Meeting Title: Miguel <> Uttam <> Amber Date: 2025-05-07 Meeting participants: Miguel De Veyra, Uttam Kumaran, Amber Lin
WEBVTT
1 00:00:53.550 ⇒ 00:00:55.110 Amber Lin: Hello!
2 00:00:55.595 ⇒ 00:00:56.080 Miguel de Veyra: Everyone.
3 00:00:56.080 ⇒ 00:00:56.810 Uttam Kumaran: Hey?
4 00:00:59.630 ⇒ 00:01:04.980 Uttam Kumaran: Okay, cool. I sent a I sent like a little agenda. If you guys wanna take a look. Yeah.
5 00:01:05.180 ⇒ 00:01:05.750 Amber Lin: Yeah.
6 00:01:06.410 ⇒ 00:01:19.079 Uttam Kumaran: So I definitely don’t want to do the majority of talking on this one, but I think this is a good framing for this conversation. I think each of us has, like different perspectives on how
7 00:01:19.680 ⇒ 00:01:44.919 Uttam Kumaran: this project is going. And I think this is a perfect time for us to just basically make sure each of us is getting what we need. I think my, what I need is going to be different than what Amber needs and different what Miguel needs and what the. And ultimately it’s whatever the company needs for this project to go. Well, I think I want to talk. I think the rag is a good is a good starting point project to like use it as example. But I do think that
8 00:01:44.950 ⇒ 00:01:52.539 Uttam Kumaran: I’ve I’ve reiterated how important this team is for us to basically not only be able to grow revenue, but also hit
9 00:01:52.640 ⇒ 00:01:59.989 Uttam Kumaran: like our margin targets and like, keep a very lean, mean team, but also grow
10 00:02:00.140 ⇒ 00:02:21.750 Uttam Kumaran: way faster than if we. You know, if we had like, basically try to grow at the pace that a 50 or 100 person team would grow right. And AI is is the way we do that. I think we have a lot of ideas. I have more ideas than we will work on in this lifetime. But I want to make sure that we actually get those done. I think we’ve
11 00:02:22.220 ⇒ 00:02:36.940 Uttam Kumaran: done a good job at kicking off some of the some of the things, but I don’t think we’ve done a good job on executing things end to end. I think there’s a couple of limitations in the way that I think we can discuss
12 00:02:37.110 ⇒ 00:02:46.641 Uttam Kumaran: today. I think part of it is on communication, I think part of it on is on what each of us need to take on for this to succeed.
13 00:02:47.300 ⇒ 00:03:03.700 Uttam Kumaran: I also think part of it is on just setting expectations really clear on when things can come out, and why they’re not coming out. You know, for me. I see it from my angle, and I know what the team is capable of. And I know I have enough technical context to know what is possible.
14 00:03:03.920 ⇒ 00:03:09.520 Uttam Kumaran: like, what is like, literally, physically possible, and what is blocked by
15 00:03:09.680 ⇒ 00:03:29.719 Uttam Kumaran: like the technology doesn’t exist. I’m seeing a lot in the market. I’m seeing a lot from our partners, from other sort of businesses. So for us to be leading edge. Like I’m my job is to push that forward. I think you guys both have a different job on the team. And I wanna make sure it’s just crystal clear
16 00:03:29.970 ⇒ 00:03:44.681 Uttam Kumaran: what those are. So that’s my perspective. I think you guys read the top of the notion. But I would love to hear. You know I have some just like introductory questions. There would love to hear your feedback on that
17 00:03:45.510 ⇒ 00:03:48.449 Uttam Kumaran: or, you know, basically, like, wherever we
18 00:03:48.630 ⇒ 00:03:51.860 Uttam Kumaran: we want to start would love to hear from each of you on like
19 00:03:52.200 ⇒ 00:03:55.379 Uttam Kumaran: how you think the project is going, and just opening thoughts.
20 00:03:57.610 ⇒ 00:04:04.430 Amber Lin: Yeah, Miguel, you wanna start. Let’s both just share our current perspective, and we’ll see where we align. We where we.
21 00:04:04.430 ⇒ 00:04:05.680 Miguel de Veyra: Hi.
22 00:04:05.950 ⇒ 00:04:20.160 Miguel de Veyra: yeah, I mean for the slack agents, for the client agents, the client hubs. I think we’ve pretty much done. We’ve pretty much initial initialized everything like there are some stuff like, for example, the slack thing that we do need to drag.
23 00:04:20.350 ⇒ 00:04:40.249 Miguel de Veyra: So. But you know we’ve already found the way which we will discuss here. We’ll just discuss it with 2 of them. And then but yeah, I think, generally speaking, this went smoothly, as I hoped it would be. It’s just that the drag thing. Of course, there’s always caveats on, you know. Surprises. But yeah, that should be.
24 00:04:40.250 ⇒ 00:04:40.860 Amber Lin: Hmm
25 00:04:42.348 ⇒ 00:05:00.299 Amber Lin: for me. I’m sorry. I’m sort of thinking about it as not just this current roadblock, because I know I have trust on you guys that it’s gonna get figured out. I’m just more thinking of on the on the bigger horizon, right? Because figuring each of these pieces out
26 00:05:00.770 ⇒ 00:05:08.499 Amber Lin: doesn’t help the business until like the end product is out. If Kyle, or if the
27 00:05:08.940 ⇒ 00:05:24.949 Amber Lin: it doesn’t have something to use, like all of our hard works, kind of like just goes to waste if they never use it so. And I’m still thinking about like all the agents when we wanna get them
28 00:05:25.410 ⇒ 00:05:45.889 Amber Lin: usable cause right now. I think it’s really good that we almost have. Every single client has an agent, even if it’s not great, because it at least can answer some basic non engineering questions. So I think that’s great. But to get it to the point where it can fulfill the data needs
29 00:05:46.605 ⇒ 00:05:55.330 Amber Lin: there’s still some stuff happening there, and we’re spending some time on rag with for slack
30 00:05:55.820 ⇒ 00:06:01.580 Amber Lin: but sorry brag is just gonna be for everything. So I think it would help to
31 00:06:02.180 ⇒ 00:06:08.190 Amber Lin: understand what what is timeframe of this becomes usable for our clients.
32 00:06:08.440 ⇒ 00:06:12.650 Amber Lin: So that’s that’s kind of where
33 00:06:12.850 ⇒ 00:06:15.960 Amber Lin: I don’t think we have that much clarity on
34 00:06:16.717 ⇒ 00:06:22.790 Amber Lin: because it is sort of a quick scope shift. Since we since we talked to the data team.
35 00:06:23.020 ⇒ 00:06:32.789 Amber Lin: and I just don’t think we have like. I haven’t set a clear roadmap for that. I haven’t like. We haven’t had clear deadlines
36 00:06:33.370 ⇒ 00:06:39.030 Amber Lin: for that as well. So that is my perception. The current status.
37 00:06:39.850 ⇒ 00:06:47.110 Uttam Kumaran: Yeah, I think maybe I’ll I’ll also just add on that. I think even between both of y’all, it’s clear that there is
38 00:06:47.320 ⇒ 00:07:08.229 Uttam Kumaran: a misalignment on how the project is going. I mean for me. The onus of this of this meeting is that one I’m even at a perspective higher than Amber’s, where I need to see these initiatives benefit our company health. Either we’re making more money or we’re saving money, right? So I’m not happy until either of those 2 things happen.
39 00:07:08.730 ⇒ 00:07:13.550 Uttam Kumaran: This is a bat. See the
40 00:07:14.800 ⇒ 00:07:37.709 Uttam Kumaran: that these are gonna save us money meaning none of the work here is actually billable, meaning. It’s not revenue generating right? And so I have to go defend this to Robert and to our team on why this is important and why this is a worthwhile potential distraction for us to do right. None of this stuff is is immediately revenue generating. And so
41 00:07:37.800 ⇒ 00:07:54.549 Uttam Kumaran: we’ve worked on this stuff way too long, and these are all capabilities that I know we could have done. You know, like, months ago, right? And I feel like we’ve always been a few weeks away from all these things happening.
42 00:07:54.560 ⇒ 00:07:57.320 Uttam Kumaran: and it’s still not there. And so
43 00:07:57.330 ⇒ 00:08:13.150 Uttam Kumaran: when I go to Amber, and part of her responsibility is to tell me when things are going to happen, and she can’t do that. Then then I go one layer deeper. Right so then I joined the meeting yesterday, and yes, although I think finally, we these teams are talking
44 00:08:13.150 ⇒ 00:08:28.929 Uttam Kumaran: one. It’s way too late. Right? We are now 6 weeks or 8 weeks since the the time where I was like. This is not moving fast enough, and that is way too long. That is 2 months. Companies only been in business 18 months.
45 00:08:29.070 ⇒ 00:08:48.579 Uttam Kumaran: Right? So do the math. That’s like way too long. Second, is technology. This stuff is easy. We’re past. We’ve been able to do this months ago. And so I’m just like surprised to see that it’s taking so long, and I can’t get a clear answer as to what is taking so long right, because I know how fast
46 00:08:48.800 ⇒ 00:09:07.790 Uttam Kumaran: the AI team can move, and so I’m struggling to understand why these slack agents weren’t done weeks ago. I’m struggling to understand why, although you’re talking to a Weish, and the team is working with data, there isn’t any plan for more than 30 days worth of slack messages
47 00:09:07.790 ⇒ 00:09:26.169 Uttam Kumaran: across. Apart from putting it into context, which we all know is not all 3 of us on this call know that is not an efficient, elegant solution that can scale. And so I’m at the point where, as the business owner, I have to make a decision onto whether this is going to succeed or not.
48 00:09:26.250 ⇒ 00:09:45.429 Uttam Kumaran: I have not had any strong indication to me that spending an additional 8 weeks on this, we’re going to get to where we want to go. So I’m at a fork in the road right? And I want to give you my perspective. And you guys know, like, I’ll tell you how I have to think about this is that costs are getting more and more important for us. Right? I think we ran the business in a lot
49 00:09:45.430 ⇒ 00:09:58.870 Uttam Kumaran: looser way, and we’re starting to think about cost more, not because we’re losing money, because, but we actually want to money. And I want. And the second thing is, I want everybody to be focused on the core company goals.
50 00:09:58.880 ⇒ 00:10:15.540 Uttam Kumaran: This is this AI piece is something that I took a bet on, and I thought would be very, very impactful you know, for a company to to spend time on, and so it would save us money. We could recruit better talent. We could be on the leading edge. But so far we haven’t seen really any
51 00:10:15.540 ⇒ 00:10:34.090 Uttam Kumaran: meaningful impact. Internally. Of course, we’ve been doing work for clients, but none of the sales hub stuff has been working. I still don’t know if anyone apart from me is using the zoom summaries. Nobody is using any of the agents or simple slack workflows. So
52 00:10:34.250 ⇒ 00:10:56.459 Uttam Kumaran: my conclusion. And again, you could tell me whether this is fair is that we haven’t succeeded at all. It’s just we failed at getting any of these adopted and failed to see any of the benefits. In fact, probably the only thing that’s worked is like, I’ve gotten everybody a chat Gpt license, and I’ve yelled at everyone to use AI. That’s probably the only thing that’s worked and I need to invest. I could.
53 00:10:56.820 ⇒ 00:10:57.870 Uttam Kumaran: So this is.
54 00:10:58.210 ⇒ 00:10:58.770 Miguel de Veyra: My!
55 00:10:58.770 ⇒ 00:10:59.369 Miguel de Veyra: Is this.
56 00:10:59.370 ⇒ 00:11:00.320 Uttam Kumaran: Is, that.
57 00:11:01.660 ⇒ 00:11:06.479 Miguel de Veyra: You’re. I’m not sure if it’s just mine, but you’re chopping off like very robotic.
58 00:11:06.760 ⇒ 00:11:08.619 Amber Lin: Your audio is a little choppy.
59 00:11:08.620 ⇒ 00:11:09.220 Uttam Kumaran: How about now?
60 00:11:09.220 ⇒ 00:11:14.700 Amber Lin: Take some of it, but sometimes it’s, I think right now should be better. Can you try again?
61 00:11:15.950 ⇒ 00:11:17.660 Uttam Kumaran: Okay, you have it now.
62 00:11:17.660 ⇒ 00:11:18.330 Amber Lin: Be good.
63 00:11:18.600 ⇒ 00:11:20.299 Miguel de Veyra: Yeah, that’s a bit better. Now.
64 00:11:22.110 ⇒ 00:11:23.510 Amber Lin: I got most of it, though.
65 00:11:23.840 ⇒ 00:11:25.140 Miguel de Veyra: Yeah, yeah, we got most of it.
66 00:11:27.920 ⇒ 00:11:28.980 Miguel de Veyra: Now you’re gone.
67 00:11:29.310 ⇒ 00:11:40.810 Uttam Kumaran: Yeah, I’m not on conferences. This is going anywhere. So I need some help here. Like I, we need to know. I need a really clear understanding that this is gonna work. Basically by the end of this week.
68 00:11:41.377 ⇒ 00:11:45.130 Uttam Kumaran: Otherwise, like, I can’t keep funding internal engineering work
69 00:11:45.820 ⇒ 00:11:50.140 Uttam Kumaran: on this stuff, because there’s no path for it to work like.
70 00:11:50.750 ⇒ 00:11:56.650 Uttam Kumaran: So like that’s that’s like my perspective on this. It’s a little bit harsh, but like
71 00:11:57.160 ⇒ 00:12:04.170 Uttam Kumaran: I don’t know I’m I’ve tried many other things to try to make this work, and it’s not happening, you know.
72 00:12:07.460 ⇒ 00:12:08.510 Uttam Kumaran: so
73 00:12:09.180 ⇒ 00:12:33.199 Uttam Kumaran: like I would love to hear. I mean, I think, Amber, you kind of hinted at a little bit of it, I think, Miguel, you know, although I heard from you that things you feel like things are going well, I think there’s clearly some misalignment in what our joint definition of well means. So like, how can we get aligned on like? How can we 1st agree on like what the state of the project is? And then do you, did you feel like any of the stuff that I mentioned was actually
74 00:12:33.390 ⇒ 00:12:34.370 Uttam Kumaran: like fair.
75 00:12:36.577 ⇒ 00:12:37.920 Miguel de Veyra: Yeah, I mean.
76 00:12:39.230 ⇒ 00:12:43.909 Miguel de Veyra: I would have done the same in your position. Probably I would have got the team a lot earlier.
77 00:12:45.040 ⇒ 00:12:50.490 Miguel de Veyra: right like looking at it from your perspective. Yeah, it makes sense. So I think it’s fair.
78 00:12:52.620 ⇒ 00:13:00.060 Uttam Kumaran: But my question is not. Do you see it as fair? Because I convinced you like? My point is that how are we so misaligned like.
79 00:13:00.180 ⇒ 00:13:05.560 Uttam Kumaran: What you just said a few minutes ago is totally different than what I said, and is
80 00:13:06.020 ⇒ 00:13:10.530 Uttam Kumaran: pretty different than what Amber said. Right? So we all had a different understanding of what
81 00:13:11.640 ⇒ 00:13:14.409 Uttam Kumaran: like, where the project status really is.
82 00:13:17.170 ⇒ 00:13:19.399 Uttam Kumaran: That’s what I want us to sort of dissect here.
83 00:13:27.830 ⇒ 00:13:28.680 Miguel de Veyra: Let me see.
84 00:13:39.250 ⇒ 00:13:47.729 Uttam Kumaran: Right, because we can go through the next exercise of who owns what blah blah. But, like fundamentally, our agents aren’t getting used.
85 00:13:48.140 ⇒ 00:14:00.790 Uttam Kumaran: I would say, a majority of your your time and Casey time and a good portion of Amber’s time. And now other people on the data team’s time is going towards is going towards this.
86 00:14:01.240 ⇒ 00:14:22.879 Uttam Kumaran: not working, so the company is not seeing a return on the investment right now, and like I can’t. I don’t know what a path forward is. I’ve tried multiple things right. I’ve tried to move amber on it. ABC is now a lot less work. So I’ve scoped that down. We have linear for tracking. We have project plans. We have the vision.
87 00:14:23.200 ⇒ 00:14:25.880 Uttam Kumaran: I just don’t know what is taking so long
88 00:14:26.070 ⇒ 00:14:29.989 Uttam Kumaran: like, I just, I just can’t articulate what’s happening.
89 00:14:30.380 ⇒ 00:14:37.240 Uttam Kumaran: And I’m not getting a clear answer from. And and we’re the only 3 people that can talk to the health of the project. There’s there’s only 3 people in the room.
90 00:14:46.417 ⇒ 00:14:49.753 Amber Lin: I can. I can start blurbing out things.
91 00:14:50.470 ⇒ 00:14:55.140 Amber Lin: or I can start on this. But Miguel kind of want to hear your perspective first.st
92 00:14:57.332 ⇒ 00:15:03.359 Miguel de Veyra: I mean one of the main things I’d say. Why, you we’re a bit misaligned because
93 00:15:04.220 ⇒ 00:15:19.300 Miguel de Veyra: the yeah for Utah. It’s like, I understand that. You know we’ve been trying to get this to work for 8 for 8 weeks, but I would say a good amount of that. A good chunk of that will stop on the data side of things like just trying to, you know, pull data out trying to make Dlt work.
94 00:15:19.530 ⇒ 00:15:26.780 Miguel de Veyra: And then on my end. We just started deploying. And then, just actually, you know, actioning stuff last week.
95 00:15:28.010 ⇒ 00:15:34.369 Miguel de Veyra: So that’s why I think it’s a bit misaligned. Because for me, we initialized all the agents we got. Even
96 00:15:35.610 ⇒ 00:15:38.369 Miguel de Veyra: linear is probably the only thing we don’t have.
97 00:15:38.580 ⇒ 00:15:43.100 Miguel de Veyra: The agents have access to right now, but the rest are like there, there!
98 00:15:43.360 ⇒ 00:15:44.520 Miguel de Veyra: So I think that’s where the.
99 00:15:44.520 ⇒ 00:15:46.759 Uttam Kumaran: But is anyone use? Is anyone using it.
100 00:15:47.380 ⇒ 00:15:55.079 Miguel de Veyra: I think the a wish is testing it out. We discussed tomorrow yesterday. Kyle is starting to test it out
101 00:15:58.870 ⇒ 00:16:02.289 Miguel de Veyra: because initially, initially sorry initializations was
102 00:16:03.072 ⇒ 00:16:05.809 Miguel de Veyra: like Tuesday and today, like some of it.
103 00:16:08.910 ⇒ 00:16:11.090 Miguel de Veyra: But that’s just from my perspective. Of course.
104 00:16:12.400 ⇒ 00:16:17.039 Uttam Kumaran: So we spent 7 weeks on data. But this is what I get. This. What I got to get to is
105 00:16:17.560 ⇒ 00:16:22.290 Uttam Kumaran: like, how can we prevent that from happening? Sure
106 00:16:23.270 ⇒ 00:16:26.090 Uttam Kumaran: and like. What part of our events?
107 00:16:27.060 ⇒ 00:16:30.380 Uttam Kumaran: Sorry, I said, what part of our plan prevents that in the future.
108 00:16:33.560 ⇒ 00:16:45.129 Miguel de Veyra: Actually a good, a good sample of that was basically last week that we yes, we did double work because I had to pull data from. I had to transform slab
109 00:16:45.827 ⇒ 00:16:58.639 Miguel de Veyra: but the the reason we had something to show for Yavi and Eden by Friday last week was because we didn’t really depend on, because I wish we’re still working on those stuff. But then, you know.
110 00:16:58.860 ⇒ 00:17:01.240 Miguel de Veyra: I was like, we can just do this.
111 00:17:01.470 ⇒ 00:17:11.250 Miguel de Veyra: Anyways, I can have it, I can pull it all on my own. But yeah, I think one of the things that we can avoid to do that is just basically to get something. We’re working 1st and then just
112 00:17:11.940 ⇒ 00:17:14.890 Miguel de Veyra: work work on improving it. Later down the road.
113 00:17:18.109 ⇒ 00:17:19.119 Uttam Kumaran: I think the challenge.
114 00:17:19.119 ⇒ 00:17:19.519 Miguel de Veyra: Like the.
115 00:17:19.520 ⇒ 00:17:23.539 Uttam Kumaran: Are you? There is that I’m like, I’m the main reason that
116 00:17:24.369 ⇒ 00:17:27.990 Uttam Kumaran: a wish got involved, and I’m the main reason that
117 00:17:28.200 ⇒ 00:17:36.150 Uttam Kumaran: everyone together and I made I forced that like, how does this project succeed without me? That
118 00:17:36.410 ⇒ 00:17:40.960 Uttam Kumaran: because I’m not gonna be like, if I didn’t check in.
119 00:17:42.100 ⇒ 00:17:46.960 Uttam Kumaran: we would have not done anything. And even 2 weeks ago, right? And I’m calling again.
120 00:17:47.520 ⇒ 00:17:51.459 Uttam Kumaran: So like, what’s I’m still not hearing like
121 00:17:51.600 ⇒ 00:17:58.940 Uttam Kumaran: how this thing actually gets into production and works and delivers an roi without me, like
122 00:17:59.140 ⇒ 00:18:05.360 Uttam Kumaran: basically micromanaging, you know, because it’s not clear that anyone else cares that this actually works
123 00:18:06.020 ⇒ 00:18:08.639 Uttam Kumaran: like, am I the only one that cares that this actually like
124 00:18:09.710 ⇒ 00:18:26.859 Uttam Kumaran: makes an impact. Because if that’s the case, then like, it’s not worth, it’s not worth doing like, even though we can do it. And the technology works. And I know it’ll make an impact. I can’t do this on my own. And so I’m I don’t want to do. I can’t work on projects anymore, or have us work on projects where I’m the point of failure.
125 00:18:27.140 ⇒ 00:18:30.419 Uttam Kumaran: Right? So that’s that’s my ultimate thing is that
126 00:18:30.680 ⇒ 00:18:39.949 Uttam Kumaran: I don’t have, although yes, I know we solve the problems we can solve any problem. If I’m and if I get involved we’ll solve the problem. But I’m not hearing like, how this happens without
127 00:18:40.230 ⇒ 00:18:42.440 Uttam Kumaran: without me, basically like
128 00:18:42.880 ⇒ 00:18:46.190 Uttam Kumaran: just asking again and again, like, Hey, where is this? Where is this? Where is this.
129 00:18:50.850 ⇒ 00:18:58.929 Amber Lin: Yeah, so let me give my perspective and how I think we should move forward. So right now.
130 00:19:01.520 ⇒ 00:19:09.480 Amber Lin: I think the problem that we have is oftentimes we get very, very zeroed in on individual steps.
131 00:19:09.950 ⇒ 00:19:18.680 Amber Lin: and then we try to figure out the individual steps and then scope kind of gets expanded. Whereas overall
132 00:19:20.225 ⇒ 00:19:37.139 Amber Lin: we some like sometimes because we have me pming, I’m not deep in engineering at work. I can say, Okay, when is this gonna get delivered. But I also get pulled into actually, the steps are actually figuring it out. And then I also lose perspective of okay, what?
133 00:19:37.780 ⇒ 00:19:54.470 Amber Lin: Like the big picture of why, this is even impactful. And what do we want to accomplish? So I think there’s there’s this constant cycle of how I’ve been feeling on this project being, we’re being dragged down into
134 00:19:54.760 ⇒ 00:19:58.039 Amber Lin: very specific small problems.
135 00:19:58.420 ⇒ 00:19:59.800 Amber Lin: And
136 00:20:01.470 ⇒ 00:20:24.040 Amber Lin: it’s kind of hard to just slash through everything. It feels like we’re trudging through much. Sometimes, like we have a lot of. We have deliverables. We are shipping. But I think if we, I think if we’re both honest with ourselves, and we zoom out into the bigger picture. Then I do think we could have
137 00:20:24.140 ⇒ 00:20:31.490 Amber Lin: one moved faster, and to moved in a way that
138 00:20:31.650 ⇒ 00:20:35.520 Amber Lin: not just getting our tasks done, but
139 00:20:35.710 ⇒ 00:20:44.370 Amber Lin: doing tasks so that they have positive impact on our clients, which is the company which is the company and people at the company.
140 00:20:46.950 ⇒ 00:20:53.849 Amber Lin: so to wrap that up. It’s like we get dragged down. And 2, we’re not client focused enough.
141 00:21:01.890 ⇒ 00:21:19.309 Uttam Kumaran: So I I mean, I think, like, this is where you know it sort of goes into the next step is, I just want to know? Like, okay, let’s say we’re we all. I think we’re in agreement, right? Like, I sort of agree with amber. And again, I have perspective. I’ve been a Pm. Before. I know sort of how this works. But I want to sort of leave
142 00:21:19.410 ⇒ 00:21:26.370 Uttam Kumaran: my perspective on that for you guys, because I want you guys to come to the real same realization and for us to get a line.
143 00:21:26.530 ⇒ 00:21:28.880 Uttam Kumaran: So I feel like at this point, like, I don’t know.
144 00:21:29.120 ⇒ 00:21:47.559 Uttam Kumaran: I feel pretty good about. Okay, this is the current state. I think, Miguel, I just want to know from you that, like everything that we’re saying and I don’t. And again, I don’t want you to just agree just to be like, I agree now that I get it like I sort of want it to click with you as well. That like, hey? This project is actually
145 00:21:47.740 ⇒ 00:21:53.664 Uttam Kumaran: not succeeding. In fact, we’re like at the brink of it, being nixed like entirely
146 00:21:54.270 ⇒ 00:21:59.700 Uttam Kumaran: right like. That’s where I I wanna make sure that you’re you’re you’re also seeing that before we
147 00:21:59.990 ⇒ 00:22:01.809 Uttam Kumaran: we move forward.
148 00:22:07.210 ⇒ 00:22:08.000 Miguel de Veyra: Yep
149 00:22:11.920 ⇒ 00:22:12.720 Miguel de Veyra: nice.
150 00:22:22.510 ⇒ 00:22:26.280 Miguel de Veyra: So I think one of the things that we need to work on is not only like the
151 00:22:26.730 ⇒ 00:22:31.979 Miguel de Veyra: the development, of course, but also how like we need to. Sorry I’m a bit.
152 00:22:33.170 ⇒ 00:22:33.859 Miguel de Veyra: No, no, you’re good.
153 00:22:35.502 ⇒ 00:22:38.760 Uttam Kumaran: Adaptation. Right? We actually have to work on adaptation.
154 00:22:45.600 ⇒ 00:22:47.070 Miguel de Veyra: Sorry? Does that make sense.
155 00:22:50.220 ⇒ 00:22:55.989 Uttam Kumaran: Right like this is where ultimately we we talked about like what each of our
156 00:22:56.700 ⇒ 00:23:13.919 Uttam Kumaran: responsibilities are. But each of us has a as a sphere of control. And so each of us has responsibilities for me. I can control making sure you guys have resources, making sure the vision is set and helping push adoption right? Like bringing people together. What I can’t do is
157 00:23:14.070 ⇒ 00:23:18.789 Uttam Kumaran: I can’t be involved in telling you how Rag works and the fact that we shouldn’t do
158 00:23:19.050 ⇒ 00:23:25.629 Uttam Kumaran: like we shouldn’t throw everything into context. I can’t go to Amber and go to every single meeting and say
159 00:23:25.910 ⇒ 00:23:39.709 Uttam Kumaran: you need to call on people and make sure we get outcomes out of these meetings. I can’t do that. I could do it once or twice, but I can’t do it every time right? And so this is where for me, if I don’t see what my.
160 00:23:39.850 ⇒ 00:23:58.279 Uttam Kumaran: what my understanding, and what I believe in the company is that I can go. I can make an immediate impact, but I try to share and coach and then leave everyone to go. Act on it on your own. But I can’t. Ultimately, if my job is to give you guys the belief that if we do this we will see Xyz impact.
161 00:23:58.390 ⇒ 00:24:03.499 Uttam Kumaran: But then I need you. Y’all to implement the execution, and then tell me where you need me to help.
162 00:24:03.680 ⇒ 00:24:25.030 Uttam Kumaran: So this is where, like just hearing that amber is like in the weeds on technical side. That’s not her responsibility. That’s not her sphere of control. That’s not her superpower right? Her superpower is that she can connect people. She runs the meeting, she gets the outcome we need, and then she communicates to me, hey? The project is moving along fine, and I have confidence in that.
163 00:24:25.220 ⇒ 00:24:47.190 Uttam Kumaran: because for the last few weeks I I know from both you guys and I joined the stand up. So I know that if I was to ask you, hey? Is the project going fine? Both of you are like, yes, but I know that I’m not seeing any adoption visibly. I know we’re not measuring anything. So there’s nothing even to look at, and I can’t even use it like as our own user. So I know that the project isn’t going well. So that’s where I’m like.
164 00:24:47.300 ⇒ 00:24:50.299 Uttam Kumaran: okay, like, there is a big disconnect. And so
165 00:24:50.400 ⇒ 00:25:09.780 Uttam Kumaran: part of it is like an ownership mentality like owning this end to end, and feeling like, Hey, I only feel pride when this thing actually goes and helps Kyle save half an hour or helps helps us like speed up a delivery deliverable for a client. Right? The second thing is understanding where your role starts and stops
166 00:25:09.840 ⇒ 00:25:29.069 Uttam Kumaran: right? And I think that’s what the next little exercise is is looking at each of the key components. And this is something I know, Miguel, you’ve expressed, and I also mentioned to Amber is that I think now that we’re a growing company. We’re starting to decide on like roles and responsibilities previously. It’s just like whoever on the team and then me taking everything else. Now that we have a Pm structure.
167 00:25:29.070 ⇒ 00:25:40.369 Uttam Kumaran: we’re we’re working on engineering management structure working on technical lead structure. I think it’s important for us to actually like very clearly articulate who owns what and why, and this we can debate.
168 00:25:40.540 ⇒ 00:25:49.529 Uttam Kumaran: and then also where the 2 roles need to work together. And that’s what the the thing on the on the bottom there is, which is basically like what I feel.
169 00:25:49.930 ⇒ 00:26:12.509 Uttam Kumaran: This is basically what I feel like each role should own. And then I kind of want us to go through and maybe even write like who owns this today. And it could be empty right? It could be like nobody’s doing this, or I thought you were doing this, or I thought we were doing this together. But does that seem like worthwhile, maybe to just spend a few minutes on on this and like, or is is it more impactful just to keep sort of
170 00:26:12.760 ⇒ 00:26:14.799 Uttam Kumaran: talking about what we’ve been talking about?
171 00:26:15.381 ⇒ 00:26:25.939 Amber Lin: Totally. I think I I would love to do the roles part, but before we start, can we align on what’s good or done for the project cause, I think
172 00:26:26.280 ⇒ 00:26:27.790 Amber Lin: ultimately
173 00:26:28.250 ⇒ 00:26:36.390 Amber Lin: there’s misalignment on even the starting point of what’s good. So can we spend like 30 seconds on this?
174 00:26:36.600 ⇒ 00:26:45.180 Amber Lin: Each person. Just say your current under current understanding. Well, Utam already said it. What should? What will good look like for this project?
175 00:26:45.410 ⇒ 00:26:46.770 Amber Lin: Mika? Would you.
176 00:26:46.770 ⇒ 00:26:57.990 Uttam Kumaran: Yeah, I think I’ll just quick thing on my side is just that people can use the slack agents, and it takes care of 20 to 40% of the most common questions about clients. That’s it.
177 00:26:58.190 ⇒ 00:26:59.849 Uttam Kumaran: Like very plain and simple.
178 00:27:09.930 ⇒ 00:27:12.010 Amber Lin: And then go. What does it look like for you.
179 00:27:18.900 ⇒ 00:27:24.179 Miguel de Veyra: good. Except for what Utam says I agree with that would have been thinking.
180 00:27:25.380 ⇒ 00:27:28.740 Miguel de Veyra: I think, like to focus on adoption
181 00:27:29.040 ⇒ 00:27:33.730 Miguel de Veyra: like no cause, you know, basically teach them on how to use it.
182 00:27:34.050 ⇒ 00:27:41.160 Miguel de Veyra: I think I mean, you know. Honestly, I would just focus on Kyle first, st because he’s he seems to be very, very interested
183 00:27:41.390 ⇒ 00:27:48.329 Miguel de Veyra: on using the agent like he was very attentive in the call yesterday he was actually participating. And he wants to know.
184 00:27:48.590 ⇒ 00:27:52.340 Miguel de Veyra: So honestly. Yeah, I think one.
185 00:27:52.620 ⇒ 00:27:52.940 Uttam Kumaran: Yeah.
186 00:27:52.940 ⇒ 00:28:03.289 Miguel de Veyra: What is good for this project would be to just have one focus on one use case 1st like one client, get that perfect and then just duplicate that into other ones.
187 00:28:05.970 ⇒ 00:28:08.500 Miguel de Veyra: More of a more of a more of an action, action.
188 00:28:08.500 ⇒ 00:28:28.489 Amber Lin: Yeah, that’s great. That’s a great path to succeeding. So just to summarize, I think you’re saying that how you would define good for this project is that people have adopted it. But, like, what does adoption look like like? What does it? What does it mean?
189 00:28:29.120 ⇒ 00:28:35.909 Miguel de Veyra: Like not only adopting it like basically actually use it to the full extent of its capabilities.
190 00:28:36.520 ⇒ 00:28:37.890 Amber Lin: Oh!
191 00:28:38.040 ⇒ 00:28:38.860 Amber Lin: Whoa!
192 00:28:39.970 ⇒ 00:28:40.360 Miguel de Veyra: And then.
193 00:28:40.360 ⇒ 00:28:41.860 Amber Lin: Its capabilities.
194 00:28:42.010 ⇒ 00:29:03.290 Miguel de Veyra: Yeah, basically, what a wish said yesterday was based on the slack messages. Can you look into Yada Yada? Basically, it’s multi tool based on the slack messages. It’ll get context from that. It’ll go get context from the Zoom Meetings and then from Github, and then it’ll combine that into one. Maybe after we implement linear, it can also find that ticket. And then Yada Yada.
195 00:29:04.170 ⇒ 00:29:04.670 Amber Lin: Okay.
196 00:29:04.670 ⇒ 00:29:06.509 Uttam Kumaran: That’s a great example of like
197 00:29:06.760 ⇒ 00:29:14.889 Uttam Kumaran: you’re you’re what you’re saying is that in a short term it needs to work, and it needs to work in the long term, right? Like, those are technical capabilities.
198 00:29:14.950 ⇒ 00:29:29.959 Uttam Kumaran: Because for me again, that’s where my mind goes. But as a business leader I’m like 20 to 40% of questions. I don’t care what it is it just needs to handle for you. I think you need to think of cool. We need to categorize those questions. Some are easier, some are hard, some are multi tool.
199 00:29:29.960 ⇒ 00:29:50.679 Uttam Kumaran: So this is where, as the technical lead, I’m what my, I think each of our expectations. And again, there are expectations you guys should have for me, and we should have them for each other. Is that the engineering is done well, and that there’s a that for any problem we face or lacking capability, there is a path forward, and there is a communicated
200 00:29:50.690 ⇒ 00:29:54.390 Uttam Kumaran: timeline and communicated engineering, right?
201 00:29:54.410 ⇒ 00:30:09.979 Uttam Kumaran: Those things are what enables amber to then come to me and say, Hey, I know there’s this blocker we’re at currently at 5%. We need to get the 10%. Here is the 5% where it comes from. Out of that. It’s gonna take 2 weeks to do this 3 weeks to do this, and I am 90% confident will get done
202 00:30:10.180 ⇒ 00:30:13.849 Uttam Kumaran: right. What I’m hearing now right to contrast is.
203 00:30:14.180 ⇒ 00:30:29.269 Uttam Kumaran: I go to. I know the linear board. I know everything that’s going on. So it’s not. We’re not working on a whole lot. And I am like, Okay, cool. I know I had to come in. We broke down all of the integration step by step by step. They’re getting taken care of. I poked into S. 3. I kind of know they’re there.
204 00:30:29.510 ⇒ 00:30:41.140 Uttam Kumaran: I don’t. I haven’t seen anyone on the data team use it yet to sort of say, it’s working. I sort of see you and Casey testing it and sort of showing signs of life.
205 00:30:41.280 ⇒ 00:30:56.095 Uttam Kumaran: But then I cast amber. She’s like, well, I don’t know. Like we’re hitting these roadblocks. I go to the rag meeting. I hear it’s only 30 days of slack messages. I’m like, what about the rest of them? It’s like that. That’s this. That’s the contrast of like where we are now. Right?
206 00:30:56.600 ⇒ 00:31:02.689 Uttam Kumaran: so, and then, I guess, amber like, what do you think.
207 00:31:04.250 ⇒ 00:31:07.889 Amber Lin: And for me, what defines good for this project is that
208 00:31:09.210 ⇒ 00:31:17.459 Amber Lin: using like usage is different than impact. Right? So we talk about that. Yes, it
209 00:31:17.690 ⇒ 00:31:29.509 Amber Lin: it it. They use it to answer their questions. But does it actually save them time? Or is it just a fun thing that they use like is, it is the. Does it provide.
210 00:31:29.510 ⇒ 00:31:30.250 Uttam Kumaran: Yeah. Good, point.
211 00:31:30.618 ⇒ 00:31:43.160 Amber Lin: That means that. Does the answer quickly, accurately. Does it save them additional steps, or do they still have to look at the same thing, even if they asked to ask the slack bot?
212 00:31:43.300 ⇒ 00:31:52.880 Amber Lin: And can the uses continue, continue, connect, continue throughout time, and can we see that the costs
213 00:31:53.130 ⇒ 00:31:59.500 Amber Lin: of a project goes down while we’re delivering the same thing. So that that means
214 00:32:00.050 ⇒ 00:32:03.390 Amber Lin: the ticket points are pretty much the same. But then.
215 00:32:03.760 ⇒ 00:32:04.310 Uttam Kumaran: Yeah.
216 00:32:04.310 ⇒ 00:32:13.520 Amber Lin: Monthly or weekly. When we look at allocations. The the ratio between velocity and cost.
217 00:32:14.376 ⇒ 00:32:18.030 Amber Lin: It’s getting better. So that I think ultimately.
218 00:32:18.170 ⇒ 00:32:25.139 Amber Lin: this is what we’re trying to achieve along with other projects along with data platform. But ultimately, it’s just, can we
219 00:32:25.790 ⇒ 00:32:31.489 Amber Lin: can we do better like? Can we can we do better as a in financial health, wise.
220 00:32:33.750 ⇒ 00:32:34.650 Uttam Kumaran: I agree.
221 00:32:36.020 ⇒ 00:32:43.363 Amber Lin: You think we both like? We all have like complementing perspectives, and all of these parts are really necessary.
222 00:32:44.460 ⇒ 00:32:48.849 Uttam Kumaran: Yeah, I think your amber, your expectations are the highest. Mine are
223 00:32:49.080 ⇒ 00:33:17.379 Uttam Kumaran: are somewhere in the middle, and Miguel’s, I think maybe somewhere in the middle. But for me, cause I just want to see that if people are using it because I know people start to use it, and they see they see usage, we’ll fix the rest of them. But this is also where, like I’m now not even confident that we can even get there right? Cause I’m like, it’s gonna be another 8 weeks like, if I spend a weekend, I think I would figure this out and we get there going. But that’s the challenge for me is, I’m like, I don’t have a clear articulation. I would
224 00:33:18.090 ⇒ 00:33:34.199 Uttam Kumaran: doing this. But I Googled slack, based rag agent over messages I learned about like embedding similarities. I learned about like all these metadata creation yesterday, like.
225 00:33:34.490 ⇒ 00:33:39.650 Uttam Kumaran: but that’s what I’m saying is, I was just surprised to know that we didn’t sort of do that before.
226 00:33:47.220 ⇒ 00:33:50.570 Amber Lin: Well, let’s move on to the roles.
227 00:33:51.860 ⇒ 00:33:52.430 Uttam Kumaran: Okay.
228 00:33:52.680 ⇒ 00:33:55.090 Amber Lin: Because I think we all have an understanding.
229 00:33:55.200 ⇒ 00:34:02.360 Amber Lin: We have okay, what’s been falling behind. And let’s clarify that on like
230 00:34:02.550 ⇒ 00:34:14.589 Amber Lin: how? Because roles make up the whole project and make it makes it running. Let’s see where we’re at, who we think is doing it, and then see where our expectations misalign
231 00:34:15.595 ⇒ 00:34:23.280 Amber Lin: and then we can allocate. Okay, who’s responsible for this falling through? And who’s responsible for this pushing like pushing forward.
232 00:34:23.580 ⇒ 00:34:25.449 Amber Lin: So let’s talk about
233 00:34:27.969 ⇒ 00:34:39.830 Amber Lin: let’s just go through each one. I I know there’s not a completely comprehensive but just, Miguel, tell me what you feel like who you feel is owning it right now.
234 00:34:45.739 ⇒ 00:34:47.559 Miguel de Veyra: Planning milestones.
235 00:34:48.860 ⇒ 00:34:53.332 Amber Lin: I I feel like right now, especially for milestones.
236 00:34:54.310 ⇒ 00:34:58.650 Amber Lin: like overall plan and milestones has been Utah.
237 00:34:59.050 ⇒ 00:35:00.170 Amber Lin: That’s been.
238 00:35:00.330 ⇒ 00:35:02.400 Uttam Kumaran: Utam has been helping on this.
239 00:35:03.030 ⇒ 00:35:04.000 Amber Lin: Yeah, I agree.
240 00:35:04.000 ⇒ 00:35:08.560 Amber Lin: Schedule. When it comes to delivery schedule, I would say that
241 00:35:12.380 ⇒ 00:35:24.406 Amber Lin: like we don’t have a completely clear schedule, because everything is kind of floating in the air. At least that’s how I feel. I I know there’s path forwards for me to improve it.
242 00:35:24.870 ⇒ 00:35:30.270 Amber Lin: And how for the team to improve it. I would say, I’m helping a bit on this, and
243 00:35:34.570 ⇒ 00:35:37.759 Amber Lin: Mutam is also helping a bit on this, Miguel. What do you feel.
244 00:35:38.540 ⇒ 00:35:39.500 Miguel de Veyra: Yeah, I agree.
245 00:35:39.890 ⇒ 00:35:40.640 Amber Lin: Okay.
246 00:35:43.030 ⇒ 00:35:49.810 Uttam Kumaran: Yeah. But like, you know, even even, I just feel like every time I ask. It’s like, Oh, I need another meeting to figure it out.
247 00:35:50.040 ⇒ 00:35:51.920 Uttam Kumaran: or like we need to do a spike
248 00:35:52.040 ⇒ 00:36:03.700 Uttam Kumaran: or like I need another week, or we just met to figure it out. It’s like nothing concrete. And I know. And again, like I’m an engineer. I know all those things are just to basically dodge
249 00:36:04.330 ⇒ 00:36:08.590 Uttam Kumaran: saying an answer which is fine, but it can’t be like that for everything.
250 00:36:09.070 ⇒ 00:36:12.459 Uttam Kumaran: Otherwise, if I don’t know like when this is gonna happen
251 00:36:12.880 ⇒ 00:36:21.400 Uttam Kumaran: like I can’t find, I can’t fund it. We don’t, and we don’t treat any of our clients this way, right like there’s no client we’ve worked for where we’re like. I don’t know when it’s gonna happen.
252 00:36:21.690 ⇒ 00:36:24.899 Uttam Kumaran: So like, honestly, we should treat ourselves
253 00:36:25.250 ⇒ 00:36:29.060 Uttam Kumaran: as good as maybe even better than we treat our our clients.
254 00:36:29.570 ⇒ 00:36:34.239 Uttam Kumaran: Which is like, yeah, we wanna crush it for us. Right? We have all the budget.
255 00:36:34.350 ⇒ 00:36:39.170 Uttam Kumaran: We have no blockers, no weird client stuff. And like we should go for gold, you know.
256 00:36:39.360 ⇒ 00:36:41.290 Uttam Kumaran: And so yeah, I agree. I think
257 00:36:41.680 ⇒ 00:36:45.680 Uttam Kumaran: you know for me. I fit in all these. I fit in some of these, but
258 00:36:46.130 ⇒ 00:36:50.410 Uttam Kumaran: I want to. I want to help. I don’t want to be the I can’t be the primary right.
259 00:36:50.600 ⇒ 00:36:57.980 Uttam Kumaran: That’s where you can leverage me the most is like I can come into any of these and assist, but that is all it can be. It can just be an assist.
260 00:37:04.070 ⇒ 00:37:06.950 Amber Lin: Let’s go on to the next one.
261 00:37:07.260 ⇒ 00:37:18.319 Amber Lin: I think this one while I’m running, I would say, save think so,
262 00:37:20.970 ⇒ 00:37:23.580 Amber Lin: Miguel, tell me. Tell me what you think.
263 00:37:26.290 ⇒ 00:37:27.410 Miguel de Veyra: You mean? I’m
264 00:37:30.280 ⇒ 00:37:44.380 Miguel de Veyra: oh, yeah, the daily things. Yes, I mean, we do do that for the past couple of weeks. So yeah, you’ve been running that sort of me. And you daily things, I’m completely out of the loop for sales or sets. Oh, it sets. Okay. Yeah.
265 00:37:44.520 ⇒ 00:37:46.800 Miguel de Veyra: So yeah, I think that’s yeah.
266 00:37:47.330 ⇒ 00:37:48.710 Miguel de Veyra: I think that’s more on you.
267 00:37:49.170 ⇒ 00:37:53.859 Amber Lin: Hmm, okay, let’s do. Let’s do this one. Then.
268 00:38:04.400 ⇒ 00:38:08.260 Miguel de Veyra: I think now it’s it’s primarily me.
269 00:38:14.620 ⇒ 00:38:15.929 Amber Lin: Oh, John, what do you feel?
270 00:38:18.110 ⇒ 00:38:19.200 Uttam Kumaran: I mean, I think
271 00:38:19.460 ⇒ 00:38:23.680 Uttam Kumaran: I mean, I think you have to own it. But I would say, I don’t see
272 00:38:23.840 ⇒ 00:38:25.950 Uttam Kumaran: this happening today, like.
273 00:38:27.340 ⇒ 00:38:33.380 Uttam Kumaran: I mean, this is where it’s like, if if I’m able, if I if for me, if I’m like cool, I can do the architecture work where I’m like.
274 00:38:33.590 ⇒ 00:38:35.759 Uttam Kumaran: we have Github. We have all these tools.
275 00:38:35.960 ⇒ 00:38:45.189 Uttam Kumaran: these need to go in these ways. We’re using super base in this way. Are we doing embedding similarity? How are we doing this kind of Brad context like, I don’t think
276 00:38:45.530 ⇒ 00:38:51.069 Uttam Kumaran: my feeling is that we’re not doing this at all like, I think we’re spending all of our time on executing
277 00:38:51.420 ⇒ 00:38:56.720 Uttam Kumaran: and no time on like broader research. Are we using the right tools?
278 00:38:56.930 ⇒ 00:39:04.060 Uttam Kumaran: How is it impacting speed cost efficiency like? Is our architecture sound
279 00:39:04.813 ⇒ 00:39:10.889 Uttam Kumaran: because, you know, I I don’t know sometimes right like I don’t know whether some some of these things are
280 00:39:11.580 ⇒ 00:39:13.220 Uttam Kumaran: right, but like
281 00:39:13.830 ⇒ 00:39:21.459 Uttam Kumaran: I have to go research. And I could go research. Right. I go, and I leverage a bunch of sources, and I find it. But like this is where I don’t think this is happening right now.
282 00:39:22.780 ⇒ 00:39:28.119 Uttam Kumaran: So we tried to do the which, like I’ve been asking for months for someone to go do a deep dive into rag.
283 00:39:28.230 ⇒ 00:39:34.649 Uttam Kumaran: And still it hasn’t happened. So I then go and find contextual that ends up being like a huge distraction, like
284 00:39:35.010 ⇒ 00:39:40.540 Uttam Kumaran: we’re not using that product. Right? So like, I just, I’m trying to get something because we can’t do it ourselves.
285 00:39:40.670 ⇒ 00:39:46.869 Uttam Kumaran: The second piece is, I go yesterday, and I can’t think nobody knows how to do rag on slack messages.
286 00:39:47.030 ⇒ 00:39:50.069 Uttam Kumaran: So then I sit with Chat gpt for 10 min, and I figure it out.
287 00:39:50.600 ⇒ 00:39:54.590 Uttam Kumaran: and I have all the answers, but like this is what was frustrating to me is like
288 00:39:55.180 ⇒ 00:39:58.200 Uttam Kumaran: I can’t do. I’m not. I can’t do that guy like
289 00:39:58.670 ⇒ 00:40:11.090 Uttam Kumaran: I can’t be the one to to solve those problems. Otherwise, like, I’m I like, you know, that’s it, they said. Can’t be the one to solve those problems. So this is where I feel like we, we should be leaning more into this.
290 00:40:11.320 ⇒ 00:40:22.430 Uttam Kumaran: And I think Miguel, more of your time needs to go into this versus like waiting to be told how to do things or waiting to be given the right instructions like this needs to be research, you know.
291 00:40:22.700 ⇒ 00:40:25.009 Uttam Kumaran: That’s my, that’s my feeling.
292 00:40:33.080 ⇒ 00:40:33.940 Uttam Kumaran: Thank you.
293 00:40:35.060 ⇒ 00:40:40.720 Miguel de Veyra: For that one, though, should we, when cause in linear right, should we?
294 00:40:41.760 ⇒ 00:40:44.429 Miguel de Veyra: Should we create tickets for research tickets or.
295 00:40:45.200 ⇒ 00:40:52.180 Uttam Kumaran: But like, this is what I’m saying. What was, what was gonna be the plan for doing larger than 30 days worth of data.
296 00:40:53.509 ⇒ 00:40:56.589 Miguel de Veyra: Basically, we embedded everything
297 00:40:57.295 ⇒ 00:41:09.339 Miguel de Veyra: we embedded every per message. But then I thought, we discussed it. And then what we’re gonna do is, basically, we’re gonna chunk, for example, 2 weeks worth of messages into one. We’re gonna basically chunk it
298 00:41:09.830 ⇒ 00:41:11.680 Miguel de Veyra: and then embed that.
299 00:41:12.030 ⇒ 00:41:16.999 Miguel de Veyra: And then, basically, you know, we can use that alphabet. We decided 2 weeks, because that’s cycle.
300 00:41:17.799 ⇒ 00:41:20.310 Uttam Kumaran: I know. But when did you decide this yesterday?
301 00:41:22.715 ⇒ 00:41:24.739 Miguel de Veyra: Yeah, earlier, today, yes.
302 00:41:25.230 ⇒ 00:41:30.189 Uttam Kumaran: But do you see, do you see what I’m saying like, unless I said something in that meeting this would not have happened.
303 00:41:33.720 ⇒ 00:41:38.150 Uttam Kumaran: because the answer I got in the meeting was, you’re gonna shove it all into context.
304 00:41:39.220 ⇒ 00:41:42.589 Uttam Kumaran: Nobody talked about embeddings. Nobody talked about metadata
305 00:41:43.170 ⇒ 00:41:49.430 Uttam Kumaran: like nobody, talked about different types of search, pre, filtering, ranking summarization. Nobody said anything like that.
306 00:41:52.720 ⇒ 00:41:55.660 Uttam Kumaran: So what I what I need to know in this meeting is like.
307 00:41:56.000 ⇒ 00:41:59.810 Uttam Kumaran: if you’re gonna take on that like playing that role
308 00:42:00.800 ⇒ 00:42:04.429 Uttam Kumaran: which is like, Hey, this is not a good good enough solution.
309 00:42:04.560 ⇒ 00:42:16.209 Uttam Kumaran: or Hey, we don’t know. And we should have spent time researching this before we got to this point, or if I need to take on that responsibility, and if I take it on, then it has, then I have to. We have to go find someone else to basically go do that
310 00:42:16.360 ⇒ 00:42:18.060 Uttam Kumaran: because it’s not happening today.
311 00:42:29.540 ⇒ 00:42:36.639 Uttam Kumaran: Well, my question is like, are, do you want to take that and can and do you feel enabled and capable of doing that?
312 00:42:36.880 ⇒ 00:42:43.699 Uttam Kumaran: It’s not the first, st not the 1st time I’ve asked like, hey? Can someone go figure out how to do rag on more than just Google Docs?
313 00:42:44.190 ⇒ 00:42:47.430 Uttam Kumaran: That’s something we all could have figured out 6 or 9 months ago.
314 00:42:47.650 ⇒ 00:42:54.730 Uttam Kumaran: But I’m I’m still not finding any support for me on figuring that out. Nobody has nobody else
315 00:42:54.880 ⇒ 00:42:58.919 Uttam Kumaran: apart from me. You or Casey and Amber now can go look into that.
316 00:42:59.030 ⇒ 00:43:01.379 Uttam Kumaran: And so who’s responsible for that?
317 00:43:07.100 ⇒ 00:43:07.750 Miguel de Veyra: Oh!
318 00:43:13.600 ⇒ 00:43:19.910 Uttam Kumaran: You could say. I don’t want to take that like, or you could say, I don’t feel like taking that, or I don’t feel capable of taking that. But then
319 00:43:20.130 ⇒ 00:43:22.109 Uttam Kumaran: I have to go hire someone that can do that.
320 00:43:23.530 ⇒ 00:43:27.839 Uttam Kumaran: or I have to. I have to basically play Angelad until we find someone
321 00:43:28.810 ⇒ 00:43:31.529 Uttam Kumaran: that’s it’s just like that’s those are the 2 routes I have.
322 00:43:39.280 ⇒ 00:43:42.200 Uttam Kumaran: I mean, what? What’s your gut instinct like? What do you think.
323 00:43:47.380 ⇒ 00:43:55.399 Miguel de Veyra: Honestly, my gut instinct is because basically, the only rag we know is the super based thing.
324 00:43:56.350 ⇒ 00:43:57.370 Uttam Kumaran: Right.
325 00:43:57.900 ⇒ 00:44:00.240 Miguel de Veyra: Like we can put it there, embedded.
326 00:44:00.820 ⇒ 00:44:02.140 Uttam Kumaran: Yeah, I know what? Not.
327 00:44:02.350 ⇒ 00:44:07.560 Miguel de Veyra: And then when we, I look, when I looked into contextual like, they built
328 00:44:07.720 ⇒ 00:44:10.980 Miguel de Veyra: their entire basically business on rug.
329 00:44:11.680 ⇒ 00:44:18.260 Miguel de Veyra: So you know, there’s some part of me. That is that something that we want to build somewhat? Because
330 00:44:19.050 ⇒ 00:44:25.710 Miguel de Veyra: if if that’s something that we wanna look into building, I don’t think that’s something you know.
331 00:44:27.020 ⇒ 00:44:29.490 Miguel de Veyra: I can build personally, or even with Casey.
332 00:44:30.700 ⇒ 00:44:33.100 Uttam Kumaran: But this is where I completely disagree like
333 00:44:34.790 ⇒ 00:44:40.590 Uttam Kumaran: I I Googled for 10 min yesterday, and I’ve I’ve been reading about Rag for more than 2 years.
334 00:44:40.960 ⇒ 00:44:48.360 Uttam Kumaran: We can build simple, pre-filtering, simple, vector embedding simple re-ranking and set up tools.
335 00:44:49.140 ⇒ 00:44:52.209 Uttam Kumaran: Right? Contextual is 20 grand a year.
336 00:44:53.720 ⇒ 00:45:05.800 Uttam Kumaran: This is where, like I I it’s just hard for me to get that like we didn’t do any sort of trade off analysis or nobody just did a simple Google search like, did you or Casey do a simple Google search on like.
337 00:45:06.100 ⇒ 00:45:13.380 Uttam Kumaran: Hey, I have all these. I have all these transactional messages that I want to do rag over. How can I do that?
338 00:45:13.490 ⇒ 00:45:21.350 Uttam Kumaran: What I’m hearing now is someone telling me that something’s impossible, and I will bet the whole company on the fact that this is not impossible.
339 00:45:22.840 ⇒ 00:45:28.030 Uttam Kumaran: In fact, I’ll bet even more that, like we could do it probably next 2 days
340 00:45:28.390 ⇒ 00:45:31.130 Uttam Kumaran: if someone just spent 10 min and googled it.
341 00:45:33.600 ⇒ 00:45:37.430 Uttam Kumaran: But what? What I’m what the point I’m trying to get to is if you don’t feel that
342 00:45:37.610 ⇒ 00:45:40.690 Uttam Kumaran: that that’s something you can take on, or if you’re like, Hey.
343 00:45:40.940 ⇒ 00:45:46.659 Uttam Kumaran: I’m just here to execute and like figuring out how to do it or the strategies is not my strong suit.
344 00:45:46.860 ⇒ 00:45:48.709 Uttam Kumaran: I that’s what I want to hear.
345 00:45:51.910 ⇒ 00:45:55.829 Uttam Kumaran: because there’s 1 thing, if you’re like, fundamentally, hey? I don’t believe we can do rag
346 00:45:56.080 ⇒ 00:45:58.190 Uttam Kumaran: on the slack messages.
347 00:45:59.280 ⇒ 00:46:07.159 Uttam Kumaran: But there’s another thing where it’s like, I don’t feel like I can go figure out how to do it. Those are 2 separate things like one is a belief. One is like
348 00:46:07.540 ⇒ 00:46:09.030 Uttam Kumaran: a truth right?
349 00:46:28.830 ⇒ 00:46:35.629 Uttam Kumaran: I mean, I’m happy to make it. I’m happy to make a decision, but I just want it. I want to give you the I want to give you the opportunity
350 00:46:36.300 ⇒ 00:46:38.860 Uttam Kumaran: if you want it like. That’s all I’m saying.
351 00:46:39.680 ⇒ 00:46:40.840 Miguel de Veyra: Yeah, I mean.
352 00:46:41.160 ⇒ 00:46:56.740 Miguel de Veyra: yeah, looking back, I don’t know why we just did Google this because part of the reason I think we just didn’t Google it because we stuck with what works best currently for ABC, because we’re basically chasing deadlines. So basically just context, everything we’ll work on improving that later on.
353 00:46:56.740 ⇒ 00:46:58.610 Uttam Kumaran: We’re not talking about ABC at all.
354 00:46:58.610 ⇒ 00:47:03.319 Miguel de Veyra: No, no, like, you know, because we want to deliver fast for multiple clients.
355 00:47:03.320 ⇒ 00:47:07.710 Miguel de Veyra: So just yeah, we’ll do what works best first, st and then.
356 00:47:07.710 ⇒ 00:47:14.579 Uttam Kumaran: Dude. How long does it? How long does Google searching like spent? I literally spent 10 min in between the meeting yesterday
357 00:47:14.880 ⇒ 00:47:17.109 Uttam Kumaran: cause I was just so surprised. I was like.
358 00:47:17.770 ⇒ 00:47:22.019 Uttam Kumaran: there’s no one on this call. Google, how to rag transactional messages
359 00:47:22.730 ⇒ 00:47:27.980 Uttam Kumaran: like not documents, not images, not videos. Transactions like
360 00:47:28.150 ⇒ 00:47:33.409 Uttam Kumaran: like row based polymer data. And then I was like.
361 00:47:34.250 ⇒ 00:47:45.669 Uttam Kumaran: I’ve read so much on this topic. I was like, cool, yeah, we just need to have tons of metadata do some pre filtering. And then, yeah, I was like, Okay, cool. I remember how re-ranking works things like that.
362 00:47:46.050 ⇒ 00:47:49.790 Uttam Kumaran: This is where I’m like, oh, it’s just like nobody Googled us.
363 00:47:51.580 ⇒ 00:47:56.160 Uttam Kumaran: Because if you googled it, I’m reading blogs from 2023, and they’re doing this.
364 00:47:58.410 ⇒ 00:47:59.430 Uttam Kumaran: It’s just fun.
365 00:47:59.810 ⇒ 00:48:03.990 Uttam Kumaran: This is where I’m just like what like to put yourself in my shoes like, what do I do at this point
366 00:48:05.140 ⇒ 00:48:11.210 Uttam Kumaran: like, what like do I go build it myself, or do I like?
367 00:48:11.490 ⇒ 00:48:14.849 Uttam Kumaran: Come to this crew? And I’m like, Hey, this is the challenge. Right now that we’re doing
368 00:48:18.920 ⇒ 00:48:20.000 Uttam Kumaran: pretty long.
369 00:48:20.680 ⇒ 00:48:30.010 Uttam Kumaran: I get your. I get your point, which is like, Hey, we want it to deliver. ABC, fine. But dude, this is 10 min of work to plan what’s gonna happen next.
370 00:48:31.300 ⇒ 00:48:33.299 Uttam Kumaran: This is where I feel like we’re.
371 00:48:35.560 ⇒ 00:48:40.079 Uttam Kumaran: This is where someone has to take this on. Because we, how are we gonna do this technology? Then.
372 00:48:41.650 ⇒ 00:48:43.479 Miguel de Veyra: Yeah, I agree. I guess I should.
373 00:48:43.710 ⇒ 00:48:48.670 Miguel de Veyra: Yeah, I think I was like, too deep to just basically just deliver.
374 00:48:48.670 ⇒ 00:48:49.900 Uttam Kumaran: Like. I think the other.
375 00:48:50.200 ⇒ 00:48:50.739 Miguel de Veyra: Like it didn’t.
376 00:48:50.740 ⇒ 00:48:51.980 Miguel de Veyra: And I’m not this.
377 00:48:52.250 ⇒ 00:49:01.640 Uttam Kumaran: Yeah, and I’m not. I’m not trying to get you to be like oh, it’s my fault. Woe is me like that’s not what the point here. But we, the 3 of us, just need to agree
378 00:49:02.030 ⇒ 00:49:06.150 Uttam Kumaran: who is taking each of those? That’s it. So otherwise the project doesn’t succeed.
379 00:49:06.760 ⇒ 00:49:10.349 Miguel de Veyra: Yeah, I think this one. I need to take a lot more responsibility.
380 00:49:10.350 ⇒ 00:49:11.509 Uttam Kumaran: It’s mostly idea one.
381 00:49:11.510 ⇒ 00:49:15.209 Miguel de Veyra: Like, I would say a good chunk of my time should be in this part.
382 00:49:16.620 ⇒ 00:49:20.810 Miguel de Veyra: instead of, like, you know, getting the agents to work. And Yada Yada.
383 00:49:20.810 ⇒ 00:49:22.200 Uttam Kumaran: Or at least that that’s what I
384 00:49:22.950 ⇒ 00:49:30.689 Uttam Kumaran: you know. That’s always been my feedback right? But I guess what I want to hear is like, why, this time you feel like it’s gonna be different, like.
385 00:49:31.050 ⇒ 00:49:36.679 Uttam Kumaran: what are some actionable steps we can actually take starting today to get you
386 00:49:36.930 ⇒ 00:49:41.360 Uttam Kumaran: to to set you up. So you you succeed and you can rip through that
387 00:49:42.460 ⇒ 00:49:44.529 Uttam Kumaran: like, what do you think needs to change?
388 00:49:50.090 ⇒ 00:49:51.880 Uttam Kumaran: Yeah. So
389 00:49:52.230 ⇒ 00:50:02.819 Uttam Kumaran: I mean, because I don’t. I think you’re more than capable of like like if you, if you read the couple of blog posts that I read, you would have been like your opinion on this would have completely changed.
390 00:50:02.930 ⇒ 00:50:06.029 Miguel de Veyra: That’s why I’m like dude. You’re 10. You’re 10 min from like.
391 00:50:07.020 ⇒ 00:50:08.349 Uttam Kumaran: Having the answer.
392 00:50:08.830 ⇒ 00:50:09.440 Miguel de Veyra: Yeah.
393 00:50:09.920 ⇒ 00:50:17.240 Uttam Kumaran: So well, we do right? Like what I guess think about, I think, what? Yeah, yeah, go ahead.
394 00:50:17.580 ⇒ 00:50:19.070 Miguel de Veyra: I think one of them.
395 00:50:19.490 ⇒ 00:50:20.040 Uttam Kumaran: Yeah, yeah.
396 00:50:20.040 ⇒ 00:50:23.210 Miguel de Veyra: The part. The things that we could definitely do is
397 00:50:24.390 ⇒ 00:50:30.819 Miguel de Veyra: instead of like, for example, when we decided to build the agents, I think we should have spent like
398 00:50:31.250 ⇒ 00:50:32.970 Miguel de Veyra: one, maybe like.
399 00:50:32.970 ⇒ 00:50:33.590 Uttam Kumaran: Interesting.
400 00:50:33.940 ⇒ 00:50:43.459 Miguel de Veyra: Say 2 to 3 days, just planning things out and what to do, what’s possible, what’s not? Instead of just going in there, basically just doing it. Whatever we think is the fastest way to do it
401 00:50:44.630 ⇒ 00:50:51.400 Miguel de Veyra: right like not just in terms of the dlt hub like that’s, you know. That was a mess on its own. But, for example, you know.
402 00:50:51.550 ⇒ 00:50:55.030 Miguel de Veyra: I think we should have like scoped it out.
403 00:50:55.030 ⇒ 00:50:56.070 Uttam Kumaran: Make a mistake.
404 00:50:56.760 ⇒ 00:51:00.149 Miguel de Veyra: And then just approach it from a very high level and then
405 00:51:00.930 ⇒ 00:51:05.779 Miguel de Veyra: dug deeper. Because when we started, the agents like, yeah, we want slack, get it from, deal
406 00:51:06.250 ⇒ 00:51:13.029 Miguel de Veyra: it from, you know the source, and then just throw it in s. 3, get it from s. 3, and then process it transform.
407 00:51:13.290 ⇒ 00:51:22.649 Miguel de Veyra: they just throw it all into context. There was nowhere in that process where, basically, hey, what technology, what should we use here? What should we, Yada Yada.
408 00:51:25.660 ⇒ 00:51:28.680 Uttam Kumaran: I mean I that’s I. I agree, like I think
409 00:51:29.560 ⇒ 00:51:34.220 Uttam Kumaran: I think this should. I think you should own this. But what what I want to know is that like.
410 00:51:34.770 ⇒ 00:51:39.800 Uttam Kumaran: and I know we’re we’re running way over. So we can continue this. But like for each of those.
411 00:51:40.370 ⇒ 00:51:44.780 Uttam Kumaran: I just need to see some change by Friday. Otherwise, like, I have to step in.
412 00:51:45.160 ⇒ 00:51:50.260 Uttam Kumaran: And the project’s gonna look a lot different like. And it’s just like, basically, I’ll probably have to. Just
413 00:51:50.630 ⇒ 00:51:53.860 Uttam Kumaran: my only options are I step in, and if I step in it’s like.
414 00:51:54.030 ⇒ 00:51:58.180 Uttam Kumaran: I don’t know. We’ll have to just like move a lot of things in a different way.
415 00:51:58.280 ⇒ 00:52:03.750 Uttam Kumaran: or we have to just put things on. Pause until I can find an end like an Eng leader
416 00:52:04.130 ⇒ 00:52:10.459 Uttam Kumaran: who can like own that sort of research process. And like pick making sure we’re using the best fundamental
417 00:52:10.800 ⇒ 00:52:14.780 Uttam Kumaran: tools, and that we’re using the best fundamental frameworks for this stuff like.
418 00:52:15.340 ⇒ 00:52:21.740 Uttam Kumaran: I will go find opportunities and tools like contextual. But like, I’ll never be like.
419 00:52:21.990 ⇒ 00:52:23.769 Uttam Kumaran: I’ll never be the one to
420 00:52:23.880 ⇒ 00:52:26.269 Uttam Kumaran: to go implement right. That’s your guys’s job.
421 00:52:26.620 ⇒ 00:52:27.470 Miguel de Veyra: Yep.
422 00:52:31.740 ⇒ 00:52:50.430 Uttam Kumaran: So maybe we continue this. I know we’re running a little bit over. But think on this. I mean, I said a lot. I don’t. It’s not really my. I don’t really want to sit here and and sort of preach that things aren’t going well, but I think we’re all in agreeance that something needs to change. I I want to see us put some plan to execute against. So we see some change by Friday.
423 00:52:50.680 ⇒ 00:52:55.520 Uttam Kumaran: and that that doesn’t just mean like, Hey, we’re gonna work really hard for the next 4 days like, that’s not what I’m interested in.
424 00:52:55.660 ⇒ 00:53:03.899 Uttam Kumaran: like, I want to know, like, hey, how are responsibilities changing? How are communications changing, and what can? What should each of us expect from each other?
425 00:53:04.300 ⇒ 00:53:09.209 Uttam Kumaran: And how can we have some sort of check in by Friday by next week.
426 00:53:09.360 ⇒ 00:53:17.820 Uttam Kumaran: and then we all agree on a point by which, hey? If this doesn’t happen, here are the next steps that need to happen right if we get to Friday or we get to next week and like
427 00:53:18.280 ⇒ 00:53:30.780 Uttam Kumaran: we’re still haven’t figured out some stuff or like there’s not been a research process. Then Xyz happens right. I think the easiest short term thing is for me to just come in and I’ll I’ll just act as engine lead, and
428 00:53:31.020 ⇒ 00:53:34.889 Uttam Kumaran: I mean, a lot of things are going to have to change, though, like stand ups will have to move
429 00:53:35.060 ⇒ 00:53:36.899 Uttam Kumaran: towards like around this time.
430 00:53:37.640 ⇒ 00:53:42.330 Uttam Kumaran: Like, I have to basically take my time away from something else.
431 00:53:42.540 ⇒ 00:53:45.220 Uttam Kumaran: But like, that’s the that’s the number one short term
432 00:53:45.390 ⇒ 00:53:59.239 Uttam Kumaran: fork in the road. The second short term fork in the road is that I just have to find someone that’s like that can take that on for me longer. Term by like within the next month. Otherwise, I don’t see this project succeeding, because we’re going to use a lot of techniques that
433 00:54:00.130 ⇒ 00:54:07.339 Uttam Kumaran: where we’ve that are that are. We’re used like year ago, that are, you know we should. We’re totally should leverage more things so.
434 00:54:07.580 ⇒ 00:54:24.450 Uttam Kumaran: And I know, I think, for amber. I think amber needs a little bit of reprieve, so she can do her job right. Sure. Job is to report back what’s going? Well, her. Her job is to destroy blockers. Anything, communication, requirements, vision. She needs to like attack and destroy those blockers.
435 00:54:25.002 ⇒ 00:54:44.980 Uttam Kumaran: And then make sure that the thing gets adopted. This is where, like in terms of the engineering. I don’t think you’re actually on the hook for adoption, although it matters. Your job is to make sure that we use the best solution given the constraints we have, and that the team is enabled, and that Amber has a great confidence in the fact that things are gonna come out when they say they’re gonna come
436 00:54:45.740 ⇒ 00:54:53.560 Uttam Kumaran: or Amber’s job is to basically be able to say, cool all of that. Now, I’m gonna package that I’m gonna destroy blockers. I’m gonna get this thing adopted.
437 00:54:53.810 ⇒ 00:54:54.770 Uttam Kumaran: and then
438 00:54:54.970 ⇒ 00:55:06.230 Uttam Kumaran: her job is to then give me the confidence to say cool. I’m gonna give you another engineer like, or I’m gonna I’m gonna I’m gonna say, hey, we have 5 more ideas in the pipeline. So here’s more budget to work on for another 6 months.
439 00:55:06.540 ⇒ 00:55:10.540 Uttam Kumaran: That’s the laddering of. That’s the domino effect, you know.
440 00:55:11.010 ⇒ 00:55:21.111 Uttam Kumaran: but it is a domino effect where it starts on the engineering side, because we’re an engineering 1st company where the engineering isn’t confident. The project manager is not confident, and then I’m a derivative.
441 00:55:21.870 ⇒ 00:55:28.130 Uttam Kumaran: But I think this is really helpful. I mean, this is one of the 1st times I think we’ve been able to see this sort of like 3 legged approach.
442 00:55:28.680 ⇒ 00:55:31.669 Uttam Kumaran: and so I hope you know, for both of you is also
443 00:55:32.400 ⇒ 00:55:38.530 Uttam Kumaran: productive and and important, you know, and it hopefully, it wasn’t too like strenuous, or
444 00:55:38.910 ⇒ 00:55:40.869 Uttam Kumaran: you know, anything like that.
445 00:55:43.060 ⇒ 00:55:43.670 Miguel de Veyra: It’s very.
446 00:55:43.670 ⇒ 00:55:49.740 Amber Lin: I would love to continue it. Do you have a few other meetings coming up? But, Utah.
447 00:55:49.930 ⇒ 00:55:58.919 Amber Lin: I’m free to meet like after the offer of the record. Kick off like I’m free free to meet. Then, if you want to continue this today.
448 00:55:59.370 ⇒ 00:56:06.469 Uttam Kumaran: Yeah, I think it’s I mean, I would love for the 3 of us to do to. I guess. I guess, Miguel, what’s your schedule like today? Or should we just plan on
449 00:56:06.900 ⇒ 00:56:09.269 Uttam Kumaran: continuing tomorrow morning, or
450 00:56:09.760 ⇒ 00:56:16.850 Uttam Kumaran: like what do you think or or I mean, if we, if we can do the rest, Async, like maybe me and Amber can take a stab at like a couple of things.
451 00:56:17.160 ⇒ 00:56:20.660 Uttam Kumaran: and then we can chat again tomorrow morning. I’m I’m down for anything.
452 00:56:21.440 ⇒ 00:56:24.040 Miguel de Veyra: Yeah, I think tomorrow morning I need to process some stuff on my end.
453 00:56:24.040 ⇒ 00:56:30.630 Uttam Kumaran: Okay, think about it. Maybe me and Amber will touch base again, and I’ll record. We’ll record the meeting and then add some notes.
454 00:56:30.810 ⇒ 00:56:34.780 Uttam Kumaran: This is like, this is the most. This is really what changes like
455 00:56:35.100 ⇒ 00:56:41.170 Uttam Kumaran: I. And I’ve said this multiple times, and I like I don’t know how you know, there’s a phrase where like, I can’t stress this enough.
456 00:56:41.360 ⇒ 00:56:43.970 Uttam Kumaran: But I really can’t not stress this enough
457 00:56:44.240 ⇒ 00:56:58.700 Uttam Kumaran: winning in this area in our company is really what’s gonna make or break us from just being an average tech consultancy that you guys have all probably seen or maybe work for. And being like a real premier
458 00:56:58.830 ⇒ 00:56:59.630 Uttam Kumaran: like
459 00:57:00.160 ⇒ 00:57:11.179 Uttam Kumaran: like leading Edge Techno AI enabled like services firm this project? Right? That’s why, like, I’m not spending like, I don’t have a conversation like this with the operations team.
460 00:57:11.360 ⇒ 00:57:17.530 Uttam Kumaran: I don’t have a conversation with the marketing team like, I don’t have a conversation like this with like a specific client team, because
461 00:57:17.780 ⇒ 00:57:22.889 Uttam Kumaran: we can. We can. We can afford that. Those things go the way they’re going.
462 00:57:23.100 ⇒ 00:57:42.510 Uttam Kumaran: and there are things that they all have to improve. Don’t get me wrong. But the reason why I’m spending time like I come to all the stand ups. I try to come to the meetings. I’m into slack. I brought you, Miguel on very early before we had anything or we’re thinking about. Casey is now growing. We’re thinking about bringing a 3.rd Their only reason is because I believe it’s possible.
463 00:57:42.830 ⇒ 00:57:56.190 Uttam Kumaran: So like this is different than like, Hey, we have a job to get done. And I’m like yelling about something that like doesn’t exist like, I’m telling you, I so confident that this is gonna be the reason we break out
464 00:57:56.380 ⇒ 00:58:04.749 Uttam Kumaran: as like a leading company in this space is this team and these initiatives. It may not be very clear from your perspective, but it’s it like
465 00:58:04.930 ⇒ 00:58:12.210 Uttam Kumaran: clear as day, like 40 40 vision for me. And I’m as soon as we nail it you’ll see like
466 00:58:12.380 ⇒ 00:58:17.570 Uttam Kumaran: I bet you I’ll I’ll buy you whatever you want to buy you a beer. Bet whatever you want on that.
467 00:58:17.800 ⇒ 00:58:25.169 Uttam Kumaran: So if if you ask me what motivates me to keep having these tough conversations, it’s it’s that realization.
468 00:58:25.626 ⇒ 00:58:39.399 Uttam Kumaran: and that realization. I’m talking to customers, partners, investors. I’m talking to advisors every single day of like a 4 or 5 of these. I explain what we’re doing. And there people are literally like, well, how can I write you a check for money?
469 00:58:39.860 ⇒ 00:58:46.959 Uttam Kumaran: It’s like, so this is what I’m saying is, we had. The vision is right. The priority is right, the execution needs to happen.
470 00:58:47.360 ⇒ 00:58:51.399 Uttam Kumaran: otherwise it’s all for nothing like we should just be a normal tech consultancy.
471 00:58:51.650 ⇒ 00:58:56.949 Uttam Kumaran: which is fine. We could do that. But like it’s no fun. And we’re gonna make a lot less money.
472 00:58:57.840 ⇒ 00:59:03.969 Uttam Kumaran: So I’ll leave you guys with that. That’s what I’m thinking about. So
473 00:59:04.100 ⇒ 00:59:12.630 Uttam Kumaran: let’s maybe amber. Let’s catch up later. Today. I’m just out about town doing sales work. So I’ll be on for a phone call and then.
474 00:59:13.400 ⇒ 00:59:14.179 Uttam Kumaran: yeah, we’ll send.
475 00:59:14.180 ⇒ 00:59:16.690 Uttam Kumaran: We’ll send notes, and then we can talk tomorrow morning. Yeah.
476 00:59:16.690 ⇒ 00:59:18.369 Amber Lin: Cool sounds, good sounds, like.
477 00:59:18.370 ⇒ 00:59:18.880 Uttam Kumaran: House.
478 00:59:18.880 ⇒ 00:59:21.338 Amber Lin: You’re you’re on a walk from like.
479 00:59:21.690 ⇒ 00:59:25.060 Uttam Kumaran: Well, I’m just outside a coffee shop. I just need to go inside at some point.
480 00:59:25.060 ⇒ 00:59:28.710 Amber Lin: Okay, okay. Sounds good. Let me know when you wanna meet.
481 00:59:29.230 ⇒ 00:59:29.830 Uttam Kumaran: Okay.
482 00:59:30.300 ⇒ 00:59:31.160 Amber Lin: We’ll call.
483 00:59:31.560 ⇒ 00:59:31.855 Uttam Kumaran: Okay.
484 00:59:32.620 ⇒ 00:59:35.380 Amber Lin: Alrighty. Thank you. All. Good meeting.
485 00:59:35.380 ⇒ 00:59:36.079 Uttam Kumaran: Thank you. Guys.
486 00:59:36.250 ⇒ 00:59:36.960 Uttam Kumaran: Bye.
487 00:59:36.960 ⇒ 00:59:37.990 Miguel de Veyra: Thanks everyone. Bye-bye.