2025-02-25_uttamjamie_92dfc404

Meeting Title: Uttam_Jamie Date: 2025-02-25 Meeting participants: Jamie Perlmutter, Uttam Kumaran

---

WEBVTT

1 00:05:17.290 ⇒ 00:05:18.400 Jamie Perlmutter: Yo dude

2 00:05:21.440 ⇒ 00:05:22.120 Jamie Perlmutter: network.

3 00:05:22.120 ⇒ 00:05:23.186 Uttam Kumaran: My guy.

4 00:05:23.720 ⇒ 00:05:24.410 Jamie Perlmutter: What up.

5 00:05:25.397 ⇒ 00:05:27.250 Uttam Kumaran: How’s it going.

6 00:05:27.610 ⇒ 00:05:35.540 Jamie Perlmutter: I’m good man. Sorry I had a nice elongated one on one with my manager, so I apologize on that front. What’s good.

7 00:05:35.750 ⇒ 00:05:36.739 Uttam Kumaran: Long time Bro.

8 00:05:37.160 ⇒ 00:05:50.490 Jamie Perlmutter: I know, crazy like I feel like time just flies and like you don’t even realize but I know I hit like one year at shopify the other week, and I was like what the fuck like I

9 00:05:50.670 ⇒ 00:05:54.750 Jamie Perlmutter: I don’t even know like it doesn’t even feel like one year since we work, although it feels like a lifetime.

10 00:05:54.750 ⇒ 00:06:01.289 Uttam Kumaran: I know it does. Yeah, it does. Yeah, that I feel like I’m still doing the same shit. So it’s like.

11 00:06:02.820 ⇒ 00:06:04.349 Uttam Kumaran: like, partly, I’m like.

12 00:06:04.540 ⇒ 00:06:12.990 Uttam Kumaran: Oh, well, we’re still doing data model, like documentation. It’s like, you’re kind of like still doing that. But

13 00:06:13.540 ⇒ 00:06:18.430 Uttam Kumaran: yeah, making now, making more money slump for like

14 00:06:18.560 ⇒ 00:06:21.100 Uttam Kumaran: while we’re working on this business for a bit. But.

15 00:06:21.100 ⇒ 00:06:21.580 Jamie Perlmutter: Yeah.

16 00:06:21.580 ⇒ 00:06:23.779 Uttam Kumaran: Stuff is good dude. Austin is sick.

17 00:06:24.100 ⇒ 00:06:25.530 Uttam Kumaran: Great place.

18 00:06:26.340 ⇒ 00:06:34.930 Jamie Perlmutter: Yeah, that’s a good friend that I met out here. He was in Austin for like 7 years before moving out here, and I mean he’d loved it like loved it.

19 00:06:34.930 ⇒ 00:06:38.980 Uttam Kumaran: Yeah, it’s sick, like a good like work, life balance like.

20 00:06:39.800 ⇒ 00:06:45.299 Uttam Kumaran: And people here don’t do. Don’t talk about work at all. Like no care in the world about like

21 00:06:45.670 ⇒ 00:06:50.159 Uttam Kumaran: what you do for work anything, and there’s not. You can’t really be too flashy here.

22 00:06:50.370 ⇒ 00:06:52.779 Uttam Kumaran: so like it’s just nice, like

23 00:06:53.250 ⇒ 00:06:55.489 Uttam Kumaran: renting a house like a big fucking house.

24 00:06:57.280 ⇒ 00:06:58.250 Jamie Perlmutter: Amazing.

25 00:06:58.250 ⇒ 00:07:02.360 Uttam Kumaran: Yeah, I don’t know. It’s just it’s pretty chill, like still working a lot. But

26 00:07:02.750 ⇒ 00:07:04.429 Uttam Kumaran: outside of that stuff is good.

27 00:07:04.930 ⇒ 00:07:10.499 Jamie Perlmutter: Okay, nice. That’s fucking awesome. Yeah. I’ve noticed that like being in la now, too, like.

28 00:07:10.660 ⇒ 00:07:22.784 Jamie Perlmutter: like, if I was in entertainment, I think it would be a little different. But like people don’t really talk about work ever like nobody really gives shit. And so and like it’s not some like Big Dick contest of how long you worked.

29 00:07:23.060 ⇒ 00:07:23.730 Uttam Kumaran: It’s actually.

30 00:07:23.730 ⇒ 00:07:38.730 Jamie Perlmutter: Like, can you work the least? And like, go enjoy your life? So it’s like, that’s definitely been a very welcome change. Not that that impacts how much I work, but it’s still. It’s like a nice culture thing.

31 00:07:38.730 ⇒ 00:07:39.670 Uttam Kumaran: Yeah, yeah, yeah.

32 00:07:39.670 ⇒ 00:07:45.200 Jamie Perlmutter: It’s like you don’t have these bankers walking around just like bragging that they haven’t slept. And it’s like nice dude like, yeah.

33 00:07:45.200 ⇒ 00:07:45.680 Uttam Kumaran: Here, like.

34 00:07:45.680 ⇒ 00:07:46.980 Jamie Perlmutter: Really? Well.

35 00:07:46.980 ⇒ 00:07:56.529 Uttam Kumaran: Yeah, people here like are very worried about like me working a lot. And like, people are like, you really need to like, leave the house occasionally. Stuff like that.

36 00:07:57.190 ⇒ 00:08:02.680 Uttam Kumaran: and I’m like Dude. I do leave the house. This is the most I love the house in my entire life, like I feel free.

37 00:08:04.038 ⇒ 00:08:15.251 Uttam Kumaran: You guys have no clue. I’m like, not addicted to caffeine at the moment. I’m like cooking like, do you understand? Like how there’s levels.

38 00:08:15.960 ⇒ 00:08:19.140 Jamie Perlmutter: The cooking thing has been fantastic, like.

39 00:08:19.140 ⇒ 00:08:20.949 Uttam Kumaran: Yeah, I’m like, although I don’t.

40 00:08:21.070 ⇒ 00:08:23.569 Jamie Perlmutter: I don’t love cooking still, but like

41 00:08:23.970 ⇒ 00:08:28.380 Jamie Perlmutter: it, I mean, it definitely saves some money like it’s worth it. But.

42 00:08:28.380 ⇒ 00:08:33.890 Uttam Kumaran: I love it, dude. I love cooking, and it’s great like that’s like my hobby outside of this stuff. So I really.

43 00:08:34.480 ⇒ 00:08:37.039 Jamie Perlmutter: I wish I caught that bug like I,

44 00:08:37.280 ⇒ 00:08:45.370 Jamie Perlmutter: Sammy and I both don’t love cooking, which means that I end up cooking and like. I’m totally fine with it, but I don’t love it.

45 00:08:45.370 ⇒ 00:08:45.760 Uttam Kumaran: Yeah.

46 00:08:45.760 ⇒ 00:08:46.879 Jamie Perlmutter: And I’d I would.

47 00:08:46.880 ⇒ 00:08:52.259 Uttam Kumaran: Once you get good. Once you get good it gets a lot better. But it took me like 5 years of

48 00:08:52.710 ⇒ 00:08:57.590 Uttam Kumaran: like. I like food a lot. And I like eating food. I watch a lot of food stuff.

49 00:08:57.690 ⇒ 00:09:00.871 Uttam Kumaran: But like cooking, you have to sort of get the

50 00:09:01.910 ⇒ 00:09:09.790 Uttam Kumaran: you gotta like, get the touch a little bit because you can’t. Otherwise you can’t like do multiple things. And if you can’t do multiple things, and you’re cooking for 3 h.

51 00:09:09.790 ⇒ 00:09:10.295 Jamie Perlmutter: Right.

52 00:09:10.800 ⇒ 00:09:15.070 Uttam Kumaran: If you can’t cook and clean at the same time, like it’s just like you have to like.

53 00:09:15.240 ⇒ 00:09:18.567 Uttam Kumaran: And then if someone God forbid! Someone’s like talking to you.

54 00:09:19.940 ⇒ 00:09:27.109 Uttam Kumaran: Book everything up. So yeah, it’s like years of burning shit, like, I’m very candid with, like, yeah, it’s not like the easiest thing. But.

55 00:09:27.110 ⇒ 00:09:27.520 Jamie Perlmutter: No.

56 00:09:27.520 ⇒ 00:09:32.239 Uttam Kumaran: Good. You’re sort of like, damn! I could just buy like raw ass ingredients, and like.

57 00:09:32.240 ⇒ 00:09:34.480 Jamie Perlmutter: Right? And just make something. Yeah.

58 00:09:34.480 ⇒ 00:09:38.019 Uttam Kumaran: Like I can look in the fridge and be like we’re like.

59 00:09:38.180 ⇒ 00:09:41.239 Uttam Kumaran: I’ll make some shit that’s like, actually decent, not just like

60 00:09:41.500 ⇒ 00:09:46.222 Uttam Kumaran: rice and salmon like, okay, cool.

61 00:09:47.973 ⇒ 00:09:55.460 Jamie Perlmutter: A 1,000%. But but yeah, how’s work? Like, I mean, you seem busy.

62 00:09:55.580 ⇒ 00:09:59.660 Uttam Kumaran: Dude. This has been the toughest thing ever

63 00:10:00.456 ⇒ 00:10:05.099 Uttam Kumaran: but also good, like, I think we’re on the cusp of like

64 00:10:05.440 ⇒ 00:10:07.909 Uttam Kumaran: getting out of this like next phase, I mean.

65 00:10:08.210 ⇒ 00:10:10.690 Uttam Kumaran: yeah, I must have called you like last year, sometime.

66 00:10:10.690 ⇒ 00:10:11.550 Jamie Perlmutter: Yeah.

67 00:10:11.840 ⇒ 00:10:16.989 Uttam Kumaran: At that point we had like a few clients. Now we’re running like almost we’re running at 7 clients in parallel.

68 00:10:16.990 ⇒ 00:10:17.710 Jamie Perlmutter: Wow!

69 00:10:17.880 ⇒ 00:10:20.510 Uttam Kumaran: We have like. There’s like 15 people on the team.

70 00:10:20.920 ⇒ 00:10:22.720 Jamie Perlmutter: That’s crazy.

71 00:10:22.720 ⇒ 00:10:23.440 Uttam Kumaran: Yeah,

72 00:10:25.410 ⇒ 00:10:26.560 Uttam Kumaran: And

73 00:10:27.230 ⇒ 00:10:34.350 Uttam Kumaran: we went so like, it’s like, all like it’s, it’s like all an evolution. And every phase takes like a complete

74 00:10:34.947 ⇒ 00:10:39.720 Uttam Kumaran: destruction of like who you were to become like what you need next.

75 00:10:39.870 ⇒ 00:10:46.190 Uttam Kumaran: like what the business needs of you next. So it was for me. It was in the beginning. It’s just like, okay, I’m just juggling contracts.

76 00:10:46.590 ⇒ 00:10:50.330 Uttam Kumaran: Same data work. Okay? Then, I’m like overloaded. Now, I’m like.

77 00:10:50.480 ⇒ 00:10:52.900 Uttam Kumaran: and I bring someone on. But like, what can I afford?

78 00:10:53.030 ⇒ 00:10:57.030 Uttam Kumaran: Then it’s like, Yeah, we brought a couple people on, but sort of a bunch of us in parallel.

79 00:10:57.220 ⇒ 00:11:04.379 Uttam Kumaran: Then it’s like, okay. Now that I have a little bit of margin, let me try to get someone to help with the website, let me get someone to help

80 00:11:04.680 ⇒ 00:11:06.559 Uttam Kumaran: with like some OP stuff.

81 00:11:07.044 ⇒ 00:11:16.109 Uttam Kumaran: Let me like incorporate the bit. Let me do all those things that sort of have to happen. And then, like, I started just making more friends along the way, like people that

82 00:11:16.310 ⇒ 00:11:24.749 Uttam Kumaran: we’re either doing like sort of like data stuff, either on one higher part of the stack, or like lower part of the stack as me, and started working with them.

83 00:11:26.080 ⇒ 00:11:29.750 Uttam Kumaran: And then it’s sort of just like started growing and like

84 00:11:29.940 ⇒ 00:11:34.150 Uttam Kumaran: tried to just hire people as like I’ve I found really good people.

85 00:11:35.170 ⇒ 00:11:38.400 Uttam Kumaran: And then learning how to do sales is like a whole thing.

86 00:11:38.800 ⇒ 00:11:42.450 Uttam Kumaran: The marketing stuff like it, the marketing stuff, I would say.

87 00:11:42.580 ⇒ 00:11:49.610 Uttam Kumaran: it’s good. It’s everything we do is mostly just to come across like we’re very big. We’re not. We’re not doing a lot of active marketing stuff like.

88 00:11:49.610 ⇒ 00:11:50.360 Jamie Perlmutter: Yeah, yeah.

89 00:11:50.360 ⇒ 00:11:56.560 Uttam Kumaran: Meaning like people when they come on the phone with us, we want them to do a quick Google and be like, okay, this is like a legit place.

90 00:11:56.560 ⇒ 00:11:57.420 Jamie Perlmutter: Yeah.

91 00:11:57.420 ⇒ 00:12:06.789 Uttam Kumaran: But the phase we’re in now is sort of moving from me, working like most of the deals to now, we’re like, basically, I’m just hiring replacements. So

92 00:12:07.280 ⇒ 00:12:17.459 Uttam Kumaran: sort of trying to move to like more, 25%, 50 to 2525 to 50% of like my time going to actual engineering. Most of my time now is more like planning and like

93 00:12:17.760 ⇒ 00:12:19.499 Uttam Kumaran: executing some hard stuff.

94 00:12:21.920 ⇒ 00:12:27.840 Uttam Kumaran: But yeah, dude, it’s been insane. I mean, it’s it’s basically like, just build, it’s, it’s like, yeah, it’s been fucking hard. It’s like.

95 00:12:30.050 ⇒ 00:12:31.260 Jamie Perlmutter: Like, regardless.

96 00:12:32.520 ⇒ 00:12:40.510 Uttam Kumaran: It’s, but it’s not hard, because, like one day is particularly hard. It’s just like every. It’s just every day constantly right? So

97 00:12:40.650 ⇒ 00:12:49.960 Uttam Kumaran: it’s like, it’s sort of like when we’re in the data team, you have multiple stakeholders in your company. I have multiple stakeholders for for multiple companies.

98 00:12:49.960 ⇒ 00:12:50.760 Jamie Perlmutter: Companies, yeah.

99 00:12:50.760 ⇒ 00:12:55.040 Uttam Kumaran: And like, I’m 1 of the few people in the company that sees like everything.

100 00:12:55.750 ⇒ 00:12:58.170 Uttam Kumaran: So it’s just a lot. It’s like a lot of context.

101 00:12:58.170 ⇒ 00:12:59.730 Jamie Perlmutter: And it’s a service business.

102 00:12:59.730 ⇒ 00:13:02.880 Uttam Kumaran: It’s a service business. So we’re always like, yeah, totally

103 00:13:03.384 ⇒ 00:13:10.490 Uttam Kumaran: and you know, but also, like, I don’t get so stressed anymore. Like, you know, we’ve been through the ringer with like people, and so.

104 00:13:10.490 ⇒ 00:13:10.850 Jamie Perlmutter: Yeah.

105 00:13:10.850 ⇒ 00:13:16.179 Uttam Kumaran: It’s not really like dealing with executives or anything. It’s actually just more about

106 00:13:16.930 ⇒ 00:13:34.170 Uttam Kumaran: like we’re not. We don’t have like Vc funding right. So I have to wait to get the money, or at least get a guarantee the money to then go hire people, and so we’re we’re always a little bit over committed, which is kind of how this is gonna go. I don’t wanna just have people that are like sort of sitting. So we’re.

107 00:13:34.170 ⇒ 00:13:35.130 Jamie Perlmutter: Yeah.

108 00:13:35.130 ⇒ 00:13:37.400 Uttam Kumaran: For the past like 3, 4 months. We’ve just been like

109 00:13:37.660 ⇒ 00:13:42.169 Uttam Kumaran: waiting for the for like. How much work we have to do to go up to then bring on more people.

110 00:13:43.031 ⇒ 00:13:47.528 Uttam Kumaran: But it’s good. It’s a lot of the same stuff like it’s Ecom sas

111 00:13:47.960 ⇒ 00:13:55.250 Uttam Kumaran: we’re standing up Snowflake, dbt, it’s varying degrees of like of like size companies. But.

112 00:13:55.250 ⇒ 00:13:55.620 Jamie Perlmutter: Yeah.

113 00:13:55.620 ⇒ 00:13:57.970 Uttam Kumaran: So I’m like we have some like

114 00:13:58.720 ⇒ 00:14:03.650 Uttam Kumaran: on an E-com side. It’s all like 20 to 100 million dollar shopify Amazon.

115 00:14:03.650 ⇒ 00:14:04.690 Jamie Perlmutter: Sick.

116 00:14:04.690 ⇒ 00:14:14.403 Uttam Kumaran: Shit. So like big companies. But like they’re screwed because they don’t. They’re never gonna bring on like repeat data people. Typically we find people where a they

117 00:14:15.210 ⇒ 00:14:29.049 Uttam Kumaran: they’re like, Oh, shit like we’re growing. And like we’re running the business off a spreadsheet like we need to figure something out. But I don’t know where to start, or they’ve already tried to bring on like a data analyst. And they’re like, Go figure this out. Of course, like you need like a full stack team.

118 00:14:29.050 ⇒ 00:14:29.629 Uttam Kumaran: It’s okay.

119 00:14:29.630 ⇒ 00:14:34.580 Uttam Kumaran: That person gets that person gets jammed. And either that person gets fired or like

120 00:14:34.980 ⇒ 00:14:50.819 Uttam Kumaran: they basically bring us in to be like this person sucks. And we’re more like that person doesn’t suck. There’s just like a whole. There’s just a bunch of stuff that needs to happen here. The nice thing about us is like they. We’re not like staff augmentation. Where, like, I’m not like, call me, and I’ll give you like 10 developers like every.

121 00:14:50.820 ⇒ 00:14:51.210 Jamie Perlmutter: Yeah.

122 00:14:51.210 ⇒ 00:14:58.099 Uttam Kumaran: It’s like more holistic, and they rarely met someone like me or on my business partner.

123 00:14:58.825 ⇒ 00:15:09.200 Uttam Kumaran: who I sort of teamed up with more recently, they never met us who we can sort of talk at that level. And then also sort of talk about how it’s gonna get done. Typically they get one or the other. They get engineers who can’t

124 00:15:09.380 ⇒ 00:15:11.190 Uttam Kumaran: describe this stuff, or they get.

125 00:15:11.190 ⇒ 00:15:11.760 Jamie Perlmutter: Happens.

126 00:15:11.760 ⇒ 00:15:15.549 Uttam Kumaran: They get like sort of hoity toity consultants who

127 00:15:16.270 ⇒ 00:15:19.409 Uttam Kumaran: they’ll spend the month building the deck, but like nothing gets done.

128 00:15:19.600 ⇒ 00:15:22.819 Uttam Kumaran: Then they get us. And I’m like every I’m like, sort of everywhere. And I’m like.

129 00:15:23.430 ⇒ 00:15:29.259 Uttam Kumaran: just like, lay it on me. And I’m sure I’ve seen it. And that’s how it’s going like

130 00:15:29.470 ⇒ 00:15:32.070 Uttam Kumaran: we’re starting to see compound benefits of like

131 00:15:32.380 ⇒ 00:15:35.960 Uttam Kumaran: every shopify customer is basically doing the same sort of issues.

132 00:15:36.290 ⇒ 00:15:45.300 Uttam Kumaran: Every multi like storefront customers are the same issues. Everybody is like some sort of Zendesk, something, some sort of others. It’s that same stuff.

133 00:15:45.510 ⇒ 00:15:49.259 Uttam Kumaran: So every customer benefits from us like kind of seeing the worst, the worst.

134 00:15:49.660 ⇒ 00:15:50.000 Jamie Perlmutter: Yeah.

135 00:15:50.000 ⇒ 00:15:52.380 Uttam Kumaran: Making the same mistakes. It’s the same.

136 00:15:52.520 ⇒ 00:15:55.120 Uttam Kumaran: it’s the same shit has always been basically

137 00:15:55.555 ⇒ 00:16:00.139 Uttam Kumaran: it’s just that like I feel like I’ve done it like a million times. So.

138 00:16:00.140 ⇒ 00:16:00.660 Jamie Perlmutter: Yeah.

139 00:16:00.660 ⇒ 00:16:04.869 Uttam Kumaran: There. If they get in touch with us and they go forward with us, they get a lot out of it.

140 00:16:05.220 ⇒ 00:16:09.110 Uttam Kumaran: and I think we’re underpriced compared to like the market.

141 00:16:10.030 ⇒ 00:16:11.759 Uttam Kumaran: and like, we’re pretty genuine like.

142 00:16:12.110 ⇒ 00:16:13.989 Uttam Kumaran: I’m like, not really a consultant.

143 00:16:14.370 ⇒ 00:16:15.900 Jamie Perlmutter: You know, it’s not like.

144 00:16:15.900 ⇒ 00:16:19.869 Uttam Kumaran: You’re not meeting some of our business friends you’re meeting like me.

145 00:16:20.020 ⇒ 00:16:20.630 Jamie Perlmutter: Right.

146 00:16:20.630 ⇒ 00:16:31.749 Uttam Kumaran: And I’m like this is what you should do. You go do it like, or you could hire us to do it so like what one thing we do is like for new clients. We do like an audit where we come in. And we basically do like 2 weeks of like.

147 00:16:32.350 ⇒ 00:16:36.780 Uttam Kumaran: give me all your sources, give me all of your tools, give me all your dashboards.

148 00:16:37.450 ⇒ 00:16:44.520 Uttam Kumaran: And we sort of give them a plan. And then I’m sort of like, this is the plan. You could go shop. This here’s our price to do it.

149 00:16:46.060 ⇒ 00:16:51.725 Uttam Kumaran: I’m not. I have no worry. And I think a lot of consulting consultants have. They’re they’re like,

150 00:16:52.640 ⇒ 00:16:55.180 Uttam Kumaran: they’re very worried about things they don’t know.

151 00:16:55.980 ⇒ 00:17:00.070 Uttam Kumaran: either, because they’re business people. And they they’re trying to sell a technical product. And they have like.

152 00:17:00.070 ⇒ 00:17:00.440 Jamie Perlmutter: Yeah.

153 00:17:00.540 ⇒ 00:17:01.599 Uttam Kumaran: They never did it.

154 00:17:01.750 ⇒ 00:17:06.560 Uttam Kumaran: or they’re like technical people, and they’re trying to sell a business thing. And then they’ve never like

155 00:17:06.660 ⇒ 00:17:10.379 Uttam Kumaran: they never understood that that business people don’t care about Dbt.

156 00:17:10.630 ⇒ 00:17:13.280 Uttam Kumaran: They care about like dashboards and like decisions, right?

157 00:17:13.390 ⇒ 00:17:15.620 Uttam Kumaran: And like, we found, find myself

158 00:17:15.800 ⇒ 00:17:18.090 Uttam Kumaran: a little bit of both. And so

159 00:17:18.250 ⇒ 00:17:25.720 Uttam Kumaran: I it’s, it’s like, it’s working. Now, there’s like a bunch of other challenges. But yeah, it’s working.

160 00:17:26.280 ⇒ 00:17:28.209 Jamie Perlmutter: Are you gonna raise your prices soon?

161 00:17:29.230 ⇒ 00:17:34.990 Uttam Kumaran: Every time every time we fucking raise right every client, because Dude Price

162 00:17:35.130 ⇒ 00:17:38.729 Uttam Kumaran: price, what do they say? It’s like value is in the eye of the beholder. Dude.

163 00:17:38.730 ⇒ 00:17:39.500 Jamie Perlmutter: Yeah.

164 00:17:39.500 ⇒ 00:17:45.619 Uttam Kumaran: So. And then the other thing is like yesterday’s price is not today’s price. The better we get at what we do.

165 00:17:45.890 ⇒ 00:17:48.149 Uttam Kumaran: the less competition there is, and.

166 00:17:48.150 ⇒ 00:17:48.640 Jamie Perlmutter: Totally.

167 00:17:48.640 ⇒ 00:17:56.020 Uttam Kumaran: We have companies that spend a lot on us and are very low touch and companies that are very crazy, and

168 00:17:56.350 ⇒ 00:17:58.510 Uttam Kumaran: they don’t spend as much. So

169 00:17:58.660 ⇒ 00:18:04.519 Uttam Kumaran: it’s sort of value to the people we now move from, like hourly pricing to like package pricing, though.

170 00:18:04.820 ⇒ 00:18:11.929 Uttam Kumaran: so I we try not to do a lot like before we were doing a lot and having the variable pricing could cause issues. So we’re like.

171 00:18:12.700 ⇒ 00:18:15.689 Uttam Kumaran: right now, we just do 15 KA month or 25 KA month

172 00:18:16.170 ⇒ 00:18:19.060 Uttam Kumaran: for like, and we sort of various scopes.

173 00:18:19.760 ⇒ 00:18:20.130 Jamie Perlmutter: Got it.

174 00:18:20.640 ⇒ 00:18:22.855 Uttam Kumaran: So, yeah,

175 00:18:23.650 ⇒ 00:18:29.310 Uttam Kumaran: the next big thing is like, how do we get clients that are more like 30, 40, 50 KA month?

176 00:18:30.200 ⇒ 00:18:32.439 Uttam Kumaran: I think it requires. It requires like a

177 00:18:32.600 ⇒ 00:18:35.550 Uttam Kumaran: either deeper technical skill or more like

178 00:18:36.410 ⇒ 00:18:38.730 Uttam Kumaran: we’re coming in to take on even more scope.

179 00:18:38.730 ⇒ 00:18:39.340 Jamie Perlmutter: Yeah.

180 00:18:40.080 ⇒ 00:18:42.310 Uttam Kumaran: But the work is the same, the work is the same.

181 00:18:45.060 ⇒ 00:18:45.980 Jamie Perlmutter: That’s awesome.

182 00:18:46.190 ⇒ 00:18:48.508 Uttam Kumaran: Yeah dude. It’s it’s like, insane.

183 00:18:48.840 ⇒ 00:18:51.400 Jamie Perlmutter: I mean, thank God, you left flow. Code. Am I right?

184 00:18:51.400 ⇒ 00:18:54.736 Uttam Kumaran: Thank God, I left Loca. Thank God! I left Prequel

185 00:18:55.770 ⇒ 00:18:57.240 Jamie Perlmutter: Yeah.

186 00:18:57.240 ⇒ 00:19:02.129 Uttam Kumaran: I totally forgot that you were at Prequel like, that’s I, actually completely.

187 00:19:02.130 ⇒ 00:19:07.789 Uttam Kumaran: Those guys fumbled. A good thing. Yeah, fumbled a good thing. Continue to fumble 2 good things.

188 00:19:10.180 ⇒ 00:19:11.040 Jamie Perlmutter: Bye.

189 00:19:11.040 ⇒ 00:19:13.449 Uttam Kumaran: Made no friends in the process.

190 00:19:13.660 ⇒ 00:19:14.679 Jamie Perlmutter: No, and all they do.

191 00:19:14.680 ⇒ 00:19:18.159 Uttam Kumaran: What are we in it for? If we can’t make friends in the process like.

192 00:19:18.160 ⇒ 00:19:21.520 Jamie Perlmutter: Dude, and all they fucking do is speak on panels.

193 00:19:21.520 ⇒ 00:19:24.151 Uttam Kumaran: It’s over. Not fool me twice.

194 00:19:24.590 ⇒ 00:19:25.800 Jamie Perlmutter: Yeah.

195 00:19:28.430 ⇒ 00:19:31.129 Uttam Kumaran: You know, it’s almost like the X. It’s the x

196 00:19:31.565 ⇒ 00:19:35.229 Uttam Kumaran: those 2 crew that’s actually its own sort of pod. Now, like.

197 00:19:35.630 ⇒ 00:19:39.089 Uttam Kumaran: I’m friends with a lot of the folks that I’ve interacted with them. And like

198 00:19:39.360 ⇒ 00:19:46.260 Uttam Kumaran: they’ve not done people really well, they didn’t do well at all. No people. They hired a data culture. They didn’t do well at all.

199 00:19:47.130 ⇒ 00:19:50.210 Uttam Kumaran: And yeah, I don’t know. It’s their loss like that. Shit comes back.

200 00:19:50.830 ⇒ 00:19:52.809 Uttam Kumaran: That shit comes back around like

201 00:19:53.070 ⇒ 00:19:55.290 Uttam Kumaran: this. Business is not easy to run.

202 00:19:55.520 ⇒ 00:20:03.339 Uttam Kumaran: and if your business is 100 engineers, or mostly engineers, and you’re not an engineer, and you really have the disdain, for like engineering.

203 00:20:04.270 ⇒ 00:20:06.589 Uttam Kumaran: it’s not a good. It’s not like a healthy workplace.

204 00:20:06.590 ⇒ 00:20:15.049 Jamie Perlmutter: No, it’s not a good recipe. No, they’re an absolute fucking disaster, and every time I see them on Linkedin, like I just chuckle, and then move on.

205 00:20:15.390 ⇒ 00:20:17.590 Uttam Kumaran: Yeah, same, yeah.

206 00:20:17.590 ⇒ 00:20:22.250 Jamie Perlmutter: It’s crazy. But I I talk to Tilo every week these days.

207 00:20:22.250 ⇒ 00:20:25.759 Uttam Kumaran: Nice. Yeah dude. How’s work like, tell me, what’s the deal.

208 00:20:25.990 ⇒ 00:20:30.280 Jamie Perlmutter: Yeah. You know, I think it’s interesting because it’s like.

209 00:20:31.420 ⇒ 00:20:50.029 Jamie Perlmutter: still fucking. I’m technically data scientist. I do the same shit. We’ve always done like it’s no different. I’d say. The difference is, I think being at shopify has been really cool to see how the business actually operates, because I’ve never really seen anything like it like

210 00:20:50.520 ⇒ 00:20:59.119 Jamie Perlmutter: spotify moved so slow. And this like I literally think I could have not shown up to work for a year, and you’d still like the playlist.

211 00:20:59.240 ⇒ 00:21:01.130 Jamie Perlmutter: So it’s like, what am I? What the fuck am I doing.

212 00:21:01.130 ⇒ 00:21:06.390 Uttam Kumaran: Yeah, I mean, I’ve been I. The music was just. It’s like the music was, is the majority of that product like.

213 00:21:07.350 ⇒ 00:21:08.710 Jamie Perlmutter: And so.

214 00:21:08.710 ⇒ 00:21:11.250 Uttam Kumaran: The queue now a queue, and it doesn’t like fuck up.

215 00:21:11.540 ⇒ 00:21:12.069 Uttam Kumaran: So that’s.

216 00:21:12.070 ⇒ 00:21:21.580 Jamie Perlmutter: Like, I, yeah, exactly. I learned a lot there about like how data can play a role like at such a huge scale cause. Like.

217 00:21:21.860 ⇒ 00:21:30.110 Jamie Perlmutter: I think they had some of the smartest data minds I’d ever met there, like just like true, like like Phds causal.

218 00:21:30.740 ⇒ 00:21:46.680 Jamie Perlmutter: you fucking name it experimentation just like Meta analysis, all of that, like, I didn’t know anything about that stuff before. So I think it was really cool to learn. And like, I really do like one guy I work with, I really believe is like the best data scientist out there, like I like, I will die on that hill like I really do.

219 00:21:47.280 ⇒ 00:22:14.170 Jamie Perlmutter: And so I learned a ton about that. But it was just like from a personal perspective. I was like, I’m not doing anything here like they like me, but I’m not doing anything, so let me get out of here, and I was just like complaining to Tilo about it, and next thing I know he’s like you should talk to my boss. I think you’d like it here, and you know I will say, like, there’s always gonna be red tape at a big company like there’s always gonna be like some slowdowns. But this place moves faster than anything I’ve seen.

220 00:22:14.620 ⇒ 00:22:17.629 Jamie Perlmutter: Gail. Oh, yeah, cause like Toby is all over everything.

221 00:22:17.630 ⇒ 00:22:19.080 Uttam Kumaran: Like, we work fast.

222 00:22:19.660 ⇒ 00:22:25.840 Jamie Perlmutter: Not quite. We work fast because it doesn’t need to be. We work fast, but it’s like, but it’s.

223 00:22:25.980 ⇒ 00:22:31.209 Jamie Perlmutter: I think, compared to any other tech company of this size, right? That has 8 to 10,000 people.

224 00:22:31.670 ⇒ 00:22:35.740 Uttam Kumaran: Yeah, it moves at a way quicker pace. Hi.

225 00:22:35.740 ⇒ 00:22:41.480 Jamie Perlmutter: The expectation I mean, Toby is so involved.

226 00:22:41.700 ⇒ 00:23:08.799 Jamie Perlmutter: and every 6 weeks does a review of every single project that’s going on in the company, and like, keeps his finger on the pulse, and has very strong opinions. And so I think it’s like, if he sees something isn’t moving. He’s like, either do something about this or kill it, because I don’t care like we don’t need it. If if it’s not moving, we probably don’t need it, and you know that the whole, like founder mode concept, which people went absolutely ape shit about, but, like

227 00:23:09.110 ⇒ 00:23:11.049 Jamie Perlmutter: shopify, does operate with that.

228 00:23:11.050 ⇒ 00:23:13.730 Uttam Kumaran: I like that guy dude. He’s sick. I mean, he’s basically like

229 00:23:14.130 ⇒ 00:23:23.390 Uttam Kumaran: dude, because I am in the same situation where more people just it’s natural that it just. It’s just like baggage. It’s just a tax.

230 00:23:23.650 ⇒ 00:23:26.309 Uttam Kumaran: And so how do you turn it from a tax to like

231 00:23:26.610 ⇒ 00:23:32.079 Uttam Kumaran: you need net output to go up, and you need revenue employee per employee to go up.

232 00:23:32.310 ⇒ 00:23:41.230 Uttam Kumaran: But you need more people to do it, and that’s sort of the mode I’m in which I I look to a lot of those people. I’m like, what are these guys doing at that level that I can sort of adapt?

233 00:23:41.250 ⇒ 00:24:05.509 Uttam Kumaran: And also it’s like for me. I also think a lot about transparency like, if you’re proud of your goals and what you’re trying to do and your people. There’s no need to hide from the wins and be very obvious how much money we’re making, or be shy from what’s fucking up like. I talk very openly about like this is not big. This is not working. This is not working, and like this is going well. This going up. All the companies I worked at

234 00:24:05.960 ⇒ 00:24:11.010 Uttam Kumaran: transparency nightmare right? Increasingly less transparent. And

235 00:24:11.320 ⇒ 00:24:16.710 Uttam Kumaran: and we knew, I just know, because I was on data. So I knew everything but like horrible like.

236 00:24:17.140 ⇒ 00:24:33.840 Uttam Kumaran: and it’s a choice. It’s a complete choice. Because I’m running the business. And I’m I. Apart from explaining people’s salaries, everybody can know anything, any single thing they want. They can. They can get access to or no. What’s the problem with them, knowing how much money we’re making, go, what you gonna go do it yourself. Good luck!

237 00:24:33.840 ⇒ 00:24:34.510 Jamie Perlmutter: Yeah.

238 00:24:35.180 ⇒ 00:24:50.799 Uttam Kumaran: That’s good. Fucking like this is not easy, like, what are you gonna steal? How we model dim customers? Good luck like it doesn’t work that way. So instead, it’s like, just try to be open with everything, and that’s helped a lot. But that’s really cool to hear.

239 00:24:51.120 ⇒ 00:25:14.480 Jamie Perlmutter: Yeah. So it’s been really cool to see that it’s been really cool to be at a company that has such a strong engineering culture, like, you know, spotify was like a very strong data culture, like they believe that product was a science, not an art. And you can measure everything. And you can make all your decisions based off of what you measure, which is really cool, shopify sort of the opposite where it’s like.

240 00:25:15.080 ⇒ 00:25:19.749 Jamie Perlmutter: it’s sort of just like a, it’s like a we. It’s product intuition. That’s back.

241 00:25:20.310 ⇒ 00:25:25.289 Jamie Perlmutter: Yeah, right? And so. But it’s been interesting to see it, because I kind of agree with the stance of like.

242 00:25:26.250 ⇒ 00:25:29.990 Jamie Perlmutter: they like to say, we’re data informed, not data driven.

243 00:25:29.990 ⇒ 00:25:30.740 Uttam Kumaran: Yeah, yeah.

244 00:25:30.740 ⇒ 00:25:37.429 Jamie Perlmutter: Where it’s like, you’re objectively gonna know that certain things that you’re working on are important and are going to be better for the user.

245 00:25:37.430 ⇒ 00:25:39.900 Uttam Kumaran: But also do use the app like.

246 00:25:40.360 ⇒ 00:25:54.430 Uttam Kumaran: If you people use shopify. If people use spotify, they’ll be like this thing is fucking, not good and like you, they fix a lot. But it took for like a company with that much money to fix like small things like, why do you need to run like a multi band.

247 00:25:54.430 ⇒ 00:25:56.440 Jamie Perlmutter: Yeah, exactly. Exactly.

248 00:25:56.440 ⇒ 00:26:00.739 Uttam Kumaran: To like, figure out where to put the play button like, put the play button, and like move on to the

249 00:26:00.850 ⇒ 00:26:06.320 Uttam Kumaran: move on to the some other thing, right? Or put more resources into the ad stuff.

250 00:26:07.454 ⇒ 00:26:14.810 Uttam Kumaran: I feel like shopify, I mean, but dude shopify is the premier to sell online, basically like the stripe for selling online.

251 00:26:14.980 ⇒ 00:26:20.500 Uttam Kumaran: I think it’s also way better platform than Amazon in terms of the data that customers get.

252 00:26:21.140 ⇒ 00:26:28.810 Uttam Kumaran: The analytics still sucks in the platform, which is why I exist with my whole business is because you guys suck at that. So I exist.

253 00:26:29.280 ⇒ 00:26:33.380 Uttam Kumaran: Also, it’s not just that. It’s like connected to your other shit. There.

254 00:26:33.670 ⇒ 00:26:37.519 Uttam Kumaran: They’re sort of Sol. But until they hit 20 million, or

255 00:26:37.820 ⇒ 00:26:43.230 Uttam Kumaran: until they hit a lot of money, they people would just run their whole business off. Of the analytics. There.

256 00:26:43.230 ⇒ 00:26:43.820 Jamie Perlmutter: Right.

257 00:26:43.820 ⇒ 00:26:51.460 Uttam Kumaran: So it works right? It’s really when it’s like, Yo, you need a bi tool because you have, like logistics, data. You have Zendesk data. You have all this stuff.

258 00:26:51.890 ⇒ 00:26:56.050 Uttam Kumaran: Some people maybe can wake up and still go into shopify orders. But like.

259 00:26:56.740 ⇒ 00:27:00.049 Uttam Kumaran: we need like, we need to build like definitions. And this.

260 00:27:01.050 ⇒ 00:27:05.589 Jamie Perlmutter: I will say they’re working on a actually a SQL. Editor within the admin.

261 00:27:05.590 ⇒ 00:27:09.389 Uttam Kumaran: I saw like the shop sequel. Yeah. Yeah. But dude, you know.

262 00:27:09.390 ⇒ 00:27:13.239 Jamie Perlmutter: That’s not gonna replace you. I’m just saying it’s like, general thing.

263 00:27:13.240 ⇒ 00:27:18.780 Uttam Kumaran: But I I see I see what they’re trying to do is basically like, I mean dude. It’s actually a really good place to own that because

264 00:27:19.010 ⇒ 00:27:24.889 Uttam Kumaran: no, none of the vendors can own analytics at that scale like you’re not logging into like.

265 00:27:25.070 ⇒ 00:27:40.720 Uttam Kumaran: you know what I mean. So how can you expect the other? These guys are making like, 50 million of their hell employees like they want a shit ton of data. But for the small folks. Yeah, 100. I mean, ideally, those guys should be able to get away with just having one data analyst.

266 00:27:40.980 ⇒ 00:27:41.320 Jamie Perlmutter: Yeah.

267 00:27:41.320 ⇒ 00:27:42.860 Uttam Kumaran: I’m still on shopify.

268 00:27:43.100 ⇒ 00:27:48.160 Uttam Kumaran: Just get a data analyst to be able to run queries there, pull shit into, excel.

269 00:27:48.160 ⇒ 00:27:50.260 Jamie Perlmutter: Right, build your little dashboard there and then.

270 00:27:50.260 ⇒ 00:27:50.660 Uttam Kumaran: Yeah.

271 00:27:50.750 ⇒ 00:27:56.070 Jamie Perlmutter: Like, no, I totally agree. But yeah, I’d say the other thing

272 00:27:56.170 ⇒ 00:28:00.679 Jamie Perlmutter: working a shopify which has been cool is like, so I work on the

273 00:28:01.100 ⇒ 00:28:04.100 Jamie Perlmutter: I, I basically work on the app store. So like the developer

274 00:28:04.430 ⇒ 00:28:06.979 Jamie Perlmutter: developer platform. And then the app store.

275 00:28:06.980 ⇒ 00:28:08.650 Uttam Kumaran: What part of that? Like? What part.

276 00:28:08.950 ⇒ 00:28:16.240 Jamie Perlmutter: Like, literally like literally, the whole platform, all the functionality to build an app like whether it’s the cli, whether it’s

277 00:28:16.240 ⇒ 00:28:21.750 Jamie Perlmutter: it’s like what happened, whatever it is, like all of that which I knew literally nothing about.

278 00:28:21.750 ⇒ 00:28:23.723 Jamie Perlmutter: I know all about. I know all about.

279 00:28:23.970 ⇒ 00:28:34.470 Jamie Perlmutter: So it’s been, really, it’s been really cool learning about that just cause. I think, like these developer platforms are only gonna become more and more relevant. Right.

280 00:28:34.470 ⇒ 00:28:42.839 Uttam Kumaran: Because, you know, I I was like, I want to develop a brain forge app for all the shopify customers so that they just install us. They don’t have to worry.

281 00:28:42.840 ⇒ 00:28:45.370 Jamie Perlmutter: Yeah. And then the data sync happens and.

282 00:28:45.370 ⇒ 00:28:49.330 Uttam Kumaran: Yeah. And like, yeah, so I will. I think about that all the time, you know, like.

283 00:28:49.500 ⇒ 00:28:56.639 Uttam Kumaran: even Snowflake has an app ecosystem, right? So it’s great to be able to do that. And a lot of people use the shopify

284 00:28:57.540 ⇒ 00:29:00.109 Uttam Kumaran: app ecosystem to build their connections.

285 00:29:00.110 ⇒ 00:29:07.010 Jamie Perlmutter: And I still think it’s pretty untapped like it’s pretty big, but it’s pretty untapped, like. I just think a lot of the apps still suck.

286 00:29:07.400 ⇒ 00:29:09.510 Jamie Perlmutter: and so, like I have a threatened. I have a friend.

287 00:29:09.510 ⇒ 00:29:11.819 Uttam Kumaran: Because if you’re not technical, it’s hella hard.

288 00:29:11.820 ⇒ 00:29:12.560 Jamie Perlmutter: Right.

289 00:29:12.560 ⇒ 00:29:13.539 Uttam Kumaran: These are like the studio.

290 00:29:13.540 ⇒ 00:29:16.930 Jamie Perlmutter: Building one right now, and I think he’s gonna absolutely crush it. It’s like.

291 00:29:18.660 ⇒ 00:29:27.429 Jamie Perlmutter: but I still think it’s pretty untapped like, if you’re an engineer, and you’re looking for like some place to just build. Like to just get some recurring revenue like I think.

292 00:29:28.010 ⇒ 00:29:29.970 Jamie Perlmutter: easy spot to get into.

293 00:29:29.970 ⇒ 00:29:30.610 Uttam Kumaran: Yeah.

294 00:29:31.250 ⇒ 00:29:39.879 Uttam Kumaran: that’s it. That’s helpful to know. I mean dude. There’s is there some shit we’re doing for some people where we’re like running python scripts on their shopify. And that’s where I’m like.

295 00:29:40.090 ⇒ 00:29:43.219 Uttam Kumaran: I should just like App. I should just make a little app for this.

296 00:29:45.500 ⇒ 00:29:51.160 Uttam Kumaran: yeah, it’s interesting. What are you? Gonna what are you gonna do? You’re gonna stay there like, what? What do you? What’s your plan?

297 00:29:52.890 ⇒ 00:30:06.070 Jamie Perlmutter: I mean, I’d like to stay here for a bit, maybe hop around teams or something, but I’d like to stay here for a bit like the thing I’m building on the side which I haven’t nailed yet. But it’s like kind of working is.

298 00:30:06.530 ⇒ 00:30:08.910 Jamie Perlmutter: I had this idea to use

299 00:30:09.250 ⇒ 00:30:14.510 Jamie Perlmutter: basically all the AI tools to help create personalized greeting cards.

300 00:30:14.950 ⇒ 00:30:15.610 Uttam Kumaran: No.

301 00:30:15.610 ⇒ 00:30:29.159 Jamie Perlmutter: So like. I had this idea to last a year ago, when I was like trying to buy Sammy Fuck because her birthday and Valentine’s Day are back to back days. Well, makes for an expensive February, but nonetheless. I

302 00:30:29.450 ⇒ 00:30:33.860 Jamie Perlmutter: I was like in Cvs looking at these cards, and they all fucking suck.

303 00:30:33.860 ⇒ 00:30:34.560 Uttam Kumaran: Right right now.

304 00:30:34.560 ⇒ 00:30:36.679 Jamie Perlmutter: So stupid, and I was like, wait a second.

305 00:30:36.870 ⇒ 00:30:45.369 Jamie Perlmutter: I’m like, there’s all this shit out there like, how can I not just make one. So like, I’ve casually been chipping away at the idea of like, basically

306 00:30:46.160 ⇒ 00:31:06.790 Jamie Perlmutter: whether it’s using fucking openai deep seek like whatever I play around with all of them. But like using that. And then some image image generation stuff to basically where you almost like fill out a mad lib of like, I’m writing a card to my mom. It’s her 50th birthday. I’m doing blah blah blah. And then it just spits out a personalized greeting card like 4 options for that person.

307 00:31:06.790 ⇒ 00:31:07.700 Uttam Kumaran: Nice.

308 00:31:07.971 ⇒ 00:31:12.309 Jamie Perlmutter: And so it’s been cool like I haven’t. I think the thing I’m missing is like

309 00:31:13.660 ⇒ 00:31:18.180 Jamie Perlmutter: there’s something I know there’s something out there that I’m missing

310 00:31:18.490 ⇒ 00:31:28.959 Jamie Perlmutter: that will make this different than somebody being able to build this app in 2 days right? Like there’s some I know there’s something that’s not quite making it like

311 00:31:29.140 ⇒ 00:31:34.239 Jamie Perlmutter: there’s something I’m missing. And it’s like, I know. It’s just one thing that when I find it.

312 00:31:34.240 ⇒ 00:31:36.570 Uttam Kumaran: Are you gonna ship it? You’re gonna ship them the card.

313 00:31:36.900 ⇒ 00:31:37.540 Jamie Perlmutter: Yeah.

314 00:31:38.390 ⇒ 00:31:41.580 Uttam Kumaran: You should also do a virtual. You could do a virtual component.

315 00:31:41.580 ⇒ 00:31:42.060 Jamie Perlmutter: Totally.

316 00:31:42.060 ⇒ 00:31:47.520 Uttam Kumaran: Maybe you should just niche down even more to like just birthdays, or like just Valentine’s.

317 00:31:47.520 ⇒ 00:31:48.980 Jamie Perlmutter: I was, I was.

318 00:31:48.980 ⇒ 00:31:49.940 Uttam Kumaran: Do it like I mean.

319 00:31:49.940 ⇒ 00:31:50.550 Jamie Perlmutter: Down.

320 00:31:50.550 ⇒ 00:31:52.190 Uttam Kumaran: I mean mother’s day is the next.

321 00:31:52.680 ⇒ 00:32:00.930 Jamie Perlmutter: As opposed to as opposed to niching down in terms of event. What I’m actually thinking about niching down to is like the style of the card.

322 00:32:01.820 ⇒ 00:32:12.120 Jamie Perlmutter: because that’s been the hardest part for me is like some of them look too cartoony and shit, and if I could just make them like a little like extra artistic right, and like a little simpler.

323 00:32:12.120 ⇒ 00:32:12.720 Uttam Kumaran: Yeah.

324 00:32:12.990 ⇒ 00:32:26.339 Jamie Perlmutter: People will love the aesthetic, and then it will still have, like the inside joke, like I like play around with it, and like I had one like Sammy’s. Mom loves her dog so like I like put in like it’s like one of the options is like, is there an inside joke or any detail.

325 00:32:26.340 ⇒ 00:32:26.980 Jamie Perlmutter: Yeah, yeah.

326 00:32:26.980 ⇒ 00:32:50.480 Jamie Perlmutter: Loves her, Shih Tzu. And then it’s just like you know her and her Shih Tzu on the front page, and whatever like it looks great. And so I think it’s actually the style I need to strip down. But I have to figure out like I keep playing around with the prompting, because the image generation stuff is like 90% of the way there and then I would say, the caption stuff, I would say, is more like 70% of the way there, because I think the captions still fucking suck like they’re still.

327 00:32:50.990 ⇒ 00:32:52.949 Jamie Perlmutter: I’m way too hard. So like, I just.

328 00:32:52.950 ⇒ 00:32:55.349 Uttam Kumaran: You should try Grok like the twitter one.

329 00:32:55.730 ⇒ 00:32:56.960 Jamie Perlmutter: I haven’t tried Grok yet.

330 00:32:56.960 ⇒ 00:32:59.230 Uttam Kumaran: People said that one is sort of like.

331 00:32:59.820 ⇒ 00:33:02.290 Uttam Kumaran: it’s like more normal. It’s more human.

332 00:33:02.590 ⇒ 00:33:04.530 Jamie Perlmutter: They just don’t have an Api yet.

333 00:33:04.970 ⇒ 00:33:06.570 Uttam Kumaran: Yeah, you’re right. Yeah. I don’t think they.

334 00:33:06.570 ⇒ 00:33:11.009 Jamie Perlmutter: So that’s the only problem is like I’ve been using Openai. And then, like a little.

335 00:33:11.010 ⇒ 00:33:21.079 Uttam Kumaran: How about I send you? Well, dude, you should try. I’ll send you this. There’s this thing called Twain AI. We were using it a lot, because it’s really good at writing.

336 00:33:21.380 ⇒ 00:33:22.560 Jamie Perlmutter: Oh, interesting!

337 00:33:22.560 ⇒ 00:33:24.890 Uttam Kumaran: Breaks down the writing sort of like

338 00:33:25.300 ⇒ 00:33:38.709 Uttam Kumaran: process really? Well, because I was having the same thing where I would put up like a bunch of stuff and have it. Write it for me. But instead, you, this one it actually like asks you for, like specific features of like tone.

339 00:33:38.850 ⇒ 00:33:39.670 Jamie Perlmutter: Oh, cool!

340 00:33:39.670 ⇒ 00:33:41.192 Uttam Kumaran: All this shit

341 00:33:41.940 ⇒ 00:33:43.040 Jamie Perlmutter: And.

342 00:33:43.600 ⇒ 00:33:46.940 Uttam Kumaran: I feel like you should maybe try a trial or something, because.

343 00:33:46.940 ⇒ 00:33:47.330 Jamie Perlmutter: Yeah.

344 00:33:47.330 ⇒ 00:33:55.969 Uttam Kumaran: This one I would felt the writing out of this one was the best out of everything, and you can have sections. You can give it tone and like stuff like that.

345 00:33:56.110 ⇒ 00:33:58.030 Uttam Kumaran: I don’t know. I haven’t seen a better one than

346 00:33:58.530 ⇒ 00:34:00.850 Uttam Kumaran: like this is like a super prompt.

347 00:34:01.600 ⇒ 00:34:02.100 Jamie Perlmutter: No, that.

348 00:34:02.100 ⇒ 00:34:05.360 Uttam Kumaran: Verticalize like a super writing prompt for this shit.

349 00:34:05.360 ⇒ 00:34:13.010 Jamie Perlmutter: Yeah, no, that’s cool. Cause I, yeah, cause like, I have like these all these different structures. For like.

350 00:34:13.435 ⇒ 00:34:13.860 Uttam Kumaran: Yeah.

351 00:34:13.860 ⇒ 00:34:19.419 Jamie Perlmutter: And like I keep playing around with them. But I will say I started using lovable the other day.

352 00:34:19.429 ⇒ 00:34:21.879 Uttam Kumaran: 2. Both is our both is our client.

353 00:34:23.159 ⇒ 00:34:24.849 Jamie Perlmutter: Lovable, is amazing.

354 00:34:24.850 ⇒ 00:34:28.059 Uttam Kumaran: Do you know, Bolt, it’s like they’re they’re one of our clients.

355 00:34:28.060 ⇒ 00:34:29.040 Jamie Perlmutter: Oh, no way.

356 00:34:29.040 ⇒ 00:34:31.010 Uttam Kumaran: Yeah, which is dude.

357 00:34:31.370 ⇒ 00:34:39.609 Uttam Kumaran: They are making the most amount of money, and every chart is like going up like this.

358 00:34:40.770 ⇒ 00:34:42.300 Jamie Perlmutter: It’s Dick!

359 00:34:42.699 ⇒ 00:34:43.369 Uttam Kumaran: Yeah. The.

360 00:34:43.370 ⇒ 00:35:11.000 Jamie Perlmutter: The way that I like, because I’ve just been using cursor, really, and like it got me enough of the way. But like the front end sucked, and I was like, and every time I would try and instruct like try and prompt cursor, it would fuck up the whole thing, and it would just break everything. And I was like, Okay, like, this is really driving me insane because I can’t develop. I don’t know Javascript like I don’t know fucking. I know what I need, but I can’t write a line of Javascript like I don’t know. The 1st thing so like

361 00:35:11.000 ⇒ 00:35:15.440 Jamie Perlmutter: lovable is so much better at the front end. Stuff that it’s like been game changing.

362 00:35:15.440 ⇒ 00:35:18.740 Uttam Kumaran: And what is it output for you, though it gives you all the code.

363 00:35:19.110 ⇒ 00:35:22.139 Jamie Perlmutter: Oh, yeah, you link it directly to Github.

364 00:35:22.140 ⇒ 00:35:23.270 Uttam Kumaran: Oh, it was sick!

365 00:35:23.270 ⇒ 00:35:30.789 Jamie Perlmutter: Yeah. So it just like automatically pushes stuff. You can either create a new branch or I just put it on main cause. I’m the only one working on this. I don’t really give fuck and.

366 00:35:30.790 ⇒ 00:35:33.069 Uttam Kumaran: And you can do pretty complicated apps like

367 00:35:33.430 ⇒ 00:35:35.940 Uttam Kumaran: like we could. I want a database for something.

368 00:35:35.940 ⇒ 00:35:39.719 Jamie Perlmutter: Totally. So they have an integration with, what’s it called super base?

369 00:35:40.090 ⇒ 00:35:44.259 Jamie Perlmutter: Yeah. So they will automatically spin up these schemas for you and.

370 00:35:44.260 ⇒ 00:35:44.580 Uttam Kumaran: Thanks.

371 00:35:44.580 ⇒ 00:35:46.039 Jamie Perlmutter: That data and like.

372 00:35:46.040 ⇒ 00:35:49.890 Uttam Kumaran: If you’re like, if your project gets super big, are you like, okay, like.

373 00:35:50.100 ⇒ 00:35:53.560 Jamie Perlmutter: When my project gets super big, then I want to hire people.

374 00:35:53.560 ⇒ 00:35:54.430 Uttam Kumaran: Yeah, okay.

375 00:35:54.430 ⇒ 00:36:00.160 Jamie Perlmutter: Do it. But I need to get from 0 to one. And I’m not a good enough developer to actually build that myself.

376 00:36:00.160 ⇒ 00:36:00.809 Jamie Perlmutter: Yeah, yeah.

377 00:36:00.810 ⇒ 00:36:06.330 Jamie Perlmutter: this is enough to get me there, because I can basically have everything hosted. I can have data stored. I can

378 00:36:07.080 ⇒ 00:36:10.569 Jamie Perlmutter: all these different functions set up like like you name it.

379 00:36:10.570 ⇒ 00:36:15.250 Uttam Kumaran: But did you know you needed some of those things like, or you’re you just like you give it like a

380 00:36:15.560 ⇒ 00:36:18.910 Uttam Kumaran: I’m trying to do this. It’s like you probably need this thing and probably need this thing.

381 00:36:18.910 ⇒ 00:36:19.970 Jamie Perlmutter: It, it prompts.

382 00:36:20.182 ⇒ 00:36:23.799 Uttam Kumaran: Don’t know all this. I don’t know all like the back end shit that what may need.

383 00:36:23.800 ⇒ 00:36:39.000 Jamie Perlmutter: So I’ve definitely learned more from just being on the team. I am at shopify like that’s definitely like, naturally giving me more knowledge. But it does prompt you where you’re like, hey? You could probably use a database to store the result. And then, like, it’s a quicker fetching process, and whatever. And you’re like, Oh, okay.

384 00:36:39.480 ⇒ 00:36:42.219 Jamie Perlmutter: so like, I wouldn’t know the 1st thing about it. But

385 00:36:42.960 ⇒ 00:36:48.152 Jamie Perlmutter: like, if nothing else, this is just giving me like a good way to explore all the shit that’s out there.

386 00:36:48.400 ⇒ 00:36:49.040 Uttam Kumaran: Yeah.

387 00:36:49.240 ⇒ 00:36:55.530 Jamie Perlmutter: But I have this like weird conviction that this is gonna work. I just actually have to get it to a place where I’m happy enough with it.

388 00:36:55.530 ⇒ 00:37:00.870 Uttam Kumaran: Yeah. But dude, that’s even like enough to just know that you can do a thing end to end like solo.

389 00:37:01.100 ⇒ 00:37:04.860 Jamie Perlmutter: Right? Right? And like I can, I can be

390 00:37:04.980 ⇒ 00:37:09.950 Jamie Perlmutter: a founder who maybe isn’t technically an engineer, but is actually technical enough.

391 00:37:09.950 ⇒ 00:37:11.699 Uttam Kumaran: Understand what’s going on. Yeah.

392 00:37:12.575 ⇒ 00:37:13.450 Jamie Perlmutter: So

393 00:37:14.020 ⇒ 00:37:20.869 Jamie Perlmutter: so no, we’ll see. But like I just keep chipping away at it. The problem with the project like that, though, is, it’s like, not like I can just like put in an hour.

394 00:37:21.050 ⇒ 00:37:21.880 Uttam Kumaran: Yeah, yeah, yeah.

395 00:37:21.880 ⇒ 00:37:30.540 Jamie Perlmutter: You know what I mean, like you need like 3 h of like real time to dive in, and it can be hard to make that when like you’ve got a full time job and like

396 00:37:30.540 ⇒ 00:37:31.620 Jamie Perlmutter: thanks for you know.

397 00:37:31.620 ⇒ 00:37:34.716 Uttam Kumaran: Yeah, yeah, I know, yeah.

398 00:37:35.513 ⇒ 00:37:42.779 Jamie Perlmutter: But but yeah, I’ve loved leaving. It was time to leave New York, and I’ve been so much happier here.

399 00:37:42.780 ⇒ 00:37:43.780 Uttam Kumaran: Really good.

400 00:37:44.000 ⇒ 00:37:50.349 Jamie Perlmutter: It’s great, and I mean I’ve never lived in good weather in my life. Dude, it’s fucking crazy how anyone chooses to.

401 00:37:50.350 ⇒ 00:38:02.420 Uttam Kumaran: I know dude, you know people, but people lie about that, and I think it’s I don’t know why, because I’m like it’s so like life is way nicer when it’s always sunny outside. And it’s like.

402 00:38:03.190 ⇒ 00:38:06.350 Jamie Perlmutter: Everything about it’s better every single thing about it, and like.

403 00:38:06.350 ⇒ 00:38:11.271 Uttam Kumaran: Like you could always sit outside and eat like what’s what a joy.

404 00:38:11.650 ⇒ 00:38:14.429 Jamie Perlmutter: Crazy like I think about it. And I’m like.

405 00:38:15.110 ⇒ 00:38:27.910 Jamie Perlmutter: I’m like, Okay, you know, it’s February 25, th like I played golf last week like like. I don’t know like people here like I would way rather my weekend revolve around the day than the night, and like that.

406 00:38:27.910 ⇒ 00:38:28.250 Uttam Kumaran: Yeah.

407 00:38:28.250 ⇒ 00:38:40.090 Jamie Perlmutter: Difference is like at night. We’re often just fucking, chilling, but during the day is when we’re trying to do stuff, whether it’s go to the beach. We’re in a volleyball league, like whatever it is and like, it’s such a better fucking lifestyle.

408 00:38:40.090 ⇒ 00:38:41.200 Uttam Kumaran: Yeah, yeah.

409 00:38:41.200 ⇒ 00:38:45.450 Jamie Perlmutter: Dark bar like, I’m know?

410 00:38:45.450 ⇒ 00:38:46.250 Jamie Perlmutter: Like, yeah.

411 00:38:46.250 ⇒ 00:38:53.709 Uttam Kumaran: Yeah, I feel like here. It’s sort of both of those and people here. It’s a lot slower. And Austin is like, it’s a city, but like

412 00:38:53.830 ⇒ 00:38:56.799 Uttam Kumaran: there’s also like a bunch of stuff, and it’s pretty calm, like

413 00:38:56.910 ⇒ 00:39:02.060 Uttam Kumaran: it’s not much traffic. You sort of. I sort of know. I’ve been to all the restaurants, and like.

414 00:39:02.060 ⇒ 00:39:02.750 Jamie Perlmutter: Yeah.

415 00:39:02.750 ⇒ 00:39:12.510 Uttam Kumaran: Nowhere. If we want a certain vibe, I’m like we should go here. We should go here, and it’s great. It’s like very predictable, like we’ll still end up going. I still travel back to New York myself.

416 00:39:13.191 ⇒ 00:39:15.539 Uttam Kumaran: And I’ll be in La, actually like April 6.th

417 00:39:15.540 ⇒ 00:39:17.419 Jamie Perlmutter: Oh, yeah, what’s bringing you to? La.

418 00:39:17.710 ⇒ 00:39:20.610 Uttam Kumaran: We just have, like 3 people down there

419 00:39:20.760 ⇒ 00:39:22.879 Uttam Kumaran: that are that are working. So we’re just

420 00:39:23.160 ⇒ 00:39:26.929 Uttam Kumaran: gonna go say, Hi, and chill. So I’ll let you know when we’re there.

421 00:39:26.930 ⇒ 00:39:30.540 Jamie Perlmutter: Yeah, please let me know. Cause I think I’m I should be around that week. So okay, cool.

422 00:39:30.540 ⇒ 00:39:31.160 Jamie Perlmutter: awesome.

423 00:39:31.780 ⇒ 00:39:32.600 Uttam Kumaran: Yeah. Dude.

424 00:39:33.130 ⇒ 00:39:35.609 Uttam Kumaran: Yeah. Dude, you should come work for us. Bro.

425 00:39:36.670 ⇒ 00:39:39.089 Uttam Kumaran: I have to find like 300 grand for you. But

426 00:39:40.120 ⇒ 00:39:42.430 Uttam Kumaran: can I find the 300 you should come work for us.

427 00:39:42.430 ⇒ 00:39:47.560 Jamie Perlmutter: If I can, if you.

428 00:39:47.560 ⇒ 00:39:49.785 Uttam Kumaran: That’s a reasonable number. Dude.

429 00:39:50.230 ⇒ 00:39:56.400 Jamie Perlmutter: If you can get if you can get me with like real upside 200 plus in Beta.

430 00:39:56.718 ⇒ 00:40:03.400 Uttam Kumaran: It’s not like, hey? The only thing I’ll copy it. I’ll give. It’s like the actual work with like actual Ceos.

431 00:40:03.400 ⇒ 00:40:03.890 Jamie Perlmutter: Yeah, yeah.

432 00:40:03.890 ⇒ 00:40:04.959 Uttam Kumaran: Makes it fun.

433 00:40:05.220 ⇒ 00:40:08.909 Jamie Perlmutter: Right. That’s what I mean that that’s why I’m saying I can take a pay cut to come work.

434 00:40:08.910 ⇒ 00:40:12.585 Uttam Kumaran: Really, dude. Don’t! Don’t sell yourself short.

435 00:40:13.110 ⇒ 00:40:15.169 Jamie Perlmutter: I’m not saying I will. I’m saying I can.

436 00:40:16.185 ⇒ 00:40:16.860 Jamie Perlmutter: But.

437 00:40:17.340 ⇒ 00:40:19.740 Uttam Kumaran: That’d be my dream. Dude.

438 00:40:19.740 ⇒ 00:40:25.000 Jamie Perlmutter: Seriously, if you like. If you actually get to the point where you’re like you, you would need.

439 00:40:25.000 ⇒ 00:40:34.559 Uttam Kumaran: So I want to bring on the best people we know and like, you’re one of them. And the the our work is gonna get harder in 2 directions. It’s gonna become

440 00:40:34.680 ⇒ 00:40:38.499 Uttam Kumaran: more strict, more strategy. Like dude. We’re going to companies that are like.

441 00:40:38.680 ⇒ 00:40:43.950 Uttam Kumaran: tell us, like what to do. It’s turning into strategy, which is like.

442 00:40:44.780 ⇒ 00:40:59.559 Uttam Kumaran: Oh, we need to run pricing tests. Oh, we should go talk to your distributors about lowering cost of goods like it’s becoming like beyond data. And then it’s also may I? I think? Part partly, it may go more like machine learning data, science.

443 00:40:59.560 ⇒ 00:41:00.070 Jamie Perlmutter: Yeah.

444 00:41:00.070 ⇒ 00:41:13.419 Uttam Kumaran: Because the work we’re doing now is sort of it’s we’ve gone fuller stack, right? We do data engineering. We also do event analytics and analysis and dashboarding. But I think it’s just gonna go like this. And then our scope is gonna go wider. So that’s what

445 00:41:13.800 ⇒ 00:41:34.389 Uttam Kumaran: I’m thinking and dude the companies we’re working with are all incredibly cool, like, again, like really fast growing. B, 2 B Sas, or large E-com, and these guys are like operating the company every day, relying on us for all the data stuff that the challenge and the data stuff is actually not that bad. It’s really just like

446 00:41:35.953 ⇒ 00:41:48.560 Uttam Kumaran: it’s it’s really just the challenge of working directly with clients. It’s fun because they we have a seat at the table. At every single conversation. We know everything about these companies. We’ve now done so much work in Ecom that like.

447 00:41:48.970 ⇒ 00:41:51.230 Uttam Kumaran: it’s like, we just learned so much.

448 00:41:51.410 ⇒ 00:41:58.369 Uttam Kumaran: And yeah, I basically what I’m maybe I’ll send you this because I want you to read it where I’m trying to work on like incentive structure for like the team.

449 00:41:58.670 ⇒ 00:42:06.959 Uttam Kumaran: And basically we’re trying. I’m trying to build kind of like what most consultancies or data teams don’t do, which is pay people more when they work harder.

450 00:42:07.446 ⇒ 00:42:19.910 Uttam Kumaran: What a what a novel concept. And I basically want to try to say, like, if you’re able to, people are able to work on multiple clients or able to handle. Like most the escalations.

451 00:42:20.050 ⇒ 00:42:22.600 Uttam Kumaran: and like maybe based on the complexity of the client.

452 00:42:22.870 ⇒ 00:42:40.870 Uttam Kumaran: people will get paid not only more just for that if we renew. Or if we upsell like, I’m just gonna pay people more, the more they take on. And we’re gonna do like company profit share or something. And like, I actually want less people and more to overpay. And naturally we will.

453 00:42:41.360 ⇒ 00:42:50.329 Uttam Kumaran: I was talking to. I was talking to Weinberger about this. I was like, we’re gonna underpay. He’s like, no, you’re not underpaying. You’re just like he. He said. I was like, yeah. But but basically like, I want to pay them less.

454 00:42:50.440 ⇒ 00:43:08.940 Uttam Kumaran: And basically they’re going to be like, I should. I can make more somewhere else. And they can leave. And then we’re just left with people that make a shit ton. And you know, it’s not like one x or 2 X, it’s like, the difference is like between a okay data person and someone who’s really good is like 5 x. If I can get that, just pay 1.5 or 2 x, then that’s.

455 00:43:08.940 ⇒ 00:43:13.379 Jamie Perlmutter: Well also, and hiring is hiring so expensive, like hiring.

456 00:43:13.380 ⇒ 00:43:21.625 Uttam Kumaran: Kind of hard. And so that’s that’s a big problem for us is like we hired a lot hired some overseas. Now we’re moving a little bit more us and

457 00:43:22.320 ⇒ 00:43:24.220 Uttam Kumaran: but I want that. Yeah.

458 00:43:24.220 ⇒ 00:43:26.829 Jamie Perlmutter: I’ve heard good things about people in Uruguay actually.

459 00:43:27.110 ⇒ 00:43:29.113 Uttam Kumaran: Yeah, dude. I was just in Argentina.

460 00:43:29.400 ⇒ 00:43:30.450 Jamie Perlmutter: Really oh, sick!

461 00:43:30.450 ⇒ 00:43:46.639 Uttam Kumaran: In Argentina. I went to Uruguay. Yeah. And Argentina people are good in Uruguay. Also, some people are good. The tough thing is just like for client facing stuff. It’s hard. Yeah. It’s just like it’s not a link. It’s not as much of a language thing as more culture. And those guys. They work. They’re just slow, like, culturally.

462 00:43:46.840 ⇒ 00:43:48.149 Uttam Kumaran: So it’s tough like.

463 00:43:48.150 ⇒ 00:43:48.870 Jamie Perlmutter: Yeah. Totally.

464 00:43:48.870 ⇒ 00:43:49.880 Uttam Kumaran: Don’t think that’s tough.

465 00:43:51.430 ⇒ 00:43:55.029 Uttam Kumaran: But yeah, I mean, like I I don’t know. I don’t know when we’ll be at that

466 00:43:55.220 ⇒ 00:44:02.210 Uttam Kumaran: level. But I again at some point. I just want to get the best people pay like the Max.

467 00:44:02.360 ⇒ 00:44:05.519 Uttam Kumaran: and then have it just be more of like a meritocracy of like.

468 00:44:05.820 ⇒ 00:44:11.710 Uttam Kumaran: if you take on more, we and then throw the hardest problems right? Like, I want to go after stuff that’s beyond data.

469 00:44:12.080 ⇒ 00:44:15.440 Uttam Kumaran: That’s a company or or like other bigger stuff.

470 00:44:15.900 ⇒ 00:44:19.852 Uttam Kumaran: and like, I don’t know, I feel like we’re equipped to. We’re equipped to do that.

471 00:44:20.580 ⇒ 00:44:30.320 Jamie Perlmutter: I want to read. I want to read what you’re putting together. But at 1st thought it almost sounds like you’re trying to do 2 things at once, which is level people, and then reward them like.

472 00:44:30.320 ⇒ 00:44:37.000 Uttam Kumaran: The leveling. Yeah, the leveling. I don’t know. It’s tough, because I actually like.

473 00:44:37.160 ⇒ 00:44:39.470 Uttam Kumaran: I don’t really care about roles.

474 00:44:39.880 ⇒ 00:44:53.360 Jamie Perlmutter: No, but it’s more. But it’s more to your point, like when you were saying like, if they take on more projects, and whatever there’s a balance there of like incentivizing people to do that, while also making sure that only people who are equipped to do that are.

475 00:44:53.360 ⇒ 00:45:01.699 Uttam Kumaran: That’s so- so that’s definitely true is like, there will be. Yeah, you’re right, and that in order to take on additional one, the the 2 you’re working on need to be like.

476 00:45:01.700 ⇒ 00:45:02.219 Jamie Perlmutter: Right. They have.

477 00:45:02.872 ⇒ 00:45:05.480 Uttam Kumaran: Capacity and quality checks.

478 00:45:05.610 ⇒ 00:45:11.130 Uttam Kumaran: So we’ll do things like, look at the amount of escalations. Look at like if the client is happy through feedback.

479 00:45:11.130 ⇒ 00:45:11.760 Jamie Perlmutter: Right.

480 00:45:12.160 ⇒ 00:45:16.719 Uttam Kumaran: But again, dude, I’m on 7 clients one way or another.

481 00:45:17.390 ⇒ 00:45:23.160 Uttam Kumaran: so if people can come in, if someone could come in and do that fuck. I’ll give them all the fucking money.

482 00:45:23.160 ⇒ 00:45:23.830 Jamie Perlmutter: Right. But I’m saying.

483 00:45:23.830 ⇒ 00:45:24.210 Uttam Kumaran: All right.

484 00:45:24.420 ⇒ 00:45:30.249 Jamie Perlmutter: You can have like a less formal process. But you also don’t want the person who can’t handle 7 clients.

485 00:45:30.250 ⇒ 00:45:31.139 Uttam Kumaran: Sure, sure, sure, sure.

486 00:45:31.140 ⇒ 00:45:32.240 Jamie Perlmutter: Sign up for 7.

487 00:45:32.240 ⇒ 00:45:37.219 Uttam Kumaran: Exactly so. The other thing we do, which again, not many people do is we only do contract to hire meaning.

488 00:45:37.410 ⇒ 00:45:39.509 Uttam Kumaran: Come, people come in part time.

489 00:45:39.800 ⇒ 00:45:57.300 Uttam Kumaran: So we both suss each other out. I don’t mind them sussing me out. It’s not like we figure out. And then we sort of move around like, Okay, cool. Why don’t you take this client on? Why don’t you seem to be handling that? Why don’t we move to 2 clients?

490 00:45:57.700 ⇒ 00:46:01.199 Uttam Kumaran: And I want those people to be able to make more, and

491 00:46:01.510 ⇒ 00:46:05.250 Uttam Kumaran: I don’t feel greedy about like that, like I actually want people to make.

492 00:46:05.250 ⇒ 00:46:05.800 Jamie Perlmutter: Totally.

493 00:46:05.800 ⇒ 00:46:07.939 Uttam Kumaran: More. I want people to wake up and be like

494 00:46:08.300 ⇒ 00:46:15.739 Uttam Kumaran: fuck. I’m getting making money here like it’s going well, I want to work on problems. The worst thing is, you start to get people build resentment

495 00:46:15.840 ⇒ 00:46:24.740 Uttam Kumaran: and shit like that, and you sort of want with the bankers have, which is a golden handcuffs. Right? It’s like that’s the tough part, and you don’t want to that extreme. But like.

496 00:46:24.740 ⇒ 00:46:25.130 Jamie Perlmutter: Yeah.

497 00:46:25.130 ⇒ 00:46:27.199 Uttam Kumaran: Feel like that? Is the psychology a little bit right?

498 00:46:27.200 ⇒ 00:46:29.179 Jamie Perlmutter: No, and I look you gotta

499 00:46:29.950 ⇒ 00:46:34.900 Jamie Perlmutter: anyone who’s had a good business or a good company will tell you that you have to pay your best people

500 00:46:35.530 ⇒ 00:46:36.049 Jamie Perlmutter: you just have.

501 00:46:36.050 ⇒ 00:46:37.800 Uttam Kumaran: I have to, because I want to.

502 00:46:37.800 ⇒ 00:46:47.580 Jamie Perlmutter: The other side of it, too, is like, if they’re not worth like Google Meta, the reason they pay all these people that much is because it’s not even how much value they’re adding. It’s how much value they could add somewhere else.

503 00:46:47.580 ⇒ 00:46:48.480 Uttam Kumaran: Yeah, yeah, sure.

504 00:46:48.480 ⇒ 00:47:01.069 Jamie Perlmutter: Competing with them right? And so that’s the other piece is like, Okay, somebody got the consulting bug, but they’re not getting what they need out of you. Well, they’re just gonna go to someone else. They’re going to be someone that’s very talented at someone else who’s doing the same thing.

505 00:47:01.070 ⇒ 00:47:11.040 Uttam Kumaran: Don’t want this to be a revolving door, like I want people to come here and be like Yo, we’re gonna blast off this company, and if, like I sell, if the company had sold that, everybody’s gonna make a shit ton of money.

506 00:47:11.040 ⇒ 00:47:11.580 Jamie Perlmutter: Yeah.

507 00:47:11.580 ⇒ 00:47:15.300 Uttam Kumaran: Like. That’s what I want. That’s why I actively tell people. And I’m actively like.

508 00:47:15.480 ⇒ 00:47:32.120 Uttam Kumaran: that’s what we’re doing here. Cause we know. I mean. Even with data culture, it’s a complete revolving door and hiring and firing people is not only emotionally, extremely exhausting, you’re just losing, you’re just you’re losing so much value out the door when you can’t repurpose or figure it out with somebody.

509 00:47:32.290 ⇒ 00:47:36.139 Uttam Kumaran: so it’s like higher, slow fire, extremely fast.

510 00:47:36.140 ⇒ 00:47:36.570 Jamie Perlmutter: Hmm.

511 00:47:36.570 ⇒ 00:47:38.169 Uttam Kumaran: Then people just sort of like.

512 00:47:38.520 ⇒ 00:47:42.619 Uttam Kumaran: Do that wishy washy? And then they like build a culture of like strangers

513 00:47:42.750 ⇒ 00:47:54.100 Uttam Kumaran: like and dude. We we meet as an analyst group we meet as an Ae group. I meet with everybody so like client pod and horizontally like we’re running it like a tech company internally

514 00:47:54.560 ⇒ 00:48:07.060 Uttam Kumaran: right? Because, like, we’re again, we’re just a data team. So everybody needs to meet for a given stakeholder. And then I want to talk about dbt, to all the aes, I want to talk about like some tableau shit to the top people.

515 00:48:07.240 ⇒ 00:48:16.369 Uttam Kumaran: and then we want to learn right like I want us to be. If you need Ecom data model, there should be no one on planet Earth faster than us of doing that right.

516 00:48:17.440 ⇒ 00:48:20.609 Jamie Perlmutter: Yeah, I it is funny. I will say.

517 00:48:21.290 ⇒ 00:48:31.370 Jamie Perlmutter: I think our data people are like solid at shopify, like, I think they’re pretty damn technical, which I’ve noticed like more so than other places. However.

518 00:48:31.820 ⇒ 00:48:35.310 Jamie Perlmutter: our Ae. Team, like we have a Central Ae team.

519 00:48:36.170 ⇒ 00:48:51.290 Jamie Perlmutter: they are so fucking incompetent that, like I actually, I cannot even put it into words. The amount of red tape that goes around shipping one fucking data model. I actually, I lose my mind every single time I have to do it like I I can’t even.

520 00:48:51.290 ⇒ 00:48:53.180 Uttam Kumaran: You can’t go in there and like, do anything.

521 00:48:53.770 ⇒ 00:49:06.419 Jamie Perlmutter: You can go in there. But you need all these layers of approvals. The Ci checks are fucking insane like it’s like they’re because the old data setup was so bad and proliferated, so much.

522 00:49:06.550 ⇒ 00:49:07.269 Uttam Kumaran: Yeah, yeah.

523 00:49:07.270 ⇒ 00:49:35.910 Jamie Perlmutter: That they’re so. They have so much scar tissue from it that they’re trying to guard against all of that. But they’ve gone way too far in the other way, and they’re too far on that end of the spectrum. So like shipping a Pr is like the hardest thing in the world to do, because they will have feedback about the most useless shit that you could possibly imagine. Like, Hey, like we typically like to say, is, instead of has as the prefix to this field. It’s like, all right. Is that going to change whether the data is right.

524 00:49:35.910 ⇒ 00:49:36.460 Uttam Kumaran: Yeah, yeah.

525 00:49:36.460 ⇒ 00:49:40.670 Jamie Perlmutter: Like, are we gonna do that like? It’s just like. And it’s like, if you want to do that fine.

526 00:49:40.670 ⇒ 00:49:43.179 Uttam Kumaran: Why don’t they just change it? Yeah. Why don’t they go ahead and just.

527 00:49:43.180 ⇒ 00:49:46.170 Jamie Perlmutter: You go, do it, put it in the linter, and have it change it for.

528 00:49:46.170 ⇒ 00:49:47.040 Uttam Kumaran: Yeah, yeah.

529 00:49:47.110 ⇒ 00:49:54.100 Jamie Perlmutter: Fuck, but like, why are you telling me that like I can’t even now to backfill you have to

530 00:49:54.360 ⇒ 00:50:18.750 Jamie Perlmutter: create a Pr with a backfill dag for your specific dag that has to get approved, to be able to backfill something. Why are we doing? What’s the risk? If it’s running in production already. That means it works so like, what are we guarding against now? The whole thing I like literally wrote a document the other day like complaining about everything that’s wrong with them. And I was like, Okay, I can’t share this yet, because, like

531 00:50:19.470 ⇒ 00:50:34.280 Jamie Perlmutter: like such an asshole right now. But it’s like it’s such a problem. But but they’re very focused on output here. And it’s not Netflix full Netflix culture, but it’s enough where it’s like they do let go, people who are local.

532 00:50:34.280 ⇒ 00:50:35.790 Uttam Kumaran: But what are they measuring.

533 00:50:37.040 ⇒ 00:50:38.952 Jamie Perlmutter: So we do.

534 00:50:40.410 ⇒ 00:50:46.610 Jamie Perlmutter: We do like our format for performance reviews is called mastery.

535 00:50:46.930 ⇒ 00:51:01.460 Jamie Perlmutter: and the idea is that essentially you get some fucking arbitrary score. No one really knows what the score means, but you’re on a distribution against everyone who’s on your level, and so that helps you kind of like gauge where you’re at. And basically your score can increase

536 00:51:01.630 ⇒ 00:51:03.780 Jamie Perlmutter: without you getting a new title.

537 00:51:03.910 ⇒ 00:51:17.930 Jamie Perlmutter: But you can still get paid more. So it’s really interesting framework where they’re trying to reward people. I sees more where it’s like, Hey, like, you’re a staff level engineer like we can’t really promote you. But like you’re doing great work, here’s more money. And so they.

538 00:51:17.930 ⇒ 00:51:20.770 Jamie Perlmutter: the score is like something esoteric.

539 00:51:21.010 ⇒ 00:51:32.030 Jamie Perlmutter: Yeah, it’s like, it’s, I really think the score is based on how much you’re being paid. I’m like, pretty sure it’s like a complete one to one relationship there. But I think.

540 00:51:32.030 ⇒ 00:51:35.777 Uttam Kumaran: You get hires, but like Oh, but then, isn’t that that’s like

541 00:51:36.570 ⇒ 00:52:05.270 Jamie Perlmutter: It’s backwards because you want a lower score, because then you’re going to get bumped up more and get more money. Yeah, it’s all weird. But it’s an interesting system that they set up of basically like, How well are you doing in your role? And what is your overall impact been? And those are the 2 variables essentially and based on the response in that. That’s how your score changes. But it’s an interesting concept, because it rewards people who maybe can’t get promoted. Maybe there’s no path for promotion right now. But hey, like we’re still going to give you more money this.

542 00:52:05.750 ⇒ 00:52:06.230 Uttam Kumaran: Yeah.

543 00:52:06.670 ⇒ 00:52:21.799 Uttam Kumaran: That’s why, even for me, dude, I’m not. I don’t want there to be in in our data teams because we’re so client focus. I’m actually gonna get rid of the notion of like titles. Basically, I’m planning on doing this because I read a thing about Palantir and how everybody was just called forward deployed engineer.

544 00:52:22.050 ⇒ 00:52:27.710 Uttam Kumaran: And I was like, fuck. Yeah, that’s sick because they basically had special. They had. They had like specializations.

545 00:52:27.710 ⇒ 00:52:28.220 Jamie Perlmutter: Yeah.

546 00:52:28.220 ⇒ 00:52:35.450 Uttam Kumaran: But like, if you can build the dashboard, go build the fucking dashboard. Why are you gonna throw it over the fence right? Because I think about us. It’s like

547 00:52:35.650 ⇒ 00:52:37.819 Uttam Kumaran: dude. I would. I’ll work the whole stack.

548 00:52:38.000 ⇒ 00:52:51.040 Uttam Kumaran: And so that’s partly of why I want to like reward people more who can handle that. And I’ll do that. Try to do that a couple of different ways. But I’m also basically like, gonna just drop titles and be like, you’re a data engineer.

549 00:52:51.990 ⇒ 00:52:55.909 Uttam Kumaran: or like your client, your client. You’re a client engineer.

550 00:52:56.930 ⇒ 00:53:05.319 Uttam Kumaran: I’m sort of the only person like a platform side, but it will probably start to have a couple of people like we’re helping all of our engineers work on stuff. But

551 00:53:05.550 ⇒ 00:53:11.019 Uttam Kumaran: and then just sort of be like that, because then it doesn’t matter what you are. It’s like you can specialize in Dbt.

552 00:53:11.190 ⇒ 00:53:14.370 Uttam Kumaran: but then you should totally be able to push the fucking dashboard like.

553 00:53:14.630 ⇒ 00:53:19.219 Jamie Perlmutter: Well, that’s what that’s what we had here, actually, at shopify.

554 00:53:19.810 ⇒ 00:53:30.269 Jamie Perlmutter: There was no such thing as a data engineer, a analytics engineer, everyone was a data scientist, and every engineer that they hire is technically just a developer.

555 00:53:30.420 ⇒ 00:53:47.720 Jamie Perlmutter: And they did that for a long, long time, because they’re like you should be full stack, and everything like we should be able to move you from. Sure you might be better at back end, but we should be able to move you to a front end project and you’d be able to go do that. And so you’re all just developers. And then everyone in data is just a data scientist. Now.

556 00:53:47.920 ⇒ 00:54:00.719 Jamie Perlmutter: we’re too big now for that to work. Right? Like, we do actually need specialization specifically from the data side, like, you need people who are data engineer like who are better at data engineering to be data engineers. And you need people who are better.

557 00:54:00.720 ⇒ 00:54:03.160 Uttam Kumaran: Definitely, definitely. But see for us the work has.

558 00:54:03.160 ⇒ 00:54:06.939 Jamie Perlmutter: That’s what I mean. Like you’re not. You’re not 10,000 employees, so like it’s not.

559 00:54:06.940 ⇒ 00:54:11.310 Uttam Kumaran: We’re we’re. We’re almost like restarting every month with new people. So

560 00:54:11.510 ⇒ 00:54:14.360 Uttam Kumaran: it’s not like the work is gonna end up being like

561 00:54:14.520 ⇒ 00:54:18.280 Uttam Kumaran: managing like a shit ton of things. It’s gonna hit a ceiling.

562 00:54:18.540 ⇒ 00:54:19.680 Uttam Kumaran: And

563 00:54:20.230 ⇒ 00:54:29.730 Uttam Kumaran: yeah, I’m sort of thinking. But also, yeah, I’m gonna do it. I’ll start looking at Prs. We’ll start looking at slacks like we’ll use AI, you know, when I start to use AI for so much shit. It’s sick.

564 00:54:29.980 ⇒ 00:54:39.119 Uttam Kumaran: So I’m gonna start pushing everybody’s slack messages into AI. Their Github pull request into AI, and like I’ll get a good sense of like who’s good.

565 00:54:39.400 ⇒ 00:54:39.900 Jamie Perlmutter: Yeah.

566 00:54:39.900 ⇒ 00:54:48.049 Uttam Kumaran: Like we’ll of course have like. Probably we’ll end up having some notion of tickets, and when it was assigned. But, like you kind of need multiple factors.

567 00:54:48.420 ⇒ 00:55:01.430 Uttam Kumaran: And then I’ll probably just have AI create a best score like, how do I tell me how to score some shit like this where you have these, and I’ll probably do what you told me, which is, put everyone on a like a

568 00:55:01.840 ⇒ 00:55:05.350 Uttam Kumaran: don’t have fix made to do by percentile or like.

569 00:55:05.350 ⇒ 00:55:05.700 Jamie Perlmutter: Yeah.

570 00:55:05.700 ⇒ 00:55:07.300 Uttam Kumaran: Everyone against each other.

571 00:55:07.780 ⇒ 00:55:12.920 Uttam Kumaran: and then maybe it’s that plus like, if a client succeeds is how you sort of get paid and everybody

572 00:55:12.920 ⇒ 00:55:14.949 Uttam Kumaran: right? Exactly. You’re always.

573 00:55:15.710 ⇒ 00:55:31.009 Jamie Perlmutter: You are always gonna need that somewhat qualitative input right? You’ll have your quantitative inputs in terms of like speed to X like ticket close rates like whatever whatever. But then you’ll be like, okay, like, did they actually do a good job with the client like.

574 00:55:31.010 ⇒ 00:55:31.840 Uttam Kumaran: Exactly.

575 00:55:31.840 ⇒ 00:55:35.020 Jamie Perlmutter: What on a scale? One to 5. What did they do?

576 00:55:35.020 ⇒ 00:55:35.510 Uttam Kumaran: Yeah, yeah.

577 00:55:35.510 ⇒ 00:55:41.940 Jamie Perlmutter: And and then from there, once you have those inputs. Yeah, then you’ll be able to like, put something together. And it’s like, Hey, like

578 00:55:42.080 ⇒ 00:55:46.879 Jamie Perlmutter: we call it master. You can call whatever, but, like your score went up. We’re paying you X amount more.

579 00:55:46.880 ⇒ 00:55:51.010 Uttam Kumaran: Yeah, and I’ll do I. The thing is for the people that know they’re good

580 00:55:51.330 ⇒ 00:55:53.439 Uttam Kumaran: come in and dominate right like.

581 00:55:53.440 ⇒ 00:55:54.100 Jamie Perlmutter: Right.

582 00:55:54.300 ⇒ 00:55:56.009 Uttam Kumaran: For the people that know they’re bad.

583 00:55:56.140 ⇒ 00:56:02.189 Uttam Kumaran: Yeah, we’re gonna find out the hard way. So don’t take the job or like fuck around, find out like.

584 00:56:02.190 ⇒ 00:56:04.090 Jamie Perlmutter: Right exactly, or get better.

585 00:56:04.090 ⇒ 00:56:13.909 Uttam Kumaran: Or get better, you know. Oh, my God, please God, that was only the that was like I didn’t even think about. I didn’t even think about that.

586 00:56:14.780 ⇒ 00:56:16.617 Jamie Perlmutter: Oh, that’s great!

587 00:56:17.340 ⇒ 00:56:21.340 Jamie Perlmutter: But, man, it sounds like sounds like you’re doing really well. I’m happy to hear it.

588 00:56:21.340 ⇒ 00:56:25.539 Uttam Kumaran: I appreciate it. Dude. Yeah, I’m excited to see you soon, and we should all link up sometime.

589 00:56:25.540 ⇒ 00:56:26.981 Jamie Perlmutter: Yeah, that would be awesome.

590 00:56:27.270 ⇒ 00:56:28.370 Uttam Kumaran: And la too.

591 00:56:28.710 ⇒ 00:56:31.530 Jamie Perlmutter: Yeah, is your is your girlfriend gonna come to la also.

592 00:56:31.530 ⇒ 00:56:34.383 Uttam Kumaran: I’m gonna ask her if she wants to come

593 00:56:34.700 ⇒ 00:56:35.960 Jamie Perlmutter: And I’ll do something.

594 00:56:35.960 ⇒ 00:56:37.629 Uttam Kumaran: Yeah, I’ll definitely let you know.

595 00:56:38.110 ⇒ 00:56:38.899 Jamie Perlmutter: Be a lot of fun.

596 00:56:38.900 ⇒ 00:56:41.519 Uttam Kumaran: That’d be dope. Yeah. And then dude. Come to Austin sometime.

597 00:56:41.640 ⇒ 00:56:45.850 Jamie Perlmutter: I know. Let me know. If you’re in New York I’ll be there ever so often. So

598 00:56:45.850 ⇒ 00:56:53.140 Jamie Perlmutter: I I know I’m gonna be in New York the 1st week of March or April. I mean not March, April, April.

599 00:56:53.440 ⇒ 00:56:54.360 Uttam Kumaran: Okay.

600 00:56:55.527 ⇒ 00:56:59.569 Jamie Perlmutter: I’m in New York. Yeah. March 30th to like April 5.th

601 00:56:59.570 ⇒ 00:57:02.349 Uttam Kumaran: It’s like tech week that week, so I may be there.

602 00:57:02.610 ⇒ 00:57:03.530 Jamie Perlmutter: Oh, nice!

603 00:57:03.530 ⇒ 00:57:04.200 Uttam Kumaran: Yeah.

604 00:57:04.770 ⇒ 00:57:05.270 Jamie Perlmutter: Sick.

605 00:57:05.270 ⇒ 00:57:09.830 Uttam Kumaran: I do. I have to start doing the rounds. Bro, wow!

606 00:57:10.660 ⇒ 00:57:11.185 Uttam Kumaran: Just.

607 00:57:11.710 ⇒ 00:57:13.729 Jamie Perlmutter: I see you commenting on Linkedin.

608 00:57:13.730 ⇒ 00:57:18.530 Uttam Kumaran: It is what it is, dude. It’s the problem is, it’s like, that’s the only way.

609 00:57:18.530 ⇒ 00:57:19.739 Jamie Perlmutter: Works. That’s what sucks.

610 00:57:19.740 ⇒ 00:57:23.639 Uttam Kumaran: And it works, and I just try to do it my way.

611 00:57:23.640 ⇒ 00:57:24.230 Jamie Perlmutter: Yeah.

612 00:57:24.230 ⇒ 00:57:27.830 Uttam Kumaran: And not be like pushing random. Shit.

613 00:57:27.830 ⇒ 00:57:29.120 Jamie Perlmutter: I mean. I’m glad you’re not.

614 00:57:29.420 ⇒ 00:57:31.229 Jamie Perlmutter: You’re not Weinberg, but.

615 00:57:31.230 ⇒ 00:57:38.320 Uttam Kumaran: Yeah. Weinberg, Weinberg Post. I like some of the stuff he post. It’s just a lot, and it’s kind of defense. It’s kind of a little bit

616 00:57:38.760 ⇒ 00:57:40.410 Uttam Kumaran: causes issues. But

617 00:57:41.570 ⇒ 00:57:49.849 Uttam Kumaran: yeah, I don’t know. Ultimately, like I’m gonna my goal is to get off Linkedin so as fast as I can hand my account off as a marketing analyst. I will, and like

618 00:57:50.710 ⇒ 00:57:56.438 Uttam Kumaran: they’ll run the show and saw your post. I like. I haven’t run that town 2 years, so.

619 00:58:00.565 ⇒ 00:58:04.175 Jamie Perlmutter: But hi man! Well,

620 00:58:04.770 ⇒ 00:58:08.079 Uttam Kumaran: Good John, let’s catch up again soon, dude. Let me know if I can help with anything or.

621 00:58:08.080 ⇒ 00:58:12.860 Jamie Perlmutter: Yeah 100%. And same here, man, like, if you’ve got any of this stuff that you want to bounce off me like whether

622 00:58:13.660 ⇒ 00:58:15.380 Jamie Perlmutter: or whatever like just start fucking.

623 00:58:15.380 ⇒ 00:58:22.499 Uttam Kumaran: I’ll send you some of this stuff because we’re trying to put in motion. I mean, it’s again, how do we manage people like us, how do people like us

624 00:58:23.020 ⇒ 00:58:25.519 Uttam Kumaran: like feel motivated on a data team?

625 00:58:25.520 ⇒ 00:58:25.990 Jamie Perlmutter: Right.

626 00:58:25.990 ⇒ 00:58:27.299 Uttam Kumaran: Really like what I’m working on.

627 00:58:27.300 ⇒ 00:58:28.649 Jamie Perlmutter: Feel rewarded for.

628 00:58:28.650 ⇒ 00:58:37.559 Uttam Kumaran: Yeah, that’s sort of all I’m working on and trying to do and trying to build this team up. So I would love. I’ll send it to you. And I just want to like exactly like, just feedback on, like

629 00:58:38.490 ⇒ 00:58:54.299 Uttam Kumaran: how to think about it. And also, it’s like, I think, about what it’s like to be an engineer at the company. And so I want to do things. I want it. And again, I’m gonna get feedback from everyone. The team before putting in place, because then nobody can be like I didn’t see that coming. But also again, if you’re for all Star, I want people to get paid more, and

630 00:58:54.440 ⇒ 00:58:56.690 Uttam Kumaran: I would love to send it to get your feedback. So.

631 00:58:56.690 ⇒ 00:58:58.900 Jamie Perlmutter: Yeah, would love that. I think that’d be great.

632 00:58:59.970 ⇒ 00:59:01.010 Uttam Kumaran: Selected.

633 00:59:01.010 ⇒ 00:59:06.879 Jamie Perlmutter: Sweet man alright. Well, good luck out there! Let’s we won’t let it be a year again.

634 00:59:06.880 ⇒ 00:59:09.430 Uttam Kumaran: Yes, no, let’s catch up. I’ll call. I’ll call you again soon.

635 00:59:09.430 ⇒ 00:59:11.030 Uttam Kumaran: Okay, good, alright sweet.

636 00:59:11.300 ⇒ 00:59:13.310 Uttam Kumaran: We didn’t talk sports. We didn’t talk.

637 00:59:13.310 ⇒ 00:59:15.700 Jamie Perlmutter: I know we haven’t. We didn’t talk, Luca. We didn’t talk.

638 00:59:15.700 ⇒ 00:59:20.679 Uttam Kumaran: We talked, Luca. We didn’t talk Drake. We didn’t talk football. Yeah. So next time.

639 00:59:20.680 ⇒ 00:59:22.369 Jamie Perlmutter: We’ll save that for the in person.

640 00:59:22.370 ⇒ 00:59:23.950 Uttam Kumaran: Yeah, yeah, okay. Okay.

641 00:59:23.950 ⇒ 00:59:24.540 Jamie Perlmutter: All right.

642 00:59:24.826 ⇒ 00:59:27.399 Jamie Perlmutter: Awesome dude. Alright! We’ll chat soon. See you, man.