Meeting Title: Brainforge x Airbnb Catch-up Date: 2026-02-27 Meeting participants: emily, Uttam Kumaran
WEBVTT
1 00:01:11.780 ⇒ 00:01:13.949 emily: Hey, Otam, how are you?
2 00:01:16.670 ⇒ 00:01:18.100 emily: Hey, can you hear me?
3 00:01:20.280 ⇒ 00:01:21.200 Uttam Kumaran: Hello?
4 00:01:21.580 ⇒ 00:01:24.250 emily: Hi! It’s been a while! How have you been?
5 00:01:24.250 ⇒ 00:01:25.630 Uttam Kumaran: Good, how are you?
6 00:01:25.630 ⇒ 00:01:27.300 emily: I am good!
7 00:01:27.570 ⇒ 00:01:28.759 Uttam Kumaran: How’s life?
8 00:01:29.390 ⇒ 00:01:35.609 Uttam Kumaran: Life is good, trying to stay warm here in Jersey, we’ve got tons of no’s. You’re in, like, Houston, right? In Austin.
9 00:01:35.840 ⇒ 00:01:40.259 emily: Okay. I see you got a sweater on, I was like, you guys must be warm down there.
10 00:01:40.260 ⇒ 00:01:46.320 Uttam Kumaran: No, it’s starting to get hot. It was cold in the morning, so I’ve just been sitting here still, so… yeah.
11 00:01:47.290 ⇒ 00:01:50.020 Uttam Kumaran: How’s everything… how’s everything going this year?
12 00:01:50.770 ⇒ 00:01:58.889 emily: Things are good, yeah, I think we’re off to a good year, just family-wise and work-wise, so yeah, how are things going on with you? I keep seeing your posts on LinkedIn.
13 00:01:59.410 ⇒ 00:02:01.160 emily: Exciting to see your business growing.
14 00:02:01.160 ⇒ 00:02:18.509 Uttam Kumaran: I appreciate it, yeah, I feel like we’re still continuing to grow, and Q4, I think we basically, like, tripled the business in one quarter. Wow. So we’re… we’re, like, we’re about 25 people now, and we are continuing to hire and grow, so it’s just…
15 00:02:18.720 ⇒ 00:02:27.189 Uttam Kumaran: keeps going. It’s getting better, it’s getting… I would say, on a emotional level, it’s getting better. Like, I think we are no longer, like.
16 00:02:27.840 ⇒ 00:02:42.640 Uttam Kumaran: nervous about, like, a lot of the small business stuff, like, you’re… you’re always, you know, as an owner, you’re always nervous about, like, okay, are we gonna hit payroll? Are we gonna, like… there’s all these extra risks. A lot of that, the noise is turning down, and we’re… both through hiring, through better processes…
17 00:02:42.640 ⇒ 00:02:56.929 Uttam Kumaran: And we’re still doing a ton of AI in the company, so we’re able to run really, really lean and still make a huge impact. And we’re still working with a lot of… we’re continuing to work with larger and larger companies.
18 00:02:57.070 ⇒ 00:03:13.189 Uttam Kumaran: So it’s great, and, like, we have an awesome team. Like, we just brought on someone this week to lead recruiting, actually, and staffing, so she’s taking a lot off my plate. We have one person that’s coming and leading a little bit of go-to-market and, like, marketing.
19 00:03:13.480 ⇒ 00:03:22.189 Uttam Kumaran: I’m still leading a lot of delivery, but Robert’s kind of more formally leading a lot of sales. Both of us still do a lot on the delivery side.
20 00:03:22.300 ⇒ 00:03:25.680 Uttam Kumaran: And so, like, slowly, like, delegating as fast…
21 00:03:25.850 ⇒ 00:03:28.750 Uttam Kumaran: as the company can handle it, you know, so…
22 00:03:30.010 ⇒ 00:03:38.080 emily: That’s amazing! Tell me more about, like, what… what kind of new businesses, like, what’s driving the growth, what is the demand that you’re seeing?
23 00:03:38.470 ⇒ 00:03:57.069 Uttam Kumaran: Yeah, I mean, I think one is just it’s a lot of compounding of all the things that we did well. You know, we do a lot of, like, case studies, we still continue to do a lot of, like, outbound sales, and a lot of the larger companies, the sales cycles are just, like, 6 to 9 months. So we just weren’t in business long enough to recognize a lot of that
24 00:03:57.200 ⇒ 00:03:59.719 Uttam Kumaran: Right? Because I would say 2023 was, like.
25 00:03:59.910 ⇒ 00:04:19.140 Uttam Kumaran: just, like, me and, like, a laptop. And so, I would say 2024 is when, like, a lot of the connections that we made and things we did in the world, they came to fruition, like, mid-2025, and it’s sort of, like, there’s, like, a delay. So, like, the more people we talk to, the more work we do,
26 00:04:19.470 ⇒ 00:04:33.170 Uttam Kumaran: you start to see the benefit of that on, like, a time horizon that’s, like, a little bit, like, that long, which wasn’t really clear to me when I started the business that it would take that long, but that’s just how long it takes. And so now we’re working around that’s, like, starting to compress, because
27 00:04:33.210 ⇒ 00:04:52.000 Uttam Kumaran: we do a lot of sales activities now. You know, we have several people who are full-time doing sales, marketing, you know, design. And then also, like, we’re continuing to deliver, and then we’ve just really tried to increase our pricing and go upmarket. So, we have several clients that we’re working for that are multi-hundred-million dollar brands.
28 00:04:52.150 ⇒ 00:05:10.839 Uttam Kumaran: And the moment we start working with them, we learn a lot, but also now, we have a lot more stories. And so, the amount of things that we’ve done has just really, really grown, and then our ability to take what we do, turn it into a case study or a story, go to the next person.
29 00:05:10.900 ⇒ 00:05:17.519 Uttam Kumaran: Has also not only increased how fast it takes for us to do that, but the amount of those conversations that we’re having.
30 00:05:17.650 ⇒ 00:05:21.920 Uttam Kumaran: So it’s just, the flywheel is starting to, you know, happen for,
31 00:05:22.490 ⇒ 00:05:41.330 Uttam Kumaran: for our company, and it’s really, really nice. Like, it’s taking a lot of pressure off, but it’s not like we’re slow… we slow down at all. So, we’ve… we’ve continued to emphasize how much we use AI for everything in the business, and it’s just gotten easier and faster to do that compared to 2 years ago, 1 year ago.
32 00:05:41.460 ⇒ 00:05:52.880 Uttam Kumaran: Where… and that’s sort of something that we’re a little bit religious about across the entire organization. So, not just engineering, but, like, sales, marketing, design.
33 00:05:53.320 ⇒ 00:06:02.180 Uttam Kumaran: Operations, like, everything is speeding up, and so we’re just able to punch, like, way, way, way harder and way higher, and…
34 00:06:02.370 ⇒ 00:06:13.499 Uttam Kumaran: Really, that’s just freeing us up to start to go build, like, enterprise relationships where we can go sell, you know, really, really, like, large, amazing work. And then lastly is partnerships, so…
35 00:06:13.500 ⇒ 00:06:22.470 Uttam Kumaran: We’re doing a lot of work with Snowflake, we’re doing a lot of work with Omni, which is a big VI tool, and they’re able to bring us clients, we’re able to deliver the client work for them.
36 00:06:22.530 ⇒ 00:06:26.939 Uttam Kumaran: And so, like, some of those relationships that I’ve been building over the last, like, 2 years.
37 00:06:27.110 ⇒ 00:06:36.810 Uttam Kumaran: Like, it just all kind of, like, cascades, you know, because they see that we’ve been trying to work with them, we’ve delivered them business, they like that our clients are successful.
38 00:06:37.080 ⇒ 00:06:41.000 Uttam Kumaran: And so that just, like… Moves, you know?
39 00:06:41.820 ⇒ 00:06:43.899 emily: That’s amazing, the snowball is getting bigger.
40 00:06:43.900 ⇒ 00:06:47.180 Uttam Kumaran: Yeah, yeah, yeah, so… it’s good, it’s good.
41 00:06:47.980 ⇒ 00:06:48.870 Uttam Kumaran: Yeah.
42 00:06:49.700 ⇒ 00:06:56.250 Uttam Kumaran: What do you, like, what’s how, like, are you… what’s been… what’s work been like? Like, tell me about how everything’s been going.
43 00:06:56.250 ⇒ 00:07:01.080 emily: Yeah, so, I actually returned to the corporate world full-time.
44 00:07:01.270 ⇒ 00:07:08.710 emily: Yeah, I’m a staff data scientist at Airbnb, working on a cross-functional team, so we look across, like.
45 00:07:08.950 ⇒ 00:07:33.030 emily: the entire journey from, like, making sure we have high-quality supply on the platform, to making sure we use data to help the hosts becoming better and provide a better guest experience. And then on the other side, we also, like, make sure that the guests are also high quality, you know, so that, you know, hosts… it’s a lot… you have a lot of stake when you put your home or even a rental property up, as an Airbnb, so making sure that
46 00:07:33.030 ⇒ 00:07:37.020 emily: You know, the guests are also high quality, they’re not going to trash your place.
47 00:07:37.020 ⇒ 00:07:40.200 emily: So yeah, it’s an exciting year for me, I just started.
48 00:07:40.200 ⇒ 00:07:43.139 Uttam Kumaran: At the end of last year, yeah. Yeah.
49 00:07:43.410 ⇒ 00:07:45.039 Uttam Kumaran: What’s it been like being back?
50 00:07:45.440 ⇒ 00:07:57.919 emily: It’s interesting, I would say… it’s really exciting to see how AI is, like, integrating into everything that we do, and that we have all this AI too, so it’s, like, I think it’s a great time to go back and, like, try to, like.
51 00:07:58.030 ⇒ 00:08:03.189 emily: pick up the craft again, but also in a way that is very AI augmented.
52 00:08:03.190 ⇒ 00:08:10.690 Uttam Kumaran: Yeah, yeah, yeah. That’s also what we’re seeing, is like, even across data, we’re using AI for analysis, for modeling.
53 00:08:10.910 ⇒ 00:08:20.759 Uttam Kumaran: I mean, like, for us, that’s… that’s, like, our competitive advantage in the market, because in professional services, it’s, like, the last people that want to adopt.
54 00:08:20.870 ⇒ 00:08:35.139 Uttam Kumaran: Because they bill hourly, or they’re just, like… it’s not the people in Airbnb and industry, it’s, like, consultants and contracts, so the talent is also… talent bars also can sometimes be lower, which just is, like, giving us such a boost, you know, I feel like.
55 00:08:36.450 ⇒ 00:08:40.880 emily: Yeah, I feel like it kind of frees me up to, like, do the insightful work.
56 00:08:40.880 ⇒ 00:08:44.480 Uttam Kumaran: Yes, the work you actually want to end up doing as a data person.
57 00:08:44.480 ⇒ 00:08:54.119 emily: Yeah, yeah, except when it takes care of the plumbing, and the fetching, and the crunching, and then I kind of just, like, look at it and try to, like, understand the story of the data. It actually makes the job more fun.
58 00:08:54.120 ⇒ 00:08:58.620 Uttam Kumaran: Yeah, yeah, yeah, it’s like, usually that’s, like, the last mile thing after months and months, you know?
59 00:08:58.620 ⇒ 00:09:00.699 emily: Yeah, now I, like, just do the last mile.
60 00:09:02.640 ⇒ 00:09:14.989 emily: I don’t know, I’d love to hear your thoughts on this, because then if, like, your whole team is focused on last mile thing, right? Then, as, like, a leader, you’ve got to think about, like, how do you think ahead, how do you run faster?
61 00:09:14.990 ⇒ 00:09:17.889 Uttam Kumaran: Yeah, that’s the thing, I was gonna ask you, like, do you feel like…
62 00:09:18.010 ⇒ 00:09:26.649 Uttam Kumaran: Because you’re just working closer to the last mile, like, you would… do you feel like the… two things could be possible?
63 00:09:26.790 ⇒ 00:09:42.440 Uttam Kumaran: like, or a couple things, maybe. Like, one is, like, can you get there with less people? Can you get to the… can you get to the final deliverable faster? And then maybe the third most interesting question is, like, can you actually affect the company, like, faster? Like.
64 00:09:42.520 ⇒ 00:09:50.129 Uttam Kumaran: does the fact that you get to the Insight affect Airbnb any faster? You know, so, like, I think about those… those kind of three things.
65 00:09:50.790 ⇒ 00:09:57.780 emily: Yeah, I absolutely think that on every specific task, we get to the deliverable faster.
66 00:09:57.780 ⇒ 00:09:58.540 Uttam Kumaran: Yeah.
67 00:09:58.540 ⇒ 00:10:04.519 emily: But your question’s about, do we need less people, and can we affect the company faster?
68 00:10:05.170 ⇒ 00:10:08.139 emily: I don’t know, because, like, do we need less people? What I’ve.
69 00:10:08.140 ⇒ 00:10:18.099 Uttam Kumaran: Less people is the most shallow, like, hypothesis, right? So, I think… I think it’s worth entertaining. I… I have… maybe I’ll let you say it, I have some… I have some thoughts, for sure, but…
70 00:10:18.770 ⇒ 00:10:28.500 emily: I would love to hear them, like, honestly, as a practitioner, I’m, like, worried, right? It’s like, when is AI gonna take over my job? But what I’m seeing is, like, it just kind of, like…
71 00:10:28.650 ⇒ 00:10:37.499 emily: you know, like, when you start digging under a rock, you then start digging under all those rocks, so that’s what I’m seeing. So, like, it kind of creates more demand.
72 00:10:38.230 ⇒ 00:10:38.770 Uttam Kumaran: Yes.
73 00:10:38.770 ⇒ 00:10:51.989 emily: different kind of skill set, but then I don’t know if we can actually have a bigger impact, because the company, even if you have AI doing all this stuff, you still have, like… I mean, in my corporate… maybe you guys are smaller.
74 00:10:51.990 ⇒ 00:11:09.209 Uttam Kumaran: Same thing, it’s… I think in my… in the company, I feel like I’ve reduced a lot of the… there’s… there’s no politics, like, there’s no barriers. But when we work for clients, it’s the same thing. It’s… there’s a… I call it, like, there’s business inertia, like, there’s inertia, like, pushing back.
75 00:11:09.510 ⇒ 00:11:12.829 Uttam Kumaran: And I don’t know whether, like, something that would have taken
76 00:11:13.090 ⇒ 00:11:16.950 Uttam Kumaran: Maybe something that would have taken a month, you can now do in 2 weeks.
77 00:11:17.540 ⇒ 00:11:26.510 Uttam Kumaran: But can you do it in a day at, like, an Airbnb scale? I don’t know. Like, that… there’s, like, a minimum… there’s a minimum, like…
78 00:11:26.830 ⇒ 00:11:30.880 Uttam Kumaran: you know, thing. You know, that’s how I feel.
79 00:11:31.850 ⇒ 00:11:43.559 emily: I think right now, we can’t do it in a day, but maybe, like, two models, 3 models down the line, in a couple years, we could. But then, like, you know, like, okay, if we want to change something for all the holes, right?
80 00:11:43.910 ⇒ 00:11:51.649 emily: if we can do the analysis quick, that means we can do more changes. But you can’t be pushing 500 changes to the whole.
81 00:11:51.650 ⇒ 00:11:55.670 Uttam Kumaran: So that’s the thing, there’s some, like, bar, some minimum bar, yeah.
82 00:11:55.670 ⇒ 00:11:56.700 emily: Yeah, so…
83 00:11:56.700 ⇒ 00:12:00.360 Uttam Kumaran: Can you at least hit that, like, are you guys… like, can you at least hit whatever that…
84 00:12:00.740 ⇒ 00:12:06.800 Uttam Kumaran: maximum velocity is, right? That’s how I could think about it, is like, how far are we from how fast
85 00:12:06.980 ⇒ 00:12:13.719 Uttam Kumaran: we can actually run. Like, you know, it’s okay, run a good experiment, we need X amount of events over some time horizon. Okay, maybe that is, like.
86 00:12:14.180 ⇒ 00:12:18.519 Uttam Kumaran: We can only run this amount. There’s, like, some max volume.
87 00:12:18.800 ⇒ 00:12:22.080 Uttam Kumaran: But, like, how far are we from that? It’s sort of probably the game.
88 00:12:22.080 ⇒ 00:12:28.300 emily: Yeah, I don’t know, and I’m worried when we get there. It’s kind of like, okay, if we’re running fast with AI, do we need all these people?
89 00:12:28.510 ⇒ 00:12:32.899 Uttam Kumaran: Yeah, I, I feel lucky in that,
90 00:12:34.010 ⇒ 00:12:38.180 Uttam Kumaran: I… I don’t know, it’s tough. I feel lucky that I’m in a growing business.
91 00:12:38.380 ⇒ 00:12:44.119 Uttam Kumaran: We are certainly not hiring as fast as we probably would have without AI.
92 00:12:44.400 ⇒ 00:12:46.520 Uttam Kumaran: So, do I feel bad for, like.
93 00:12:47.190 ⇒ 00:12:57.250 Uttam Kumaran: the person I didn’t hire, I’m not sure. Right? But, in a big company, it is certainly… tough, because…
94 00:12:57.940 ⇒ 00:13:03.690 Uttam Kumaran: if you… I mean, there’s two ways. One is, like, in my company, we’ve actually, like, decomposed
95 00:13:03.930 ⇒ 00:13:08.119 Uttam Kumaran: different roles, and, like, all the things that people do, and we’re actively being, like.
96 00:13:08.130 ⇒ 00:13:17.169 Uttam Kumaran: let’s go and make sure you can do these things in an AI augmented way, one by one by one by one. In a data world, right, like, can you spin up… can you, like, create
97 00:13:17.170 ⇒ 00:13:28.160 Uttam Kumaran: Snowflake warehouse? Can you do ETL? Can you create a data model? Can you… when you push a PR, can you review the before and after, how it affects the columns?
98 00:13:28.210 ⇒ 00:13:33.560 Uttam Kumaran: one by one by one, I’m like, this thing typically takes 15 to 20 hours.
99 00:13:33.950 ⇒ 00:13:47.730 Uttam Kumaran: we have someone on the team that’s like, can I build either, like, cursor tools, or can we build a GitHub action that’s sort of AI-native, or can we build some type of agentic solution that cuts that in half?
100 00:13:48.690 ⇒ 00:13:57.030 Uttam Kumaran: And then I said, go, one by one by one, we cut it in half. The thing I tell them, though, is I said, I don’t care about getting it to 100%.
101 00:13:57.200 ⇒ 00:14:05.950 Uttam Kumaran: If you can get it 50 or 80% off, then stop, right? And we’re doing that systematically, you know? So, in that sense, like.
102 00:14:06.400 ⇒ 00:14:11.580 Uttam Kumaran: if I’m doing that with no funding, no, like, Nothing.
103 00:14:11.690 ⇒ 00:14:16.969 Uttam Kumaran: I wonder what the big companies are doing, where they have teams working night and day just to, like.
104 00:14:17.220 ⇒ 00:14:19.010 Uttam Kumaran: Do these time studies.
105 00:14:20.030 ⇒ 00:14:28.389 Uttam Kumaran: And then it kind of becomes, like, okay, is your backlog big enough where actually you do what they, like, the romantic thing, which is like, oh, we can work on more things?
106 00:14:28.610 ⇒ 00:14:30.500 Uttam Kumaran: Or is there actually, like.
107 00:14:30.910 ⇒ 00:14:34.340 Uttam Kumaran: Hey, we’re actually at a pace to accomplish everything we want to accomplish, and…
108 00:14:34.620 ⇒ 00:14:39.679 Uttam Kumaran: there’s no more room. And I don’t think it… I don’t think it affects the engineers first.
109 00:14:40.050 ⇒ 00:14:45.690 Uttam Kumaran: I think it affects the support staff, design, Marketing sales first.
110 00:14:48.150 ⇒ 00:14:58.800 Uttam Kumaran: Because I think there’s still a lot of complication in the work that we do, that it’s, like, it’s… it’s like… it’s at least not at the point now where you can perfectly encompass it, but…
111 00:14:59.150 ⇒ 00:15:07.400 Uttam Kumaran: it is a require… it is getting there fast. Like, that sounds more like… it’s like a requirements issue. Like, the requirements need to be better.
112 00:15:07.670 ⇒ 00:15:12.089 Uttam Kumaran: In my… in my business, like, I feel…
113 00:15:12.720 ⇒ 00:15:17.579 Uttam Kumaran: I feel lucky because I’m going to a lot of companies where this is not even, like.
114 00:15:18.010 ⇒ 00:15:19.629 Uttam Kumaran: They’re not even close to, like.
115 00:15:19.630 ⇒ 00:15:19.990 emily: Mmm.
116 00:15:19.990 ⇒ 00:15:37.200 Uttam Kumaran: rolling out AI, like, Airbnb, Uber, you guys are, like, that’s, like, the top, right? This is… these are, like, yes, it’s, like, big companies, but they’re just trying to do the normal data work. Like, we’re still setting up dbt. So then I’m like, I live in both worlds.
117 00:15:37.240 ⇒ 00:15:42.269 Uttam Kumaran: But what I hope is that my company, because we are using it.
118 00:15:42.640 ⇒ 00:16:01.789 Uttam Kumaran: and then we’re deploying the same things, how do we make the margin? And so that’s, like, what I focus on. And then I’m like… and then we also do this for companies. We go in and we’re like, here’s how we’re using AI. We’d love to help bring that to your company. But you’ll still see, like, a lot of companies have no clue, like, or they reject it, or it’s just, like, not on their roadmap.
119 00:16:01.940 ⇒ 00:16:06.890 Uttam Kumaran: But the volume for the leadership is getting louder and louder and louder for them.
120 00:16:07.200 ⇒ 00:16:23.229 Uttam Kumaran: And I just want to be the person they call when they’re, like, thinking about how to do it, you know? Because they’re like, oh yeah, you mentioned that you guys are using a lot of AI, or your sales team is using Cursor, like, how did you figure that out? Or, like, how do you measure that? I’m like, would love to talk about that, you know?
121 00:16:23.910 ⇒ 00:16:30.380 emily: Yeah, you’re at an interesting spot, you know, go into those companies and help them use AI that’s gonna…
122 00:16:30.380 ⇒ 00:16:31.979 Uttam Kumaran: I feel, like, I…
123 00:16:32.110 ⇒ 00:16:38.620 Uttam Kumaran: I’ve… a lot of people ask me about, like, the ethical piece that you mentioned, and I just feel like our… the companies we work for
124 00:16:39.180 ⇒ 00:16:42.279 Uttam Kumaran: We… number one, we don’t advertise that we come in and, like.
125 00:16:42.480 ⇒ 00:16:46.219 Uttam Kumaran: Our… when we come in, we advertise, we want to help you grow more efficiently.
126 00:16:46.460 ⇒ 00:16:51.350 Uttam Kumaran: So it’s never like, oh, let’s come in and, like, help you fire everybody. We won’t… I won’t take that kind of work.
127 00:16:51.500 ⇒ 00:17:03.410 Uttam Kumaran: And then also, a lot of these companies are just struggling to just do the things that they need to do. Their people are overworked, so their retention’s lower, and so part of it is just, like, can we just speed up, like.
128 00:17:03.760 ⇒ 00:17:07.660 Uttam Kumaran: moving things between PDFs, like, filling it, form filling, like…
129 00:17:08.050 ⇒ 00:17:14.680 Uttam Kumaran: speed up just your understanding of data? Maybe an end-to-end data platform would have taken a year. Can we do that faster?
130 00:17:15.200 ⇒ 00:17:17.130 Uttam Kumaran: Those are the kinds of things.
131 00:17:17.130 ⇒ 00:17:17.740 emily: Yeah.
132 00:17:17.740 ⇒ 00:17:21.499 Uttam Kumaran: But it’s inter… it’s interesting, for sure.
133 00:17:22.200 ⇒ 00:17:27.049 emily: And I think sometimes we’ve got to be realistic. If your work does lead to, like.
134 00:17:27.339 ⇒ 00:17:31.570 emily: people having to reshape their job, that is fine, too.
135 00:17:31.570 ⇒ 00:17:32.400 Uttam Kumaran: Yeah, yeah.
136 00:17:32.400 ⇒ 00:17:44.999 emily: And I feel like, as, like, a leader in the industry, like, you can, like, you do a lot of, like, events, right? You can think about, like, hey, data practitioner, this is how you should change or, like, evolve in this stage. You know, if.
137 00:17:45.000 ⇒ 00:17:45.560 Uttam Kumaran: Yeah.
138 00:17:45.560 ⇒ 00:17:51.010 emily: comes to you, and, you know, job changes is part of the project, I think it’s okay to be realistic.
139 00:17:51.190 ⇒ 00:17:55.759 Uttam Kumaran: Yeah, or maybe even, like, how do you get ahead of that as a data leader or a data person?
140 00:17:55.760 ⇒ 00:17:56.530 emily: Yeah.
141 00:17:56.530 ⇒ 00:18:05.730 Uttam Kumaran: Because also, all of us just want to do our job… I mean, I would say, like, for the most part, I hope people just want to do their job, and do their job well, and get paid, and so…
142 00:18:05.980 ⇒ 00:18:14.220 Uttam Kumaran: I think it’s like, how can you come to your company before they come to you? Like, how can you come, hey, I’m thinking about how we can bring AI into our team, or like.
143 00:18:14.280 ⇒ 00:18:34.239 Uttam Kumaran: hey, there’s, like, every time we do a business review, it takes 15… it takes 15 hours to put this thing together, and, like, can we just use the slide generator? Because it’s not even about the slides, it’s, like, right? So can I just use the AI slide generator? But I’m still putting together, like, all the insights into a doc that then goes into that as a prompt.
144 00:18:35.030 ⇒ 00:18:38.219 Uttam Kumaran: You know, that’s what I… I think about. It’s,
145 00:18:38.920 ⇒ 00:18:44.879 Uttam Kumaran: I don’t know, it’s a tough problem, like… It’s, yeah, I,
146 00:18:45.930 ⇒ 00:18:57.450 Uttam Kumaran: it’s tough, like, when we go to companies, though, again, I just feel like we’re bringing them a breath of fresh air by saying, like, we can find these efficiencies for you, and, like.
147 00:18:57.860 ⇒ 00:19:03.040 Uttam Kumaran: Sometimes it’s in call centers, sometimes it’s in just, like, really, like, traditional sales staff.
148 00:19:03.230 ⇒ 00:19:10.260 Uttam Kumaran: And they’re struggling, it’s, like, so hard for them, and they don’t have the talent that we’re used to be working around.
149 00:19:10.470 ⇒ 00:19:25.719 Uttam Kumaran: And when we come in, we’re like, we have answers that not only are data-driven, but then we’re like, hey, we… we use this tool internally, like, we’d love to implement it for you guys. And so they’re very happy. So I still think that, like, in our tech bubble, the noise is really loud, but in the rest of the world.
150 00:19:26.790 ⇒ 00:19:29.420 Uttam Kumaran: not even… like, I don’t even think it’s, like…
151 00:19:29.820 ⇒ 00:19:35.419 Uttam Kumaran: people are using ChatGPT for whatever they can use it for, and that’s it, you know?
152 00:19:35.950 ⇒ 00:19:37.470 emily: That’s a good place to be, I think.
153 00:19:37.470 ⇒ 00:19:46.480 Uttam Kumaran: It’s a good place, yeah, yeah, yeah. One day, they’ll be like, oh, I wish, like, one ChatGPT could talk to another, and I’m like, yes, let me help you do that.
154 00:19:46.480 ⇒ 00:19:47.250 emily: What goes off.
155 00:19:47.250 ⇒ 00:19:49.430 Uttam Kumaran: Yeah, yeah, yeah.
156 00:19:51.160 ⇒ 00:19:52.490 Uttam Kumaran: Yeah, so…
157 00:19:52.740 ⇒ 00:19:59.280 Uttam Kumaran: Yeah, I mean, we’re… on the data side, we’re also starting to do a lot more, like, sincere strategy work.
158 00:19:59.470 ⇒ 00:20:08.830 Uttam Kumaran: And the reason I think we have an advantage is that, like, because we do a lot of the data warehousing, data modeling, we’re starting to hire more
159 00:20:08.930 ⇒ 00:20:13.620 Uttam Kumaran: Like, senior data analysts and people that are, like, able to put together, like.
160 00:20:13.900 ⇒ 00:20:23.119 Uttam Kumaran: really crisp insights with, like, the recommendation on the action, and going to one place to get that is… has been really, really powerful, because
161 00:20:23.520 ⇒ 00:20:27.490 Uttam Kumaran: we’re one team, so, like, we model it for our analysts. It’s not like a…
162 00:20:27.490 ⇒ 00:20:27.870 emily: Yeah.
163 00:20:27.870 ⇒ 00:20:33.060 Uttam Kumaran: oh, like, this column is missing, or, like, they never told me that this, like, the definition… it’s like…
164 00:20:33.210 ⇒ 00:20:36.210 Uttam Kumaran: One-stop shop, and, like, we’re verticalized.
165 00:20:36.390 ⇒ 00:20:46.719 Uttam Kumaran: And so I think both sides benefit, like, the modelers get really good requirements, and the analysts get exactly what they want, and we could do that really fast.
166 00:20:47.040 ⇒ 00:20:52.159 Uttam Kumaran: And then the client wins for that, right? And so, for me, it’s like, okay, how do I continue to just, like.
167 00:20:53.080 ⇒ 00:21:09.700 Uttam Kumaran: how do I… how do I share that that is a… that is, like, a great competitive offering of an environment versus, like, being a data scientist, or a data analyst, or a data engineer in a company? And, like, how do I continue to share that, like, the way we do… the way we do data work
168 00:21:09.970 ⇒ 00:21:28.600 Uttam Kumaran: I think it’s different than a lot of other consultancies, and, like, I think we’re actually more of a product company internally, in that there’s a lot more camaraderie and intermingling than a typical consulting company. Meaning, like, I think a typical consulting company, it’s like, oh, I’m working on this client, you, like, don’t know anything about any other clients or anything, like…
169 00:21:28.700 ⇒ 00:21:39.879 Uttam Kumaran: This is not like that. I feel like it’s almost like consulting as a service, in that internally, you know, like, what it’s like to be in a product company with Slack and all the different teams and learnings.
170 00:21:40.020 ⇒ 00:21:44.320 Uttam Kumaran: And we’re more like that, probably just because that’s my background.
171 00:21:44.720 ⇒ 00:21:52.269 Uttam Kumaran: But externally, the client just sees the service that we’re delivering. So I think that makes it fun to have, like, team members where
172 00:21:52.510 ⇒ 00:22:03.569 Uttam Kumaran: We have all the benefits of, like, what it’s like to work on a platform, like shared learnings, we do something here, someone can use it here, people actually become, like, friends, and it’s like, you know,
173 00:22:04.210 ⇒ 00:22:12.030 Uttam Kumaran: And yeah, I think that’s been… that’s been really… compared to the other, like… I worked at a couple consultancies, and I have a lot of friends in them, and I think it’s a lot different.
174 00:22:13.480 ⇒ 00:22:18.969 emily: Yeah, I work at consultancy too, it’s very different, like you say, like, I only know my client, I don’t know your client.
175 00:22:19.380 ⇒ 00:22:21.550 Uttam Kumaran: Yeah, and it’s so formal, and I’m like, this is…
176 00:22:21.850 ⇒ 00:22:25.819 Uttam Kumaran: Internally, where it’s a lot more, like, a typical tech…
177 00:22:25.950 ⇒ 00:22:31.139 Uttam Kumaran: company. Or at least we’re heading that… that direction in terms of the culture, you know?
178 00:22:31.510 ⇒ 00:22:35.190 emily: Yeah, I think you definitely give off that vibe, like, every time I talk to you.
179 00:22:36.150 ⇒ 00:22:40.619 emily: Robert, it just definitely feels like a very collegiate environment, you know?
180 00:22:40.620 ⇒ 00:22:49.609 Uttam Kumaran: Yeah, well, it’s also… it’s our… it’s because that’s who we are, and we’re data people, and so we’re not, like, MBAs or, like, private equity, so…
181 00:22:49.730 ⇒ 00:22:57.580 Uttam Kumaran: I’m… like, this is all… this is all I know, and we’re doing it, I think, in an industry that is used to such formality.
182 00:22:58.790 ⇒ 00:23:07.680 Uttam Kumaran: And it’s used to such formality because you tend to bring on people that, like, are stretched super thin, or, like.
183 00:23:07.870 ⇒ 00:23:17.469 Uttam Kumaran: Or, like, oh, there’s, like, there’s 100 people on something, and so the requirements, so we need to add so much project management to make sure it’s all formal, all buttoned up.
184 00:23:17.760 ⇒ 00:23:24.069 Uttam Kumaran: And I’m like, well, without that, I think I could deliver way faster, and I think the client is actually gonna appreciate how, like.
185 00:23:24.570 ⇒ 00:23:26.660 Uttam Kumaran: We build a relationship.
186 00:23:27.060 ⇒ 00:23:40.749 Uttam Kumaran: And I think we’re gonna… we’re moving pretty significantly away from the billable hour, and we’re making the spread, you know, because we’re trying to drive towards, hey, we’re gonna deliver an outcome. Like, this outcome of, like.
187 00:23:40.930 ⇒ 00:23:51.529 Uttam Kumaran: ultimately, even Robert talks to me about, like, how do we move our team to, say, like, we’re gonna deliver data warehouse, or to, like, we’re gonna help drive you 20%,
188 00:23:51.640 ⇒ 00:23:52.990 Uttam Kumaran: Higher click-through rate.
189 00:23:53.630 ⇒ 00:24:01.339 Uttam Kumaran: And we are a part of that, and if we deliver that, how can we get… start to get a piece? You know? Or, like, we are going to deliver you, like.
190 00:24:01.540 ⇒ 00:24:07.940 Uttam Kumaran: A 50% increase in attributable leads, because we improved, like, the attribution model.
191 00:24:08.130 ⇒ 00:24:09.539 Uttam Kumaran: And, like, what is that worth?
192 00:24:10.170 ⇒ 00:24:15.990 Uttam Kumaran: And I never thought about data that way, right? You’re always, like, just a… it’s almost like a cost… a cost center, like.
193 00:24:16.190 ⇒ 00:24:23.680 Uttam Kumaran: it’s just this much for this people. You don’t… you tend to think that, like, wow, if I make a change on something, and it drives 30% traffic, and I’m like.
194 00:24:23.900 ⇒ 00:24:37.549 Uttam Kumaran: I’m thinking so underpaid. So we’re like, okay, we are driving towards that, and if we are moving away from the billable hour, how can we do something where it’s a good deal for everybody, where
195 00:24:37.550 ⇒ 00:24:48.399 Uttam Kumaran: they give us latitude to drive towards a shared outcome of a KPI that they also want, right? And so they can stop thinking of us as, like, a tech team, IT,
196 00:24:48.430 ⇒ 00:24:50.610 Uttam Kumaran: And start thinking of us as, like, almost like a…
197 00:24:50.810 ⇒ 00:25:05.209 Uttam Kumaran: closer to a business team, and our weapon is the data work that we do, you know? It’s the end-to-end data work that we do. Like, we built the models, we landed things, we then did the insights, and then we did the recommendation, and we saw the lift.
198 00:25:05.460 ⇒ 00:25:10.089 Uttam Kumaran: And how can we then get a piece of the lift? It’s like, and some of our clients were…
199 00:25:10.640 ⇒ 00:25:13.440 Uttam Kumaran: Slowly trying to drive them towards that.
200 00:25:13.560 ⇒ 00:25:19.370 Uttam Kumaran: But it’s a new model, like, you know, it’s a new way of, I feel like, thinking. Not a lot of people get that type of deal.
201 00:25:19.650 ⇒ 00:25:30.229 emily: I love that thinking, because it aligns really well with your end-to-end model, too, right? All the way from, like, the data model, the pipelines, to the insights. I think it’s really good.
202 00:25:30.230 ⇒ 00:25:33.720 Uttam Kumaran: And it’s risky on us, I think, you know? I think we take the risk on.
203 00:25:34.830 ⇒ 00:25:37.349 emily: Yes and no, but you’re good, so you will deliver.
204 00:25:38.410 ⇒ 00:25:39.659 emily: Why are you getting the higher margins?
205 00:25:39.660 ⇒ 00:25:44.090 Uttam Kumaran: You say no because, like, you guys will… you guys will do it. I say… I’m saying the opposite, I’m like.
206 00:25:44.360 ⇒ 00:25:51.410 Uttam Kumaran: I don’t know, sometimes I’m like, there’s 101 reasons why, like, we may not hit it, but…
207 00:25:51.580 ⇒ 00:26:03.270 Uttam Kumaran: I also think that, like, I… as a… as a… for the business owners and the operators that we… I don’t think they get a deal like that. They’re never offered something like that, where they’re like, hey, we’re gonna come in and find you that 5% lift.
208 00:26:03.850 ⇒ 00:26:13.510 Uttam Kumaran: And I know you want that lift, and you have maybe other people on your team doing that, so let’s contribute to that, versus, oh, it’s $30K a month to do data platform. It’s like…
209 00:26:14.030 ⇒ 00:26:17.050 Uttam Kumaran: that doesn’t end up touching your OKRs.
210 00:26:17.400 ⇒ 00:26:25.089 Uttam Kumaran: And, I think ultimately, like, a really great data team, or maybe a data leader, like, I’m curious how they do it at Airbnb, like.
211 00:26:25.300 ⇒ 00:26:36.250 Uttam Kumaran: that’s how they should be pitching the data work, right? It’s not like, oh, we’re just here to do… like, oh, some people ask me, like, what data… I’m like, data warehouse, like, whatever. Like, it’s not about… that’s not the star.
212 00:26:36.400 ⇒ 00:26:45.750 Uttam Kumaran: that’s… and that’s my whole career, is, like, in modeling and warehousing, but I’m like, the star is the insight, and the star is if you can take that and actually make some money with it.
213 00:26:45.930 ⇒ 00:26:47.969 Uttam Kumaran: You know? And so, like…
214 00:26:48.200 ⇒ 00:26:53.029 Uttam Kumaran: Can we talk about that? Don’t worry about it, I’ll get you about doing this other stuff, you know?
215 00:26:53.630 ⇒ 00:27:00.150 emily: Yeah, tie it to a KPI, get people excited, and then you want them to ask you, what does it take to make it happen?
216 00:27:00.150 ⇒ 00:27:06.579 Uttam Kumaran: than your data warehouse. Yes, yes, exactly. How does it work at Airbnb? Are you guys given, like.
217 00:27:06.730 ⇒ 00:27:12.180 Uttam Kumaran: OKRs around, like, experiments, or impact, or, like… Yeah.
218 00:27:12.180 ⇒ 00:27:17.390 emily: Yeah, so as the data scientist, I actually helped set the OKR, so it’s like…
219 00:27:17.390 ⇒ 00:27:18.290 Uttam Kumaran: Nice.
220 00:27:18.290 ⇒ 00:27:37.439 emily: I guess the leadership comes up with kind of, like, a vision, like, things like grow supply, or, like, maybe it’s Wall Street, grow revenue, or how do we grow revenue, we have new product ideas, and then I get specifics, okay, this product is going to reduce cancellation rate by X percent, you know, so then I set it that way. Okay.
221 00:27:37.440 ⇒ 00:27:42.530 Uttam Kumaran: And then that sort of triggers to, like, okay, exactly what does product need to build? I see.
222 00:27:42.530 ⇒ 00:27:51.859 emily: I just need to code, and then to track this OKR, we want to track cancellation rate by, like, platform, region, so then that have, like, requirements for the data.
223 00:27:51.860 ⇒ 00:27:52.260 Uttam Kumaran: Yeah.
224 00:27:52.260 ⇒ 00:28:01.190 emily: So, like, to summarize, it starts at, like, a vision level, at the leadership level, and then my job is to basically translate that vision into, like, numbers.
225 00:28:01.630 ⇒ 00:28:07.740 Uttam Kumaran: Interesting, interesting. And you guys do, like, pacing and all the forecasting on those KPIs?
226 00:28:08.790 ⇒ 00:28:23.789 emily: Yes and no, definitely do pacing, but not on the KPI levels. The KPI levels tend to be very detailed, like, the goal for them is to be detailed enough so that our team can operationalize against them, but then when we do pacing, it’s more for, like, reporting, so then it doesn’t…
227 00:28:23.950 ⇒ 00:28:24.750 Uttam Kumaran: Yeah, okay.
228 00:28:24.750 ⇒ 00:28:27.569 emily: It doesn’t make sense to be that granular.
229 00:28:27.570 ⇒ 00:28:28.820 Uttam Kumaran: Makes sense, makes sense.
230 00:28:30.480 ⇒ 00:28:31.490 Uttam Kumaran: Interesting.
231 00:28:34.370 ⇒ 00:28:40.540 Uttam Kumaran: Nice, well, tell me how, like, how can I be helpful? Like, what, I’m… I just, I’m glad to even reconnect, you know?
232 00:28:40.540 ⇒ 00:28:52.129 emily: Me too, I mean, I’m gonna pose the same question to you, how can I help you? I just really like our chats, and I just want to stay in touch, you know, like, it’s very encouraging to follow your journey.
233 00:28:52.130 ⇒ 00:28:56.200 Uttam Kumaran: I appreciate it. Well, I think one ask is, like, if you have any…
234 00:28:56.360 ⇒ 00:29:00.679 Uttam Kumaran: I’ll put it even broadly, if you have any smart people in your life that you think
235 00:29:00.880 ⇒ 00:29:10.140 Uttam Kumaran: may be interested in working at Brainforge, or even just having a conversation and hearing about it, I am 100% all ears.
236 00:29:10.140 ⇒ 00:29:15.710 emily: Is there, like, a particular profile you’re looking to work with?
237 00:29:15.710 ⇒ 00:29:19.329 Uttam Kumaran: Yeah, I can send you sort of our careers, but, like, we have… we’re hiring across
238 00:29:19.790 ⇒ 00:29:23.040 Uttam Kumaran: most roles in data and in AI.
239 00:29:23.310 ⇒ 00:29:32.729 Uttam Kumaran: And, frankly, the AI piece, I’m, like, less concerned if the people have actually done hardcore AI work, but, like, it’s mostly, like, full-stack developers or back-end developers.
240 00:29:33.230 ⇒ 00:29:39.509 Uttam Kumaran: And then the data, it’s kind of across the board, like, data engineers, analytics engineers.
241 00:29:39.650 ⇒ 00:29:42.959 Uttam Kumaran: And, like, really, like, senior data analysts, you know?
242 00:29:43.520 ⇒ 00:29:55.689 Uttam Kumaran: But I’ll send you the JDs, and so you have something to link to, but again, I put it… I put it in that way, where I’m like, if you have, like, yo, yeah, there’s someone who I know I’m thinking about, like, I would love to just say hi, and tell them about what we do.
243 00:29:55.690 ⇒ 00:29:56.830 emily: I will give it some thought.
244 00:29:56.940 ⇒ 00:29:59.219 Uttam Kumaran: Okay, okay, that would be great, yeah.
245 00:29:59.220 ⇒ 00:30:00.939 emily: It’s so nice connecting, I hope…
246 00:30:00.940 ⇒ 00:30:09.869 Uttam Kumaran: Yeah, yeah, and I’m hopeful that we can do something in New York soon, so I’ll totally let you know if we do a happy hour and an event. Like, would love to say hi.
247 00:30:09.870 ⇒ 00:30:11.899 emily: Yeah, so it’ll be good to meet in person.
248 00:30:11.900 ⇒ 00:30:17.020 Uttam Kumaran: Yeah, definitely, definitely. Alright, keep in touch, take care. 100%, thank you.
249 00:30:17.020 ⇒ 00:30:17.720 emily: Aye.
250 00:30:17.720 ⇒ 00:30:18.280 Uttam Kumaran: Bye.