to inspire, yeah.
220 00:19:37.300 ⇒ 00:19:43.970 Uttam Kumaran: So, so this is also, like, consider, consider that, that not everybody’s gonna be on. Yeah. So…
221 00:19:45.850 ⇒ 00:19:49.149 Uttam Kumaran: Yeah. Hold on, let me pull up there. So this is…
222 00:19:49.200 ⇒ 00:19:59.589 Samuel Roberts: Week 1… Yeah, first week of January, what does that put us on the week?
223 00:20:04.550 ⇒ 00:20:09.280 Samuel Roberts: Yeah, I mean, if you wanna, you know, Really conservative than the 10th.
224 00:20:09.790 ⇒ 00:20:10.960 Samuel Roberts: Probably good.
225 00:20:11.260 ⇒ 00:20:14.680 Samuel Roberts: Just because I know we’ll be back and running with it.
226 00:20:17.370 ⇒ 00:20:18.779 Uttam Kumaran: Okay, let’s do that.
227 00:20:19.590 ⇒ 00:20:32.739 Uttam Kumaran: So then, what I’m gonna say is I would like to go through this Gantt chart with what I just said, because we just took something that was listed as 2 days, and now it’s 3 weeks. So…
228 00:20:32.990 ⇒ 00:20:36.039 Uttam Kumaran: Is the rest of this doc, like, accurate?
229 00:20:38.540 ⇒ 00:20:40.759 Samuel Roberts: What do we have? Okay, so…
230 00:20:40.760 ⇒ 00:20:44.329 Uttam Kumaran: Because there are… there are certain things in here that are only listed for 2 days.
231 00:20:44.910 ⇒ 00:20:49.619 Uttam Kumaran: and I… and I just want to confirm that, like, ev- like, this should be, like.
232 00:20:50.110 ⇒ 00:20:58.279 Uttam Kumaran: And think of me as, like, if I was a product manager on this client, I… this needs to really represent what I can go to a client and be like, here’s what we’re gonna hit.
233 00:20:58.460 ⇒ 00:21:08.520 Uttam Kumaran: Otherwise, you’re gonna put me in a spot where I’m gonna say something to ABC, and then we’re not gonna hit it, you know? So, I… my ask for you guys on the technical side is to tell me what the technical…
234 00:21:08.980 ⇒ 00:21:11.629 Uttam Kumaran: Timeline is on all these items.
235 00:21:12.400 ⇒ 00:21:14.669 Samuel Roberts: Back to the FaceTime. Okay.
236 00:21:17.190 ⇒ 00:21:23.240 Samuel Roberts: Yeah, I can probably give this a going over after, unless you want to go through it all now, but…
237 00:21:24.490 ⇒ 00:21:29.639 Uttam Kumaran: I mean, can you… can we at least plan out what is going to be worked on until the end of the year?
238 00:21:30.800 ⇒ 00:21:31.830 Uttam Kumaran: Now?
239 00:21:34.730 ⇒ 00:21:40.930 Samuel Roberts: Yeah. So I think we can finish up that testing, that we talked about, Mustafa.
240 00:21:41.450 ⇒ 00:21:42.120 Mustafa Raja: Yep.
241 00:21:43.070 ⇒ 00:21:45.770 Samuel Roberts: So, I mean, end of the year, we’re talking through
242 00:21:46.380 ⇒ 00:21:50.890 Samuel Roberts: Tomorrow, or are people gonna be on? I don’t wanna make assumptions about what people are doing here, but…
243 00:21:51.980 ⇒ 00:21:54.580 Uttam Kumaran: I guess Casey Mustafa, are you guys gonna be on at all next week?
244 00:21:54.580 ⇒ 00:21:55.950 Samuel Roberts: What are you guys thinking?
245 00:21:56.610 ⇒ 00:22:00.680 Mustafa Raja: Yeah, for 24th and 25th, I might be off, but…
246 00:22:01.040 ⇒ 00:22:04.030 Samuel Roberts: But yeah, back, the rest of the time, I’ll be out.
247 00:22:05.020 ⇒ 00:22:06.390 Samuel Roberts: Did you last my voice?
248 00:22:06.390 ⇒ 00:22:14.740 Casie Aviles: Yeah, I’ll be on… After Christmas, so 26. Next week, yes, I’ll be on, except for the first.
249 00:22:15.170 ⇒ 00:22:15.830 Samuel Roberts: Okay.
250 00:22:16.800 ⇒ 00:22:17.809 Samuel Roberts: Now, open up.
251 00:22:17.810 ⇒ 00:22:19.839 Uttam Kumaran: And I’ll be, like, just kind of…
252 00:22:20.090 ⇒ 00:22:22.830 Uttam Kumaran: Yeah, I’ll be on, but, like, it’s…
253 00:22:23.770 ⇒ 00:22:27.990 Uttam Kumaran: So that’s why I think consider that, and let’s think about what we can get done.
254 00:22:28.230 ⇒ 00:22:33.199 Uttam Kumaran: I’m not saying… like, I don’t want this to be, like, we have to get every… I just need an accurate view, so I can say…
255 00:22:33.200 ⇒ 00:22:33.600 Samuel Roberts: Right, right.
256 00:22:33.600 ⇒ 00:22:34.419 Uttam Kumaran: to that, yeah.
257 00:22:34.830 ⇒ 00:22:38.509 Samuel Roberts: Yeah, I mean, the test agent makes sense, the rebuild routing logic…
258 00:22:39.180 ⇒ 00:22:50.120 Samuel Roberts: As master routing logic? What is this? Which routing logic are we talking there? We’re talking the one we discussed earlier about… Nope, this is, so we added the new routing logic in it, and right?
259 00:22:50.120 ⇒ 00:22:51.320 Mustafa Raja: This is that.
260 00:22:51.320 ⇒ 00:22:52.510 Samuel Roberts: Yes, okay, okay.
261 00:22:52.510 ⇒ 00:22:54.679 Mustafa Raja: For the vector… vector agent.
262 00:22:54.840 ⇒ 00:23:00.619 Samuel Roberts: Okay, that makes sense. And that leads to implement master entry endpoint for main request handling.
263 00:23:05.440 ⇒ 00:23:06.460 Samuel Roberts: Okay.
264 00:23:06.810 ⇒ 00:23:12.060 Samuel Roberts: Monster Entry Point,
265 00:23:12.880 ⇒ 00:23:17.290 Samuel Roberts: I’m a little nervous about doing that midweek then, but we can get it going.
266 00:23:18.900 ⇒ 00:23:21.520 Samuel Roberts: Is that what you’re talking GCP, Casey, for 5?
267 00:23:22.170 ⇒ 00:23:23.060 Casie Aviles: Yes, yes.
268 00:23:23.060 ⇒ 00:23:24.160 Mustafa Raja: Yes.
269 00:23:26.100 ⇒ 00:23:28.040 Samuel Roberts: You got, what, 3 days on that?
270 00:23:29.330 ⇒ 00:23:30.449 Samuel Roberts: Probably alright.
271 00:23:30.990 ⇒ 00:23:35.060 Samuel Roberts: Everything else after that, I don’t have a good sense of yet.
272 00:23:35.830 ⇒ 00:23:38.390 Samuel Roberts: But that gets us through end of the year, right?
273 00:23:38.790 ⇒ 00:23:39.999 Samuel Roberts: So we could get the monster.
274 00:23:40.000 ⇒ 00:23:40.730 Mustafa Raja: Oh, yes.
275 00:23:40.730 ⇒ 00:23:42.529 Samuel Roberts: stood up on our own GCP, maybe?
276 00:23:43.500 ⇒ 00:23:46.000 Mustafa Raja: Well, we have their GCP, though. The…
277 00:23:46.000 ⇒ 00:23:47.350 Samuel Roberts: Team set it up.
278 00:23:47.920 ⇒ 00:23:48.630 Casie Aviles: Yes.
279 00:23:49.140 ⇒ 00:23:51.179 Casie Aviles: Oh, okay. We have access to those.
280 00:23:51.550 ⇒ 00:23:53.269 Samuel Roberts: That’s right. Okay. Yeah, then I think…
281 00:23:54.050 ⇒ 00:23:58.589 Samuel Roberts: I don’t know about by end of the year, but then by the second. This doesn’t seem off for those three.
282 00:24:00.020 ⇒ 00:24:00.760 Mustafa Raja: Yeah.
283 00:24:00.760 ⇒ 00:24:02.990 Samuel Roberts: Does that seem accurate to you guys, Mustafa Casey?
284 00:24:03.710 ⇒ 00:24:08.780 Mustafa Raja: Yes, this looks good, because this piece we already have in, and then we just have to translate it to.
285 00:24:08.780 ⇒ 00:24:11.120 Samuel Roberts: Yeah, that shouldn’t be… it seems fine for that.
286 00:24:12.730 ⇒ 00:24:13.570 Samuel Roberts: Okay.
287 00:24:15.500 ⇒ 00:24:22.190 Samuel Roberts: So, end of the year, we’ll be not quite getting it up and running on their GCP as a…
288 00:24:22.860 ⇒ 00:24:24.030 Samuel Roberts: As the request handling?
289 00:24:24.030 ⇒ 00:24:25.590 Mustafa Raja: Yeah.
290 00:24:25.590 ⇒ 00:24:29.420 Samuel Roberts: Okay. Well, that makes sense, because I wouldn’t want to flip the switch anyway before we’re all back, so…
291 00:24:29.840 ⇒ 00:24:30.500 Mustafa Raja: Yeah.
292 00:24:31.100 ⇒ 00:24:31.930 Samuel Roberts: Okay.
293 00:24:32.220 ⇒ 00:24:34.880 Samuel Roberts: Scroll down a little bit for me, just while we’ve got that.
294 00:24:35.740 ⇒ 00:24:39.899 Mustafa Raja: For this section, or do you want to look at these ones?
295 00:24:40.350 ⇒ 00:24:43.490 Samuel Roberts: I don’t want to go past the end of the year yet. I want to look at what else.
296 00:24:43.490 ⇒ 00:24:44.080 Mustafa Raja: Okay.
297 00:24:44.080 ⇒ 00:24:50.069 Samuel Roberts: bottom, what we had going through there. So we just did the forms, the link service area… wait, why did we move the link service areas?
298 00:24:53.550 ⇒ 00:24:56.189 Mustafa Raja: Salik, do we not want to move it?
299 00:24:56.900 ⇒ 00:25:00.830 Samuel Roberts: I mean, I… is that something that can happen? There’s no dependency on that, right?
300 00:25:01.750 ⇒ 00:25:04.730 Casie Aviles: I don’t think it’s super urgent. Okay. That’s not true.
301 00:25:04.730 ⇒ 00:25:17.459 Samuel Roberts: Yeah, then next week’s fine. And then the forms, yeah. Add up to the home improvement technicians, that’s, okay, that’s also what we discussed. Yeah, I mean, I think everything that we just discussed, SANS forms, makes sense through end of the year.
302 00:25:22.340 ⇒ 00:25:23.110 Samuel Roberts: Yeah.
303 00:25:23.370 ⇒ 00:25:27.239 Samuel Roberts: Alright, so then we can go through the ones in January after.
304 00:25:27.360 ⇒ 00:25:29.400 Samuel Roberts: And get a more accurate Gantt going.
305 00:25:33.950 ⇒ 00:25:35.429 Samuel Roberts: Any other thoughts, guys?
306 00:25:35.660 ⇒ 00:25:37.279 Samuel Roberts: Does that sound like a plan for…
307 00:25:37.440 ⇒ 00:25:38.179 Samuel Roberts: This weekend.
308 00:25:38.180 ⇒ 00:25:40.350 Mustafa Raja: This sounds like a… this sounds like a plan.
309 00:25:40.350 ⇒ 00:25:40.770 Samuel Roberts: Okay.
310 00:25:40.770 ⇒ 00:25:44.919 Uttam Kumaran: Can I add one more item, is, establishing BigQuery.
311 00:25:46.480 ⇒ 00:25:47.380 Samuel Roberts: Mmm, okay.
312 00:25:47.380 ⇒ 00:25:48.030 Mustafa Raja: Yeah…
313 00:25:49.460 ⇒ 00:25:50.210 Uttam Kumaran: So…
314 00:25:50.210 ⇒ 00:25:54.800 Mustafa Raja: Did you… It should be migrating from Snowflake, right?
315 00:25:55.550 ⇒ 00:25:59.839 Uttam Kumaran: Yeah, so let’s add two tasks here. One is gonna be initializing BigQuery.
316 00:26:01.650 ⇒ 00:26:05.049 Samuel Roberts: Is this BigQuery for the ZipsDB, or for the logging?
317 00:26:05.690 ⇒ 00:26:06.630 Uttam Kumaran: logging.
318 00:26:06.630 ⇒ 00:26:07.230 Samuel Roberts: Okay.
319 00:26:07.400 ⇒ 00:26:15.969 Uttam Kumaran: And I need this for the discovery work, too. So in it, BigQuery, let’s also put migrate logging from Snowflake to BigQuery.
320 00:26:18.060 ⇒ 00:26:21.459 Uttam Kumaran: Yeah. And then we also are gonna migrate RIL
321 00:26:22.110 ⇒ 00:26:25.840 Uttam Kumaran: from pointing to Snowflake to BQ,
322 00:26:31.190 ⇒ 00:26:37.260 Uttam Kumaran: And then we’re most likely gonna also have to, established DBT.
323 00:26:37.700 ⇒ 00:26:38.370 Samuel Roberts: Hmm.
324 00:26:40.110 ⇒ 00:26:41.320 Uttam Kumaran: on BigQuery.
325 00:26:43.060 ⇒ 00:26:46.879 Uttam Kumaran: So you can just say init dbt or something on BQ.
326 00:26:47.680 ⇒ 00:26:55.749 Uttam Kumaran: And all of these, I think, Mustafa, you have, like, all the experience to do these, so if you want to do these, or if you want to split with KC, but I feel like you kind of have…
327 00:26:56.980 ⇒ 00:27:05.699 Uttam Kumaran: Awareness here, so, I think in it, BQ, can we do by end of next week?
328 00:27:06.360 ⇒ 00:27:07.970 Uttam Kumaran: Let’s just drag that.
329 00:27:09.490 ⇒ 00:27:10.670 Samuel Roberts: Yeah. And then…
330 00:27:11.980 ⇒ 00:27:14.549 Uttam Kumaran: So, meaning from, like.
331 00:27:15.240 ⇒ 00:27:20.969 Uttam Kumaran: You could just do 22nd to the f- to basically the 31st, and then same with init dbt on BQ.
332 00:27:22.040 ⇒ 00:27:26.689 Uttam Kumaran: But both of these are the most important, and then the rest you could do…
333 00:27:27.430 ⇒ 00:27:30.739 Uttam Kumaran: we could basically do first week of Jan.
334 00:27:31.910 ⇒ 00:27:34.680 Uttam Kumaran: So, the first one… Yeah, yeah, yeah.
335 00:27:35.450 ⇒ 00:27:39.159 Uttam Kumaran: So that one, yep, and then the 113.
336 00:27:39.950 ⇒ 00:27:44.140 Uttam Kumaran: You can do probably halfway until… the second.
337 00:27:44.400 ⇒ 00:27:47.959 Uttam Kumaran: I don’t think it’s gonna take… it’s probably gonna take 2 days.
338 00:27:48.860 ⇒ 00:27:49.930 Uttam Kumaran: Max.
339 00:27:49.930 ⇒ 00:27:50.640 Mustafa Raja: Okay.
340 00:27:51.430 ⇒ 00:27:54.440 Mustafa Raja: Yeah. So this is after that I tried to…
341 00:27:54.440 ⇒ 00:28:00.110 Uttam Kumaran: Yeah, so, I mean, I would put… I’d probably put it to, like, starting midweek next week. Yeah, yeah, yeah.
342 00:28:01.270 ⇒ 00:28:07.860 Uttam Kumaran: Okay, cool, and then… and then you can just basically move 10 and 11 to sometime in the… in the first week of Jan.
343 00:28:11.300 ⇒ 00:28:14.850 Samuel Roberts: with Dylan Greece.
344 00:28:15.710 ⇒ 00:28:17.699 Mustafa Raja: 1011 or 11 and 12?
345 00:28:20.650 ⇒ 00:28:23.710 Uttam Kumaran: 11 and 12 are the ones that are gonna be first week of Jan.
346 00:28:23.710 ⇒ 00:28:25.590 Samuel Roberts: Yeah, 10 needs to be above 9.
347 00:28:32.560 ⇒ 00:28:33.989 Samuel Roberts: I said that there’s a number
348 00:28:37.730 ⇒ 00:28:42.310 Samuel Roberts: Yes, that’s it, yeah, that’s it. Okay, I have the right one, and the color is Fawn Malton.
349 00:28:43.880 ⇒ 00:28:45.539 Samuel Roberts: That’s what you wanted, correct?
350 00:28:46.490 ⇒ 00:28:47.920 Samuel Roberts: Yes, that’s good.
351 00:28:50.220 ⇒ 00:28:51.260 Uttam Kumaran: Okay, cool.
352 00:28:51.500 ⇒ 00:28:55.060 Uttam Kumaran: Anything else?
353 00:29:00.330 ⇒ 00:29:01.310 Samuel Roberts: On ABC.
354 00:29:02.670 ⇒ 00:29:03.330 Uttam Kumaran: Yeah.
355 00:29:03.700 ⇒ 00:29:08.999 Samuel Roberts: I don’t have anything else. I’ll go through some of the Gantt post-New Year.
356 00:29:09.190 ⇒ 00:29:09.840 Uttam Kumaran: Okay.
357 00:29:13.850 ⇒ 00:29:18.520 Uttam Kumaran: Because I would also like… I’m considering Sam plugging Pranavin to help with this.
358 00:29:18.520 ⇒ 00:29:19.110 Samuel Roberts: Okay.
359 00:29:19.290 ⇒ 00:29:21.209 Uttam Kumaran: So I just really need to…
360 00:29:21.340 ⇒ 00:29:26.159 Samuel Roberts: Yeah. To do the quarter… first quarter planning, I just really need to know, like, how many hours are needed.
361 00:29:26.310 ⇒ 00:29:27.240 Uttam Kumaran: Yeah.
362 00:29:27.240 ⇒ 00:29:27.870 Samuel Roberts: Okay.
363 00:29:27.990 ⇒ 00:29:28.830 Samuel Roberts: I get it.
364 00:29:31.880 ⇒ 00:29:36.289 Samuel Roberts: Alright, and then… okay, anything else, ABC, Casey, Mustafa? I think we’re good?
365 00:29:37.640 ⇒ 00:29:39.369 Casie Aviles: Yeah, I think that’s all for me.
366 00:29:41.320 ⇒ 00:29:41.950 Samuel Roberts: Alright.
367 00:29:42.260 ⇒ 00:29:45.910 Samuel Roberts: Lilo… Update, Pranav.
368 00:29:46.320 ⇒ 00:29:58.510 Samuel Roberts: Back tomorrow, I got the MCP good, or working enough. I got the… and then we had the, like, a call with them on Friday, or one of them, I should say.
369 00:29:58.910 ⇒ 00:30:05.739 Samuel Roberts: Today, I was gonna make a little more progress, probably, on the multi-chat.
370 00:30:06.080 ⇒ 00:30:06.850 Uttam Kumaran: Okay.
371 00:30:06.870 ⇒ 00:30:16.220 Samuel Roberts: So that it’s not just one chat per brand, but I think after doing some of the MCP stuff, I think I have a better sense of how to better migrate things now.
372 00:30:16.410 ⇒ 00:30:19.240 Samuel Roberts: Took me a little bit getting up to speed on the repo, but.
373 00:30:19.240 ⇒ 00:30:19.780 Uttam Kumaran: Okay.
374 00:30:19.950 ⇒ 00:30:31.410 Samuel Roberts: I can get that done tomorrow. And then Pranav being back tomorrow, we can keep running. I don’t know what his plans are for the week, if he… is he just tomorrow and then off? So, we’ll figure that out, but…
375 00:30:33.040 ⇒ 00:30:40.609 Samuel Roberts: Yeah, I mean, we’re making good progress there. I don’t know, they said they were kind of on and off, during the holidays a bit, so they’re probably…
376 00:30:40.800 ⇒ 00:30:46.460 Samuel Roberts: gonna be around, they said, so I don’t know how best to manage that when we’re out, but…
377 00:30:47.770 ⇒ 00:30:48.610 Samuel Roberts: Excuse me.
378 00:30:48.890 ⇒ 00:30:53.099 Samuel Roberts: yeah, I think that’s good there.
379 00:30:53.580 ⇒ 00:30:55.880 Samuel Roberts: And then, yeah, game?
380 00:30:58.960 ⇒ 00:31:05.719 Gabriel Lam: Yeah. So… A couple things. The first thing I’ll bring up is the…
381 00:31:06.800 ⇒ 00:31:11.000 Gabriel Lam: the learnings and the push, this is for later today, I have…
382 00:31:11.000 ⇒ 00:31:13.580 Uttam Kumaran: Notion Docs written up for…
383 00:31:14.200 ⇒ 00:31:18.400 Gabriel Lam: the… basically the list of GitHub repos that we have in our org.
384 00:31:19.320 ⇒ 00:31:21.970 Gabriel Lam: I did an audit, I think it’d be good to…
385 00:31:22.730 ⇒ 00:31:24.490 Gabriel Lam: I’ve, like, taken out all those, sort of.
386 00:31:25.290 ⇒ 00:31:34.299 Gabriel Lam: client repos that we’re not gonna touch. But there’s a list that I would like to go through later today and just be like, hey, is this something we need? Is this something we don’t need? We’ll just…
387 00:31:34.810 ⇒ 00:31:39.070 Gabriel Lam: Put a, you know, Deprecate, tag on it for now.
388 00:31:39.340 ⇒ 00:31:44.569 Gabriel Lam: Another thing that I’m… I’ve got, written up is…
389 00:31:45.040 ⇒ 00:31:52.020 Gabriel Lam: thinking about adoption for Cursor, so… I’ve been testing…
390 00:31:52.190 ⇒ 00:31:56.680 Gabriel Lam: like, the linear MCP and cursor, and it seems to be able to pretty accurately
391 00:31:57.050 ⇒ 00:32:09.599 Gabriel Lam: detect and, like, associate actions with tickets and, like, update ad comments, update names, update parameters. Sam, you mentioned last week that there might be a worry of
392 00:32:09.700 ⇒ 00:32:15.779 Gabriel Lam: like, context overload, or too many, like, tool… tool limits.
393 00:32:16.040 ⇒ 00:32:20.179 Gabriel Lam: But that… I don’t think we would hit that at the moment, I think.
394 00:32:22.970 ⇒ 00:32:28.780 Gabriel Lam: I think we’re fine on that end, but if you have, any thoughts there?
395 00:32:29.350 ⇒ 00:32:35.869 Samuel Roberts: My only thought is there’s, like, what is it, like, 20-something tool calls that I can get from Linear?
396 00:32:36.230 ⇒ 00:32:50.590 Samuel Roberts: as long as we need them, like, all, that’s good. If there’s ones we don’t need actively, like, turning off whatever ones that are, like, never gonna be used would be kind of a smart thing, because who knows what other MCPs we will start, you know, utilizing.
397 00:32:50.590 ⇒ 00:32:58.109 Uttam Kumaran: I mean, the alternative also is, like, we develop our own MCP, or some way for people locally to hit something else.
398 00:32:58.110 ⇒ 00:33:02.150 Samuel Roberts: Yeah, well, that’s what I was talking about, the CLI, potentially. It doesn’t look like.
399 00:33:02.150 ⇒ 00:33:03.490 Uttam Kumaran: Yeah, yeah, yeah, yeah, yeah.
400 00:33:03.490 ⇒ 00:33:06.329 Samuel Roberts: But even if we do something like that, and have just, like.
401 00:33:06.540 ⇒ 00:33:10.130 Samuel Roberts: you know, cursor generate some scripts that are reusable, so it’s not…
402 00:33:10.130 ⇒ 00:33:10.880 Uttam Kumaran: Yeah.
403 00:33:10.880 ⇒ 00:33:18.600 Samuel Roberts: Yeah, exactly, and if it’s using the same… like, if it has the docs for the CLI, it can generate those calls pretty… probably pretty well.
404 00:33:18.600 ⇒ 00:33:31.929 Uttam Kumaran: Yes, so that’s why I don’t… I’m not married to the MCP, I’m… I’m… what I am feeling pretty confident is the fact that it needs to plan, the user needs to, like, see what it’s gonna do, and then…
405 00:33:32.230 ⇒ 00:33:39.799 Uttam Kumaran: it goes and takes action. If the CLI is just… because no… our team, like, nobody, I don’t want them to care much about
406 00:33:40.060 ⇒ 00:33:51.129 Uttam Kumaran: the MCP, and I don’t want to deal with the MCP limits, because it is just a glorified, like, endpoint call, so if using the CLI is just much more robust, then let’s just do that, I don’t care, you know?
407 00:33:51.130 ⇒ 00:33:54.660 Samuel Roberts: Okay, yeah, I mean, it’s worth testing out,
408 00:33:55.180 ⇒ 00:34:03.010 Samuel Roberts: Again, this wasn’t an official one, so I haven’t played with it yet, I don’t know how good it is, but it looked like it was specifically built for, like, coding agents more.
409 00:34:03.120 ⇒ 00:34:14.180 Samuel Roberts: So, like, I think this is the use case they had in mind for it, because coding agents are good at, like, making those CLI calls and reading the docs for how to do it all.
410 00:34:14.360 ⇒ 00:34:23.820 Samuel Roberts: And the nice thing about, I think, maybe doing it this way over the MCP is, like, it only goes in the context… like, we can point it to a file so that it knows how to use the CLI when we need it.
411 00:34:24.580 ⇒ 00:34:34.159 Samuel Roberts: Rather than every time you make a call, all the MCP tools get fed to the agent, or the LLM, for whatever you have enabled in Cursor.
412 00:34:34.800 ⇒ 00:34:43.740 Samuel Roberts: Which, again, might not be an issue, we might not be getting to context issues like that, but it also will depend, like, what model you’re using and how good it is at making the tool calls and stuff.
413 00:34:45.260 ⇒ 00:34:46.760 Samuel Roberts: So, I’m…
414 00:34:46.960 ⇒ 00:34:53.730 Samuel Roberts: I think it’s worth testing out, at the very least. Like, if the MCP is working, that’s, like, that’s great, I don’t want to, like, really mess with that too much, but…
415 00:34:53.739 ⇒ 00:34:54.429 Uttam Kumaran: Okay.
416 00:34:55.060 ⇒ 00:35:04.850 Samuel Roberts: it’s worth knowing if this CLI is good, and basically, if there’s gonna be, like, a lot of the same type of behavior, we could… we could document that pretty well in an AgentsMD.
417 00:35:05.330 ⇒ 00:35:10.010 Samuel Roberts: Which is gonna basically do the same thing as the MCP, is my thought, without…
418 00:35:10.240 ⇒ 00:35:12.230 Samuel Roberts: Being there every time, but…
419 00:35:12.230 ⇒ 00:35:13.630 Uttam Kumaran: I agree.
420 00:35:14.240 ⇒ 00:35:14.860 Samuel Roberts: Nope.
421 00:35:15.560 ⇒ 00:35:19.040 Gabriel Lam: So… if I… yeah.
422 00:35:19.390 ⇒ 00:35:22.009 Gabriel Lam: I… I have, like, a…
423 00:35:24.430 ⇒ 00:35:33.419 Gabriel Lam: a Notion doc written up, mostly because my cursor is down at the moment, and my GitHub is down. I, like, ran out of space on this laptop, so… I’m gonna…
424 00:35:33.420 ⇒ 00:35:35.600 Uttam Kumaran: Oh, figure that out.
425 00:35:35.600 ⇒ 00:35:42.779 Gabriel Lam: Okay. But… Yeah, so if you guys, maybe have time before the meeting this afternoon, we can…
426 00:35:44.280 ⇒ 00:35:45.849 Gabriel Lam: Like, have a quick read through it.
427 00:35:46.250 ⇒ 00:35:52.319 Gabriel Lam: Before, and then we can discuss the actual exercise of… of everything there. I also have the doc for…
428 00:35:52.700 ⇒ 00:35:56.370 Gabriel Lam: AI slides, and I think the two key…
429 00:35:56.540 ⇒ 00:36:01.349 Gabriel Lam: ones to look at are Gamma and Notebook LM.
430 00:36:01.350 ⇒ 00:36:01.990 Samuel Roberts: Okay.
431 00:36:03.960 ⇒ 00:36:04.670 Samuel Roberts: Cool.
432 00:36:05.190 ⇒ 00:36:12.359 Gabriel Lam: So, it seems like… Ashwini is already, or Pranav is already using Gamma.
433 00:36:12.360 ⇒ 00:36:16.470 Uttam Kumaran: A bunch of people randomly are using it, because I just bought it one day, and I gave people, so…
434 00:36:16.470 ⇒ 00:36:17.020 Samuel Roberts: Right.
435 00:36:17.020 ⇒ 00:36:20.289 Uttam Kumaran: I can give you my login if you want to go in there and poke around.
436 00:36:20.760 ⇒ 00:36:22.539 Gabriel Lam: Yeah, yeah. So that’d be…
437 00:36:22.540 ⇒ 00:36:23.130 Mustafa Raja: Yeah, I…
438 00:36:23.130 ⇒ 00:36:23.549 Gabriel Lam: to look at.
439 00:36:23.550 ⇒ 00:36:28.659 Mustafa Raja: I did use it in my university also when I… when I was still studying.
440 00:36:29.900 ⇒ 00:36:31.610 Mustafa Raja: For some presentations.
441 00:36:32.410 ⇒ 00:36:33.450 Gabriel Lam: Yeah, okay.
442 00:36:34.520 ⇒ 00:36:35.239 Samuel Roberts: And this is…
443 00:36:35.240 ⇒ 00:36:36.989 Uttam Kumaran: Also, I… yeah, yeah, go ahead.
444 00:36:36.990 ⇒ 00:36:41.820 Samuel Roberts: I was gonna ask about the UI versus, like, how we can utilize it in, like, Cursor, potentially.
445 00:36:42.890 ⇒ 00:36:44.089 Samuel Roberts: Because I know they had some.
446 00:36:44.090 ⇒ 00:36:51.110 Uttam Kumaran: Yeah, most of the ex… most of the exploration is gonna… is gonna be based on two things. It’s gonna be, like, can we,
447 00:36:51.320 ⇒ 00:37:00.840 Uttam Kumaran: set up Gamma in a way where if people use the UI, they’re, like, using all of our frameworks, and second, and honestly more important to me, is do they have an API?
448 00:37:00.860 ⇒ 00:37:03.509 Samuel Roberts: Yeah. And can you programmatically create stuff?
449 00:37:04.030 ⇒ 00:37:08.130 Uttam Kumaran: like… Yeah, so I feel like Gamma…
450 00:37:08.640 ⇒ 00:37:13.160 Uttam Kumaran: maybe in a better spot than Notebook LM for having, like, programmatic access.
451 00:37:13.290 ⇒ 00:37:20.249 Uttam Kumaran: For the most part, what I’m seeing on Twitter is that people are using Notebook LM, and that the outputs are good, but you have to do everything via the UI.
452 00:37:21.610 ⇒ 00:37:22.650 Gabriel Lam: Yeah. You know, so…
453 00:37:22.750 ⇒ 00:37:23.789 Samuel Roberts: Yeah, when we looked at it…
454 00:37:23.790 ⇒ 00:37:25.290 Gabriel Lam: an official API.
455 00:37:25.290 ⇒ 00:37:27.690 Samuel Roberts: But Gamma, I believe, has.
456 00:37:27.880 ⇒ 00:37:29.400 Gabriel Lam: has an API offering.
457 00:37:29.850 ⇒ 00:37:36.330 Samuel Roberts: Okay, yeah, I remember looking a little bit for Interlude, and it seemed like there was something, but I wasn’t sure how robust it was at that point.
458 00:37:38.800 ⇒ 00:37:39.540 Samuel Roberts: Okay.
459 00:37:42.380 ⇒ 00:37:43.080 Samuel Roberts: Cool.
460 00:37:43.510 ⇒ 00:37:44.260 Samuel Roberts: Nope.
461 00:37:44.410 ⇒ 00:37:45.240 Samuel Roberts: Excuse me.
462 00:37:45.680 ⇒ 00:37:47.589 Samuel Roberts: The other thing, like, what are the…
463 00:37:47.700 ⇒ 00:37:54.679 Samuel Roberts: And this is just because we looked at a little bit of this for Interlude when we were talking, like, deck generation. There’s a tool
464 00:37:55.140 ⇒ 00:38:11.440 Samuel Roberts: that, like, outputs mark… or can take in Markdown and make presentations. It’s, like, for, you know, like, developers making presentations kind of thing. It’s not as, like, robust and pretty, but I’m wondering if that’s a good, like, lo-fi to pass into these tools, or, like, do these tools want that kind of thing, or do they just want, like.
465 00:38:11.560 ⇒ 00:38:12.610 Samuel Roberts: a prompt.
466 00:38:13.600 ⇒ 00:38:15.549 Gabriel Lam: I see two… I see two…
467 00:38:15.930 ⇒ 00:38:31.779 Gabriel Lam: ways that we use slide generation. The first is, like, for technical and, like, engineers, because we’ll work in, like, documentation and markdowns, like, that gets passed into a lo-fi version that, you know, design or marketing can then be like, okay, I have all the copy, I have all the content of every slide.
468 00:38:31.880 ⇒ 00:38:42.050 Gabriel Lam: The other version would be, like, hey, we have templates. Like, existing templates, we have good examples of what a deck for, you know,
469 00:38:42.210 ⇒ 00:38:54.200 Gabriel Lam: like, a data analysis might look like. So we can put that in, and we can either compare or review it to what we have now, or use those as guidelines. So I think there’s, like, two main things. I don’t really see it as, like, prompting
470 00:38:54.320 ⇒ 00:38:55.989 Gabriel Lam: And seeing, like, oh, you know.
471 00:38:56.790 ⇒ 00:39:03.749 Gabriel Lam: like, make me a slide of this, because I think we have enough Existing documentation that works.
472 00:39:04.050 ⇒ 00:39:12.559 Gabriel Lam: and… It’s, like, the two personas that end up using it are, they’re, like, highly technical folks.
473 00:39:12.960 ⇒ 00:39:18.079 Gabriel Lam: Who want to go from the markdown to… The slides versus, like, the…
474 00:39:18.470 ⇒ 00:39:22.310 Gabriel Lam: Just, you know, you’re… you’re in a slide… a slide deck.
475 00:39:24.250 ⇒ 00:39:24.980 Samuel Roberts: Okay.
476 00:39:25.350 ⇒ 00:39:37.110 Samuel Roberts: Yeah, I was just trying to find that tool that might generate, like, lo-fi ones, I wasn’t sure. It’s called SSLI.DEV SlideDev. This is just something we found when we were trying to figure out,
477 00:39:37.610 ⇒ 00:39:40.180 Samuel Roberts: What we might do for…
478 00:39:40.440 ⇒ 00:39:44.860 Samuel Roberts: Interlude, which was actually us generating copy for decks, and maybe a lo-fi first pass.
479 00:39:44.970 ⇒ 00:39:51.839 Samuel Roberts: The only reason I liked this was just because it could be… Output from another model.
480 00:39:52.090 ⇒ 00:39:59.939 Samuel Roberts: Like, because it’s just Markdown. And so, I don’t know how customizable this gets, if it’s something that we could, like, get a few basic layouts defined, and then…
481 00:40:00.670 ⇒ 00:40:09.139 Samuel Roberts: that gives us a, like, unbranded, maybe, version that then gets passed through something? I don’t know. There’s a little more there that I’m not sure of.
482 00:40:09.140 ⇒ 00:40:23.829 Uttam Kumaran: Yeah, we do have structure, and we do already have, like, an internal dock with all the different layouts that are commonly used. And as you guys know, you guys are engineers, and you guys are working on decks. I don’t want you guys focused on format, I want focus on content.
483 00:40:26.200 ⇒ 00:40:39.789 Uttam Kumaran: And… yeah, like, that’s the main objective, is, like, I want our people to spend more time using AI or whatever to, like, think about what the actual… is gonna go into the slides, versus, like, finagling with…
484 00:40:39.790 ⇒ 00:40:40.320 Samuel Roberts: Yeah.
485 00:40:40.320 ⇒ 00:40:41.600 Uttam Kumaran: Formats and stuff.
486 00:40:41.960 ⇒ 00:40:51.429 Uttam Kumaran: that’s, like, my… that’s my priority. Ideally, I want to get to the point where the whole thing can be done basically programmatically, and our engineers turn into more of, like.
487 00:40:51.650 ⇒ 00:40:52.540 Uttam Kumaran: last…
488 00:40:52.660 ⇒ 00:40:58.599 Uttam Kumaran: Mile tweaks and things, because all the context will be in meeting notes and transcripts and stuff like that.
489 00:40:58.910 ⇒ 00:41:10.169 Uttam Kumaran: You know, like, think about a world where we take all of the transcripts from the week for a client, and then on Thursday morning, I use this system to generate the deck, the overview deck.
490 00:41:10.280 ⇒ 00:41:13.340 Uttam Kumaran: Right, so, like, that’s, like, a good North Star to hold in your brain.
491 00:41:14.570 ⇒ 00:41:16.900 Samuel Roberts: Yeah, Gabe, I’m sending… this is the link to the.
492 00:41:16.900 ⇒ 00:41:17.560 Gabriel Lam: Awesome.
493 00:41:17.840 ⇒ 00:41:20.830 Samuel Roberts: There’s probably other tools out there that do something similar.
494 00:41:21.010 ⇒ 00:41:27.019 Samuel Roberts: But again, the benefit here was just that any model that’s a language model can do this.
495 00:41:27.370 ⇒ 00:41:28.980 Samuel Roberts: And then…
496 00:41:30.470 ⇒ 00:41:37.399 Samuel Roberts: you know, we don’t have to rely on, like, a specialized thing, like Notebook LM or Gamma or something, at least for, like, a first, you know.
497 00:41:38.000 ⇒ 00:41:48.279 Samuel Roberts: But as I’m digging through this just now, it looks like there are ways to customize it and everything, so even this might be something that’s worth investing a little bit of time to see, like, how close we can match this to, like, the Google Slides that we have, maybe?
498 00:41:48.400 ⇒ 00:41:49.980 Samuel Roberts: And see what it outputs, but…
499 00:41:51.290 ⇒ 00:41:58.879 Samuel Roberts: Okay. Yeah, just a thought, again, we spent a little bit of time spiking on this for Interlude before that stops.
500 00:41:59.520 ⇒ 00:42:00.100 Gabriel Lam: department.
501 00:42:01.960 ⇒ 00:42:05.589 Uttam Kumaran: One other piece, that maybe…
502 00:42:06.220 ⇒ 00:42:12.650 Uttam Kumaran: Okay, we could just… you could just create the Notion doc for me to write here, is I want to also talk about,
503 00:42:14.450 ⇒ 00:42:18.730 Uttam Kumaran: what I would call… like,
504 00:42:18.920 ⇒ 00:42:31.620 Uttam Kumaran: browser automation solutions, and this is going to mainly affect people on the operations and the sales side, but it could also affect delivery, which is, I want us to explore tools like Comet.
505 00:42:31.820 ⇒ 00:42:41.920 Uttam Kumaran: Atlas, There’s a tool called Manus for basically, like,
506 00:42:42.350 ⇒ 00:42:48.479 Uttam Kumaran: browser automation, but I don’t want this to be confused with Playwright or,
507 00:42:49.480 ⇒ 00:42:59.520 Uttam Kumaran: browser-based. Instead, this is, like, for example, some of our stuff on the operations and the sales side is literally going and, like, filling out a form, or, like.
508 00:42:59.780 ⇒ 00:43:16.110 Uttam Kumaran: submitting something on… like, this is a good example. Ryan submits bids on Upwork. I know that what he’s doing is he probably takes the scope of the proposal, throws it into a custom GPT, and then is copying and pasting it over into these forms.
509 00:43:16.220 ⇒ 00:43:32.950 Uttam Kumaran: Instead, I’m sure he can actually probably just do all of that within one browser interface. Additionally, think about an example, where we don’t understand, like, what the form is beforehand. And so, there’s a lot of tasks that we do that are so UI-based,
510 00:43:33.390 ⇒ 00:43:47.979 Uttam Kumaran: that are just, like, ripe for automation, but the problem is we may not know what the format is ahead of time, and so I want to just have a doc. You know, we don’t have to work on this today, but I’m just gonna brain dump a bunch of stuff around general
511 00:43:48.040 ⇒ 00:43:55.329 Uttam Kumaran: like, browser automation. So for our operations team, for our sales team, if they’re doing stuff on the internet.
512 00:43:55.330 ⇒ 00:44:13.559 Uttam Kumaran: and they’re finding themselves, like, filling out a form constantly, or, like, updating a cell in a spreadsheet, or, like, any of that type of stuff, I want them to potentially consider using an AI browser solution, or… for example, Claude just released a browser extension, or they beefed up their existing one.
513 00:44:13.580 ⇒ 00:44:19.320 Uttam Kumaran: That you basically can say, like, fill out this form in this way, and it’ll go Poke on things.
514 00:44:19.540 ⇒ 00:44:25.980 Uttam Kumaran: Cursor also has browser automate, so I just want us to think about something around this… this world of…
515 00:44:26.300 ⇒ 00:44:32.520 Uttam Kumaran: Things. There’s, for example, like, Casey, I think a lot of what you’re doing for insomnia could potentially get
516 00:44:32.940 ⇒ 00:44:38.279 Uttam Kumaran: done by this, right? Like, opening up Uber Eats, sitting down, moving it, right? So, that type of stuff.
517 00:44:42.170 ⇒ 00:44:45.709 Uttam Kumaran: So if you just create the doc, I’ll throw stuff in there. This is a long story short, but…
518 00:44:45.710 ⇒ 00:44:46.300 Gabriel Lam: Yeah.
519 00:44:46.710 ⇒ 00:44:48.790 Uttam Kumaran: Wanted to get you guys to think about that.
520 00:44:50.730 ⇒ 00:44:51.320 Gabriel Lam: Okay.
521 00:44:56.230 ⇒ 00:45:01.100 Uttam Kumaran: Cool. Anything else? So, Sam, you’re gonna work on that for Lilo,
522 00:45:02.180 ⇒ 00:45:09.120 Uttam Kumaran: maybe I could also poke at, potentially, how to get Replit hooked up with our GitHub.
523 00:45:09.720 ⇒ 00:45:10.130 Samuel Roberts: Yeah.
524 00:45:10.130 ⇒ 00:45:11.369 Uttam Kumaran: Do you want me to work on that?
525 00:45:11.580 ⇒ 00:45:17.489 Samuel Roberts: Yeah, I don’t know exactly… yeah, I don’t know where they’re, like, if they just pull from it, or if they can actually, like, be in the repo.
526 00:45:17.900 ⇒ 00:45:19.389 Samuel Roberts: That would be good to know, at least.
527 00:45:20.220 ⇒ 00:45:25.089 Samuel Roberts: As a first pass of, like, yeah, it works, yeah, no, it doesn’t, or it only pulls one way, like the, you know…
528 00:45:28.280 ⇒ 00:45:30.490 Uttam Kumaran: Anything else on Lilo you want me to take?
529 00:45:30.900 ⇒ 00:45:33.580 Samuel Roberts: Trying to think.
530 00:45:36.090 ⇒ 00:45:38.840 Samuel Roberts: Not that I can think of offhand right now…
531 00:45:42.400 ⇒ 00:45:44.449 Samuel Roberts: I think that’s a good one,
532 00:45:49.710 ⇒ 00:46:00.609 Samuel Roberts: Yeah, just run with that for now, because I’m not… I’m… honestly, I’m not sure how long these, like, multi-chat things are gonna take. I feel like some of it’s already tough to do it, so, like, I’ll… I’ll give you… we’ll get a sense, we can maybe reconvene later, but, like, just on Slack.
533 00:46:01.230 ⇒ 00:46:02.170 Samuel Roberts: See where we’re at.
534 00:46:04.970 ⇒ 00:46:13.529 Samuel Roberts: Because there’s a couple other things, like, right now, that Klaviyo MCP is the only one that’s actually, like, set up, but all the other ones are, like, look like they are, so I might just do a few things to make it clearer.
535 00:46:13.650 ⇒ 00:46:17.919 Samuel Roberts: And the multi… yeah, I think we’re good with that. The replic will be helpful.
536 00:46:18.250 ⇒ 00:46:18.840 Samuel Roberts: That way.
537 00:46:18.840 ⇒ 00:46:19.450 Uttam Kumaran: Okay.
538 00:46:20.890 ⇒ 00:46:21.960 Uttam Kumaran: Okay, great.
539 00:46:23.260 ⇒ 00:46:25.770 Uttam Kumaran: Okay. Cool, guys, I think that’s all I had.
540 00:46:25.770 ⇒ 00:46:26.560 Samuel Roberts: Alright.
541 00:46:27.850 ⇒ 00:46:30.069 Samuel Roberts: Thank you all, okay.
542 00:46:30.550 ⇒ 00:46:33.189 Samuel Roberts: Talk to you all at some point later, Slack or whatever.
543 00:46:33.620 ⇒ 00:46:34.120 Uttam Kumaran: Okay.
544 00:46:34.120 ⇒ 00:46:35.120 Samuel Roberts: Alright, alright.
545 00:46:35.120 ⇒ 00:46:36.480 Uttam Kumaran: Thank you. Bye.