Meeting Title: Bi-Weekly—Uttam <> Brian Date: 2024-02-29 Meeting participants: Brian Yang, Uttam Kumaran
WEBVTT
1 00:05:35.260 ⇒ 00:05:36.309 Uttam Kumaran: Thank you.
2 00:05:39.740 ⇒ 00:05:42.550 Uttam Kumaran: Yup! Yup. yo! How’s it going?
3 00:05:43.950 ⇒ 00:05:45.849 Brian Yang: Not bad! How are you?
4 00:05:46.080 ⇒ 00:05:49.969 Uttam Kumaran: Not bad, what? Not? Great?
5 00:05:50.310 ⇒ 00:06:03.779 Uttam Kumaran: Well, I haven’t made any progress on any of like my projects. So Taiwan is pretty fun, pretty cool country. Oh, shit no way. That’s great. How long you been there.
6 00:06:04.360 ⇒ 00:06:09.409 Brian Yang: I got here a week ago. I think it’s
7 00:06:09.870 ⇒ 00:06:17.610 Brian Yang: yeah. It’s like pretty UN underrated. It’s like, it’s like, kind of like Japan, mixed with like the messiness of like a Vietnam or like a.
8 00:06:17.870 ⇒ 00:06:19.770 Brian Yang: it’s like controlled chaos. Yeah.
9 00:06:20.320 ⇒ 00:06:37.819 Uttam Kumaran: yeah, what I mean, like, what’s the history of Taiwan? I guess I really didn’t. I don’t really know a lot. I know a lot of the history of Hong Kong. I. The only thing I know about Taiwan cause I’ve been reading recently is because of, you know. Tsmc, and I’ve learned a little bit about the I know about the politics there. But
10 00:06:38.140 ⇒ 00:06:39.070 Uttam Kumaran: yeah.
11 00:06:40.080 ⇒ 00:06:43.380 Brian Yang: yeah, it’s like, well, I think, basically.
12 00:06:44.160 ⇒ 00:06:55.330 Brian Yang: there’s a I mean, II probably shouldn’t know more, being Chinese and everything. But I don’t know. I think basically, there was a revolution like the revolution. And then.
13 00:06:55.440 ⇒ 00:07:02.840 Brian Yang: like the like Union or party, like lost, like Democrats or
14 00:07:02.920 ⇒ 00:07:15.790 Brian Yang: Democracy versus like Communism, I guess. And then they kind of fled to Taiwan, and they just like they’re like 3 guys. We’re just gonna make a new country here. So then they just kind of fled. They just came here, and they just kinda like did their own thing. Now.
15 00:07:17.240 ⇒ 00:07:19.810 Uttam Kumaran: damn, that’s interesting. How’s the food and everything?
16 00:07:20.280 ⇒ 00:07:22.649 Brian Yang: The food is good.
17 00:07:22.680 ⇒ 00:07:24.679 Brian Yang: it’s really good. There’s a
18 00:07:24.930 ⇒ 00:07:28.490 Brian Yang: been having this like popcorn, fried chicken.
19 00:07:28.630 ⇒ 00:07:30.120 Brian Yang: which I think it’s like.
20 00:07:30.660 ⇒ 00:07:34.000 Brian Yang: It’s like fried chicken, but it’s got like
21 00:07:34.570 ⇒ 00:07:35.820 Brian Yang: what was that?
22 00:07:36.020 ⇒ 00:07:37.630 Uttam Kumaran: No, no, I’d say.
23 00:07:37.980 ⇒ 00:07:48.399 Brian Yang: Oh, they use they use a different batter. So like it’s like lighter. And then there’s if you’re into like the weird stuff there’s like stinky toe for here, too, which which everyone I think should turn on.
24 00:07:48.650 ⇒ 00:07:51.010 Brian Yang: Yeah.
25 00:07:51.840 ⇒ 00:07:53.630 Uttam Kumaran: Nice. Are you in hotel?
26 00:07:54.220 ⇒ 00:07:55.789 Brian Yang: No, Airbnb.
27 00:07:56.980 ⇒ 00:08:02.159 Uttam Kumaran: cool. Okay. How’s the Airbnb process? There is like very similar to here.
28 00:08:03.030 ⇒ 00:08:07.450 Brian Yang: Yeah, you just bring something online in the area, like just book it same thing.
29 00:08:09.040 ⇒ 00:08:12.890 Uttam Kumaran: you know. And do you end up meeting the the owners, or there’s just like a check in.
30 00:08:13.150 ⇒ 00:08:14.809 Uttam Kumaran: Thank you. Forget the keys.
31 00:08:15.120 ⇒ 00:08:18.649 Uttam Kumaran: It’s like a smart lock, you know, through a code.
32 00:08:18.770 ⇒ 00:08:20.050 Brian Yang: Yeah, it’s crazy
33 00:08:20.970 ⇒ 00:08:33.280 Uttam Kumaran: dude. I was. Gonna say. the Austin home prices are pretty much doing what I thought was going to happen. Stuff is like really coming down here.
34 00:08:33.480 ⇒ 00:08:38.770 Uttam Kumaran: It’s almost close to trend. But basically, we’re about like 30,
35 00:08:39.100 ⇒ 00:08:46.860 Uttam Kumaran: we’re gonna we’re gonna near like a probably 20 or 30% off the highs. pretty soon.
36 00:08:46.880 ⇒ 00:08:49.119 Uttam Kumaran: And I think this
37 00:08:49.260 ⇒ 00:08:55.559 Uttam Kumaran: this summer in particular, there’s a ton of deals, not only on like condos and apartments.
38 00:08:55.820 ⇒ 00:09:00.479 Uttam Kumaran: but also on like houses. A lot of stuff is coming back
39 00:09:00.690 ⇒ 00:09:09.609 Uttam Kumaran: to the ground like, and prices are very, very good. II wish I was in a capacity to buy. but I’m I’m probably gonna try and rent a house
40 00:09:09.800 ⇒ 00:09:16.990 Uttam Kumaran: and then for a year, and then kind of see if I can build enough. Build up enough cash to try to get something next year.
41 00:09:17.290 ⇒ 00:09:21.109 Uttam Kumaran: But planning planning the the
42 00:09:21.490 ⇒ 00:09:25.859 Brian Yang: sec. Yeah.
43 00:09:29.070 ⇒ 00:09:35.000 Uttam Kumaran: yeah, I don’t. II feel like, finally, stuff is coming, becoming a little bit more cheaper here.
44 00:09:35.020 ⇒ 00:09:36.889 Uttam Kumaran: And it seems like.
45 00:09:37.400 ⇒ 00:09:43.910 Uttam Kumaran: especially in the rents, like a lot of places, are offering a lot of concessions. And there’s a lot of supply coming online. So
46 00:09:45.090 ⇒ 00:09:54.700 Brian Yang: yeah, I heard, like, if they ban, like, basically hedge like funds from buying houses, that’s the only way that will basically fix it like housing crisis. Right?
47 00:09:55.010 ⇒ 00:10:03.620 Uttam Kumaran: Well, that’s happening a lot in like other parts of the Us. Where the peep. So what happened is because there’s no supply of people
48 00:10:03.830 ⇒ 00:10:10.890 Uttam Kumaran: hedge funds and not hedge funds, but like large institutional like real estate trust, came in and bought
49 00:10:11.090 ⇒ 00:10:22.629 Uttam Kumaran: like Black rock. And a ton of these guys. And basically, what they’re able to do is because they have all these other profits from the other parts of their business. They just can keep prices artificially high.
50 00:10:22.650 ⇒ 00:10:27.119 Uttam Kumaran: And there’s also they just turn those into like rental properties.
51 00:10:27.220 ⇒ 00:10:32.989 Uttam Kumaran: Pretty turnkey. And so yeah, there’s some propositions that are pushing towards
52 00:10:33.710 ⇒ 00:10:39.699 Uttam Kumaran: like, not requiring, like making sure that you can’t do that, or you have to have only a limited amount.
53 00:10:39.940 ⇒ 00:10:53.760 Uttam Kumaran: But Austin isn’t too afflicted by that. I feel like I think that’s we have a pretty healthy supply of new buildings. both residential and like residential homes and apartments.
54 00:10:54.110 ⇒ 00:11:05.530 Uttam Kumaran: Like. There’s a shit ton of supply dude like, there’s so many new buildings are all the buildings that probably were just getting started when you were here getting built over the last 2 years and are now coming online.
55 00:11:05.590 ⇒ 00:11:09.170 Uttam Kumaran: like, there’s so much supply. So
56 00:11:09.280 ⇒ 00:11:10.430 Uttam Kumaran: what happened?
57 00:11:10.710 ⇒ 00:11:13.390 Brian Yang: Huh? Yeah.
58 00:11:13.650 ⇒ 00:11:31.430 Uttam Kumaran: Yeah. Cause it takes a while for it to catch up right? Like, yeah, no, it takes. But the thing is, the nice thing about Austin is not like New York, where it takes like how long to build. They built these things in like 2 years, like the quality is one thing, but you know they’re building like pretty big luxury buildings. The problem is is so much supply
59 00:11:31.440 ⇒ 00:11:39.239 Uttam Kumaran: that the exodus to Texas and the Austin isn’t as high as before but also affordability like
60 00:11:39.460 ⇒ 00:11:56.010 Uttam Kumaran: for our like salary bands, it’s really affordable. But think about if you’re in Austin, and you’re making like 70 k. 80 k. Like, it should be affordable for you. And that’s the people that are really have a hard time moving down here people who are looking for apartments like 1,500 or less.
61 00:11:56.370 ⇒ 00:12:10.120 Uttam Kumaran: which again, me and you’re gonna be like, Yo, that’s this fucking steel, but I have so many friends here that are like, I want an apartments like 1,300 bucks, and they could find it like north of Austin. And so I think the price is gonna help those people move closer to city
62 00:12:10.400 ⇒ 00:12:15.399 Uttam Kumaran: because a lot of people, I think from out of state are
63 00:12:15.450 ⇒ 00:12:20.479 Uttam Kumaran: like slowing down moving here, and it’s the people that are here, I think, are people that are gonna be here for a while.
64 00:12:20.710 ⇒ 00:12:31.070 Uttam Kumaran: But there’s a ton of supplies. So it’s nice. And you know they’re building a lot of buildings everywhere, and all the buildings are pretty good. So I have friends at a bunch of them
65 00:12:31.780 ⇒ 00:12:34.209 Brian Yang: cool. You think you’ll
66 00:12:34.530 ⇒ 00:12:38.129 Brian Yang: I mean, if you buy a place you’re thinking long term, Austin, right? At least for a little bit.
67 00:12:38.140 ⇒ 00:12:50.969 Uttam Kumaran: Yeah, I’d like to. I mean, my ultimate like, my dream dream is to be able to have a place here and have a place in New York. We’re gonna see if I can accomplish my.
68 00:12:51.010 ⇒ 00:13:03.710 Uttam Kumaran: But yeah, I I would. I would love to have a place here like long term for sure. And then, ideally, like, you know, as soon as stuff gets stable for me, I’m gonna travel way more
69 00:13:03.770 ⇒ 00:13:05.799 Uttam Kumaran: and I’ll be able to get out
70 00:13:05.910 ⇒ 00:13:19.690 Uttam Kumaran: and travel like normally. But Austin’s a really good place, because you’re you’re within. you know. You can go to both coasts pretty easily. I think quality of life is really good here. And I do most of my stuff on the computer, anyways.
71 00:13:19.890 ⇒ 00:13:23.059 Uttam Kumaran: That if I need to go to New York or something, I can
72 00:13:23.550 ⇒ 00:13:30.950 Uttam Kumaran: go manage I mean, what I would love to do is like, have a place here, and then find some way to get a place in New York.
73 00:13:31.130 ⇒ 00:13:34.990 Uttam Kumaran: like with someone like I have a friend in San Francisco who’s like dude. We should
74 00:13:35.050 ⇒ 00:13:38.990 Uttam Kumaran: try to buy up like a apartment or something in New York
75 00:13:39.050 ⇒ 00:13:41.959 Uttam Kumaran: that way. We have it, and then we can rent it out because
76 00:13:42.210 ⇒ 00:13:47.740 Uttam Kumaran: New York is never going anywhere, and you know I’d love to spend more time there because all my friends are there.
77 00:13:47.820 ⇒ 00:13:55.259 Uttam Kumaran: But buying there is a real real tough thing. but that’s like that’s like a long-term goal. Ideally, I would like to have a house
78 00:13:55.480 ⇒ 00:14:03.759 Uttam Kumaran: here, especially sometime in the next year or 2, because I think this place is just gonna slowly creep up like I don’t know what if the returns are gonna
79 00:14:04.290 ⇒ 00:14:06.460 Uttam Kumaran: match the market. But
80 00:14:06.620 ⇒ 00:14:16.250 Uttam Kumaran: I would say like, it’s a no brainer to buy a house closer to downtown here because there’s so much business happening that it’s a good bet. It’s not like
81 00:14:16.590 ⇒ 00:14:19.279 Uttam Kumaran: it’s not like you’re buying in the Bay area at the peak.
82 00:14:19.330 ⇒ 00:14:29.250 Uttam Kumaran: This is like about it’s like, basically just starting here. I feel like. And compared to Dallas and Houston, I’ve met a lot of people moved from Houston or Dallas here.
83 00:14:29.270 ⇒ 00:14:42.399 Uttam Kumaran: and they like it way more. So I don’t know. I just have like a lot of anecdotal evidence about this place that I’m kind of like. Damn. I kind of see the future a little bit. Ii just like don’t have like a ton of liquid cash to
84 00:14:42.470 ⇒ 00:14:50.150 Uttam Kumaran: and I don’t. I don’t wanna buy house that like I don’t like. I just thought for me. I wanna be able to live there. So I’m kind of just like, maybe I’ll wait a year and try to
85 00:14:50.470 ⇒ 00:14:52.429 Uttam Kumaran: save the money to do so. So
86 00:14:53.050 ⇒ 00:14:55.980 Brian Yang: yeah. yeah, that makes sense. I mean.
87 00:14:56.250 ⇒ 00:15:14.940 Brian Yang: like, what you want to do is like kind of similar to what I want to do, except I want to go like I wanna chip out and go to different countries like, I want, like 3 or 4 cities that area like, and and just like Airbnb about, or something when I’m not there and just hop between them, because it’s like it’s kind of tiring to do it. I mean, I mean, I think often is a good.
88 00:15:15.160 ⇒ 00:15:19.659 Uttam Kumaran: I think states overall. Do I mean, I think.
89 00:15:19.900 ⇒ 00:15:31.959 Uttam Kumaran: like, unless you want to be in like a Atier city, which is like Sf. La New York, Miami. If you want to be in like a cooler city where it’s like it actually feels like you’re coming somewhere like home.
90 00:15:32.330 ⇒ 00:15:37.179 Uttam Kumaran: I really fucking love this place. It’s like it’s been really really nice.
91 00:15:37.210 ⇒ 00:15:47.609 Uttam Kumaran: I think the weather is like nice year round, and I think it’s easy to buy property here and turn into Airbnb very, very easily, and there’s still a lot a lot of tourism.
92 00:15:48.350 ⇒ 00:15:51.540 Uttam Kumaran: so I don’t know. I think this is a good place to kind of have that
93 00:15:51.830 ⇒ 00:16:03.930 Uttam Kumaran: like kind of like base. I don’t know where else, and I don’t know honestly don’t know where else in the Us. I would consider like I think, Denver, and stuff like that, like it’s so crowded there now, and it’s way too cold for me.
94 00:16:03.960 ⇒ 00:16:09.159 Uttam Kumaran: Miami, or some. The problem here is that you don’t have a coastline. So that’s a little bit of a
95 00:16:09.190 ⇒ 00:16:17.109 Uttam Kumaran: a little bit rough, but it makes up for it in like the lifestyle, and the people, and the food and the tourism and stuff like that. So
96 00:16:17.580 ⇒ 00:16:18.490 Uttam Kumaran: yeah.
97 00:16:18.960 ⇒ 00:16:29.859 Brian Yang: yeah, Austin is definitely like, very like in the Us. Is probably highest on my list. Right? Yeah, like for what you’re saying. New York and Sf, for like out of reach. Austin.
98 00:16:31.120 ⇒ 00:16:35.360 Brian Yang: yeah, that’s good. So that
99 00:16:35.880 ⇒ 00:16:45.669 Brian Yang: so doing some, we work stuff doing some stuff for India. Just kind of coasting right now until I get back to the Us with
100 00:16:45.930 ⇒ 00:16:50.529 Brian Yang: with more regular schedule. And I can actually do some real real work again.
101 00:16:50.630 ⇒ 00:16:52.420 Uttam Kumaran: When is that?
102 00:16:53.150 ⇒ 00:16:55.489 Brian Yang: I fly back on?
103 00:16:55.530 ⇒ 00:16:57.089 3, 14.
104 00:16:57.960 ⇒ 00:17:00.410 Uttam Kumaran: Oh, okay. So you’re it’s like, it’s coming up.
105 00:17:00.700 ⇒ 00:17:03.599 Brian Yang: Yeah, I’m going to Chicago for a wedding.
106 00:17:03.840 ⇒ 00:17:07.790 Brian Yang: for we’re for a wedding and then back to Maryland for like
107 00:17:07.950 ⇒ 00:17:13.309 Brian Yang: a month month and a half. Ish. I’m half sitting for my parents.
108 00:17:14.810 ⇒ 00:17:15.619 Brian Yang: Yeah.
109 00:17:15.770 ⇒ 00:17:17.260 Uttam Kumaran: okay, okay, cool.
110 00:17:17.609 ⇒ 00:17:25.690 Uttam Kumaran: Might. Am I trying to make a trip up to Austin? If if you’re well, I was gonna say, I mean, I don’t know if you if only a month I don’t wanna like
111 00:17:25.859 ⇒ 00:17:31.259 Uttam Kumaran: put that. But if you have time dude, and you’re here like definitely come visit. I am.
112 00:17:31.310 ⇒ 00:17:44.470 Uttam Kumaran: Gonna be around all that month, even if you want to just come for a weekend, or whenever I mean, I basically I basically work from home and it’s pretty chill. So
113 00:17:44.860 ⇒ 00:17:46.540 Uttam Kumaran: I’m definitely
114 00:17:47.300 ⇒ 00:17:56.449 Uttam Kumaran: definitely love it. And you know we could. We could also do some work together, and I have some progress on some of the provider stuff. So yeah, I mean, it’d be it’d be pretty fun.
115 00:17:56.660 ⇒ 00:18:04.369 Uttam Kumaran: I have like I have a ton of monitors, and like, you know, I have a bunch of places like, go sit and like we get gym and play tennis and stuff like that. So
116 00:18:04.800 ⇒ 00:18:12.679 Brian Yang: yeah, yeah, I could. Probably I think I can do that stuff to figure out the dates. But yeah, I think that would be a lot of fun.
117 00:18:12.690 ⇒ 00:18:17.049 Uttam Kumaran: Yeah, text me text me like one day. So maybe I’ll send you like a couple of weekends that
118 00:18:17.120 ⇒ 00:18:21.900 Uttam Kumaran: that work best for me. Yeah, we should do something like that.
119 00:18:22.350 ⇒ 00:18:24.049 Brian Yang: Yeah.
120 00:18:24.760 ⇒ 00:18:29.740 Brian Yang: yeah, I’ll I’ll text you. Not sure the dates yet? But yeah.
121 00:18:29.770 ⇒ 00:18:30.920 Uttam Kumaran: okay, cool.
122 00:18:32.100 ⇒ 00:18:32.890 Brian Yang: Cool.
123 00:18:33.250 ⇒ 00:18:51.890 Uttam Kumaran: How’s that? I saw this? Yeah, you’re like, you’re like, posted real meetings and stuff on stock. On one hand, I’m still me like it’s not like
124 00:18:52.180 ⇒ 00:18:54.520 Uttam Kumaran: I’m not like changed, but
125 00:18:54.870 ⇒ 00:19:01.840 Uttam Kumaran: I do have like people that are now working for the company part. Most of the people are part time. But basically.
126 00:19:01.970 ⇒ 00:19:05.979 Uttam Kumaran: I was able to kind of like, just get some more consistent income coming in.
127 00:19:06.030 ⇒ 00:19:13.159 Uttam Kumaran: and I threw more my money in. So I cause I wanted some of this stuff done. And so basically I have.
128 00:19:13.400 ⇒ 00:19:24.250 Uttam Kumaran: I just found some really really good people in each like stack part of the stack for data, like, I found a really really good analyst person who’s based in Seattle is a friend of a friend.
129 00:19:24.340 ⇒ 00:19:27.840 Uttam Kumaran: He’s like a really, really rock star, like kind of analyst.
130 00:19:27.930 ⇒ 00:19:38.749 Uttam Kumaran: I found a really really good like dashboard ux, like information architecture, person. and then me and one other person kind of cover, like modeling, like I’ll do like.
131 00:19:39.060 ⇒ 00:19:44.979 Uttam Kumaran: I could move really, really quickly on models, and but I just don’t have a ton of time. And I found someone who’s like, probably like a Bee
132 00:19:45.380 ⇒ 00:19:48.199 Uttam Kumaran: B level modeler. But he’s really really cheap
133 00:19:48.770 ⇒ 00:19:59.009 Uttam Kumaran: like 35 bucks an hour for like modeling stuff which is really good. And then I found someone who’s helping me with like more like data, and on the on the provider side. So
134 00:19:59.180 ⇒ 00:20:08.340 Uttam Kumaran: basically, I’m having him not only write some Etl first for one of my clients directly in Snow Park. I’m also having him
135 00:20:08.470 ⇒ 00:20:10.649 Uttam Kumaran: do some of the provider work.
136 00:20:10.770 ⇒ 00:20:16.630 Uttam Kumaran: And then I’m just kind of like, gonna keep pushing on that, because I really think there’s a huge opportunity.
137 00:20:16.940 ⇒ 00:20:29.420 Uttam Kumaran: and yeah, so it’s like, it’s just a couple of people. It’s going really. Well, I think it’s I’m just not in the real as everybody kinda like started getting into the motions the last 2 weeks.
138 00:20:29.490 ⇒ 00:20:35.289 Uttam Kumaran: So I’m just like now, remembering how hard it is to like keep track of like a team of people and make sure everybody
139 00:20:35.470 ⇒ 00:20:44.109 Uttam Kumaran: has tasks and is getting shit done. So I’m like trying to find like, Okay, do I need? I don’t really want to do fucking daily stand ups.
140 00:20:44.450 ⇒ 00:20:45.500 And
141 00:20:45.670 ⇒ 00:20:55.870 Uttam Kumaran: but I’m trying to get everybody to do Async. But then people don’t respond to them to the message which you know I get. So I’m trying to find like better ways. I may actually just have
142 00:20:56.040 ⇒ 00:21:09.429 Uttam Kumaran: people present their work at the end of the week, or do demo days or something like that that way? It kind of pushes people to like have something to present. But yeah, it’s just a small team again. Everybody’s like, very, very cool, and everybody’s engineering. So
143 00:21:09.520 ⇒ 00:21:14.480 Uttam Kumaran: it’s as it’s as little business as I can get it to be. So
144 00:21:16.620 ⇒ 00:21:17.989 Brian Yang: yeah, well, I think.
145 00:21:18.370 ⇒ 00:21:21.519 Brian Yang: yeah, doing a good job. Yeah, I think.
146 00:21:21.930 ⇒ 00:21:35.349 Uttam Kumaran: I don’t know. Feel like if I had to run the team. But I probably like run it with all the practices I like hate like daily stand ups and stuff like that. No dude, that’s the thing. But some of the some of it works. Some of it is. So
147 00:21:35.520 ⇒ 00:21:47.870 Uttam Kumaran: II just want II just wanted to get work done. And I tell people dude, if we’re able to get good work done. I think it’ll get us more business. So like, I try to set up the incentives in a way where it’s like really positive meaning, like.
148 00:21:47.950 ⇒ 00:21:54.689 Uttam Kumaran: if we get work done, I’m able to get higher rates, get more hours. Everybody gets paid more, and everybody has a full bill of work
149 00:21:54.800 ⇒ 00:22:00.250 Uttam Kumaran: like that’s what I’m hoping. And so I’m trying to set up the narrative in a way where it’s pretty positive. But
150 00:22:01.430 ⇒ 00:22:02.270 Brian Yang: yeah.
151 00:22:04.840 ⇒ 00:22:05.740 Brian Yang: yeah.
152 00:22:05.910 ⇒ 00:22:12.039 Brian Yang: I’m just like I’m working on this like week? Query stuff with like the chatty Pt. Wrapper on like
153 00:22:12.290 ⇒ 00:22:29.750 Brian Yang: our snowflake. And then I’m working with these like other engineers who like do nothing. And then I just it’s just like me. And like this, it’s Guy called Ruben, who’s like he does. He does stuff. So he’s like very junior. But he like, does a lot of stuff.
154 00:22:29.820 ⇒ 00:22:35.389 Brian Yang: So he’s like, really good. So what’s the what’s is the week query thing working?
155 00:22:36.260 ⇒ 00:22:42.680 Uttam Kumaran: Yeah, it works. It’s a. It’s a streamlit. So basically, it’s a streamlit app. Have you played around with streamlit on Snowflake a little bit
156 00:22:43.040 ⇒ 00:22:57.870 Brian Yang: so like it’s like you have Chat Gp, right? Like the the Ui, it’s like basically trying to rebuild that ui on chat, Gp on Snowflake. Except it cannot. You can ask it questions on your data be, but because we’re passing metadata off off to like
157 00:22:57.930 ⇒ 00:23:08.799 Brian Yang: Chat Gvt for us to answer sequel queries. So someone that’s a Vi guy who doesn’t know any sequel, can ask the question about a table, and it will spit back a sequel query and then rent on Snowflake.
158 00:23:09.030 ⇒ 00:23:14.520 Uttam Kumaran: So how hard would it be. I have a client that has a bunch of Zendesk tickets that I want to use
159 00:23:14.640 ⇒ 00:23:22.320 Uttam Kumaran: Chatbot or Claude, or something to categorize and summarize. How hard do you think that’s that would be to build
160 00:23:24.950 ⇒ 00:23:32.460 Uttam Kumaran: like it’s it’s not a chat interface, though. I just want to create like a function that takes text and like passes it to one of those Llms.
161 00:23:33.500 ⇒ 00:23:47.659 Brian Yang: I don’t think it’s that hard to build. The the tricky part is you, if it’s a client, and you have to like like cause. If you if you send it to Microsoft, it’s on like Microsoft servers. So you have all this like data for privacy. Bs you have to deal with.
162 00:23:48.010 ⇒ 00:23:51.840 Uttam Kumaran: So what do you think is Beth like?
163 00:23:51.960 ⇒ 00:23:58.270 Uttam Kumaran: I mean, I could ask them whether they care or not. But like, would that just be an external function, or had like, what’s the mechanics?
164 00:23:59.650 ⇒ 00:24:06.819 Brian Yang: So if I was to build that, you don’t care about that being on Snowflake right. It’s not like Zendesk is in Snowflake.
165 00:24:06.900 ⇒ 00:24:15.249 Uttam Kumaran: Well, the data is in Snowflake, but I prefer to be there, because then it’s like II run up with snowflake usage. That’s what I want to do.
166 00:24:16.000 ⇒ 00:24:21.489 Brian Yang: Oh, okay. yeah. So you want to go through Snowflake or no, you don’t.
167 00:24:25.050 ⇒ 00:24:26.180 Brian Yang: Please hear me.
168 00:24:28.780 ⇒ 00:24:29.660 Brian Yang: Hello!
169 00:24:34.700 ⇒ 00:24:36.480 Uttam Kumaran: Hello! Hello.
170 00:24:37.130 ⇒ 00:24:38.309 Brian Yang: yeah. Can you hear me?
171 00:24:43.990 ⇒ 00:24:44.820 Hello.
172 00:24:57.810 ⇒ 00:24:58.990 Uttam Kumaran: can you hear me? Now?
173 00:25:00.130 ⇒ 00:25:01.580 Brian Yang: Yeah. Barely.
174 00:25:07.420 ⇒ 00:25:08.850 Uttam Kumaran: What? The fuck?
175 00:25:09.140 ⇒ 00:25:10.470 Brian Yang: Okay. I heard that.
176 00:25:10.850 ⇒ 00:25:12.079 Uttam Kumaran: Okay, how about now?
177 00:25:12.150 ⇒ 00:25:13.590 Brian Yang: Yeah. I can hear you now.
178 00:25:13.740 ⇒ 00:25:15.870 Uttam Kumaran: Okay, okay, cool. Alright. Go for it.
179 00:25:15.900 ⇒ 00:25:17.910 Brian Yang: Okay? So if if
180 00:25:18.280 ⇒ 00:25:28.350 Brian Yang: if if it’s not in Snowflake and you have a bunch of like random Csv’s. Then you could just write like a simple python script that, like, you know, just
181 00:25:28.620 ⇒ 00:25:29.690 Brian Yang: passes.
182 00:25:29.900 ⇒ 00:25:35.759 Brian Yang: Just use the python like SDK, to like. Take that data and just like, feed it in. Have a
183 00:25:36.290 ⇒ 00:25:37.110 Brian Yang: oh
184 00:25:37.230 ⇒ 00:25:38.650 Uttam Kumaran: for historical data.
185 00:25:39.240 ⇒ 00:25:44.630 Brian Yang: Yeah, so just write like a hypothetical thing with an SDK, or use whatever SDK you want, like, whatever language.
186 00:25:45.170 ⇒ 00:25:48.339 Brian Yang: if it’s on Snowflake and you wanna do it through Snowflake.
187 00:25:49.110 ⇒ 00:25:54.300 Uttam Kumaran: So he’s let me tell you about the snow keeping around
188 00:25:54.610 ⇒ 00:25:58.009 Uttam Kumaran: 2. But I have gone to
189 00:25:58.290 ⇒ 00:26:02.349 Brian Yang: wait. II can. I can hear like everything. Can you hear me?
190 00:26:03.160 ⇒ 00:26:14.390 Uttam Kumaran: I can hear, like every third word you’re saying. Okay, hold on. Let me run to my car. And there’s construction like, it’s a nightmare for me.
191 00:26:16.800 ⇒ 00:26:19.540 Brian Yang: How about now?
192 00:26:19.980 ⇒ 00:26:21.750 Brian Yang: Yeah, that’s better. You didn’t break up.
193 00:26:21.860 ⇒ 00:26:27.859 Uttam Kumaran: Okay? So basically, I also want to try and give their customer success team
194 00:26:27.880 ⇒ 00:26:32.120 Uttam Kumaran: a way to categorize these tickets as they come in.
195 00:26:35.640 ⇒ 00:26:38.050 Uttam Kumaran: So like, I want live tickets
196 00:26:38.610 ⇒ 00:26:48.760 Uttam Kumaran: to come in, and maybe through a streamlet app, we build them. They could see the ticket, and then they could get produced like an automatic summary based on all previous responses.
197 00:26:48.820 ⇒ 00:26:52.230 Uttam Kumaran: and I like to build all that on
198 00:26:52.900 ⇒ 00:27:04.640 Uttam Kumaran: streamlit, if possible. So there’s like a historical use case that I want to do some analysis. but there’s also like a sorry, Jesus Christ. There’s like a fucking massive fan above my car.
199 00:27:04.810 ⇒ 00:27:09.389 Uttam Kumaran: there’s there’s a historical use case. But there’s also a
200 00:27:09.650 ⇒ 00:27:16.219 Uttam Kumaran: like a go live used case, which is like as a ticket comes in. I want to summarize it and provide them with
201 00:27:16.250 ⇒ 00:27:18.590 Uttam Kumaran: hey? Here’s probably what you should respond with.
202 00:27:21.050 ⇒ 00:27:22.130 Brian Yang: Got it.
203 00:27:24.210 ⇒ 00:27:30.850 Brian Yang: see? So I’ve only played around with like, open ais like chat, like completions, basically chat, Gp.
204 00:27:32.060 ⇒ 00:27:33.240 Uttam Kumaran: and okay.
205 00:27:33.840 ⇒ 00:27:49.569 Brian Yang: And the hard part with that is, you can’t really like feed it historical data and then have it like augment the model. There’s stuff called like, let me show you this thing, rag and stuff like that. So I have to like, I’ll have to do embeddings and things like that. I think.
206 00:27:49.970 ⇒ 00:27:56.980 Brian Yang: yeah, exactly. But again, like, what’s the interface versus like external functions? Cause? I still want to build on streamlist.
207 00:27:57.230 ⇒ 00:28:05.070 Uttam Kumaran: I think I could have stuff externally do that. But like, have you played around with calling external Apis like, are there limitations?
208 00:28:06.740 ⇒ 00:28:11.989 Brian Yang: Yeah. That’s what the snowflake thing was doing right? Like, I’m calling a snowflake
209 00:28:12.270 ⇒ 00:28:18.589 Brian Yang: function and passing it a string with what I basically, I would ask, Chatty, Pt.
210 00:28:19.640 ⇒ 00:28:29.280 Uttam Kumaran: but I guess what I’m saying is like, you’re calling open AI, and that lives externally like, are there problems calling external functions from Snowflake?
211 00:28:29.760 ⇒ 00:28:30.560 Brian Yang: No.
212 00:28:30.680 ⇒ 00:28:33.550 Brian Yang: the okay. Let’s totally open. You can do whatever you want.
213 00:28:33.840 ⇒ 00:28:36.510 Brian Yang: Yeah, there are
214 00:28:37.250 ⇒ 00:28:43.949 Brian Yang: I don’t know. There. there might be some limitations with like government cloud, and like snowflakes like
215 00:28:43.960 ⇒ 00:28:51.660 Brian Yang: networking side. But I’m just able to make a you know, any restful request and endpoint. And it’s fine.
216 00:28:53.490 ⇒ 00:29:00.189 Uttam Kumaran: Okay? Cause that’s what I want to do. So I want to give them something where. And I have. Like, I’m emailing the CEO back and forth. And
217 00:29:00.540 ⇒ 00:29:09.920 Uttam Kumaran: I was like, Yo, I think we could build this on all your Zendesk historical data. Once the one thing I want to build is like I want to do back. I want to do some historical analysis on all the tickets.
218 00:29:10.190 ⇒ 00:29:17.980 Uttam Kumaran: So that’s one thing. The second thing I wanna do is have like a build, a quick little streamlit app for their customer success team
219 00:29:18.200 ⇒ 00:29:20.470 Uttam Kumaran: to be able to put in
220 00:29:20.550 ⇒ 00:29:24.769 Uttam Kumaran: like a ticket and then get out like, here’s what you should respond with.
221 00:29:25.720 ⇒ 00:29:42.019 Uttam Kumaran: or something like that, where it can pull in the customer it could pull in their previous orders like it has all this context that the customer success person would have to go get right? Because what does the customer success person? Do? They get an email for someone. They go look up their order, they go look up their past tickets.
222 00:29:42.090 ⇒ 00:29:50.189 Uttam Kumaran: they go look up like, okay, what the fuck is even going on. What follow up questions do I need to ask? I want to do all that with AI. I want to do all in streamlit.
223 00:29:52.670 ⇒ 00:29:59.149 Brian Yang: Yeah, so the data is landing in Zendesk. How’s it getting into Snowflake?
224 00:29:59.230 ⇒ 00:30:01.009 Uttam Kumaran: It’s just 5 tram.
225 00:30:02.120 ⇒ 00:30:05.620 Brian Yang: Okay? So the so the schemas and stuff are pretty clean. Right?
226 00:30:05.770 ⇒ 00:30:09.679 Uttam Kumaran: Yeah, yeah, I mean, I’m we’re using it for analysis. And so that’s like, super clean.
227 00:30:11.900 ⇒ 00:30:13.089 Brian Yang: Yeah, so like
228 00:30:14.550 ⇒ 00:30:16.270 Brian Yang: a ticket ticket comes in.
229 00:30:16.420 ⇒ 00:30:20.570 Brian Yang: you can have the user open up
230 00:30:22.020 ⇒ 00:30:27.700 Brian Yang: the streamlit app that you’ve like, basically hosting on Snowflake. Right? They open that up
231 00:30:27.930 ⇒ 00:30:34.079 Brian Yang: and you’ll have to keep track of like the tickets they’ve worked on. And the tickets that just came in right?
232 00:30:34.120 ⇒ 00:30:36.500 Brian Yang: So like you can throw them at.
233 00:30:37.540 ⇒ 00:30:46.189 Brian Yang: maybe it’s not even streamlined or dashboard of basically, these are the new tickets. And here’s the context for the new ticket. Here’s a suggested response. But you’ll you’ll have to build that
234 00:30:46.650 ⇒ 00:30:58.319 Brian Yang: based on so like a ticket will come in. The user opens the app. Then queries go off on snowflake to find historical tickets made by this person. Basically, the context.
235 00:30:58.640 ⇒ 00:31:04.239 Brian Yang: You you take that right, that’s all. On Snowflake, on the qui, on the query. You spit that into like
236 00:31:04.280 ⇒ 00:31:14.329 Brian Yang: Openai. and then open air. We saw return something which is like your for like, suggested response based on the historical context.
237 00:31:14.340 ⇒ 00:31:15.700 Brian Yang: Yeah.
238 00:31:16.140 ⇒ 00:31:22.840 Brian Yang: And then, like, it’s not the the I mean this, this, the response isn’t gonna be perfect. So you you’ll need some sort of interface for the
239 00:31:22.850 ⇒ 00:31:31.940 Brian Yang: for your like response agent to like click through the context, to check, to make sure the response is right. And stuff like that, right? So, yeah, yeah, I mean, I don’t want to take them entirely out of the loop.
240 00:31:32.120 ⇒ 00:31:36.649 Uttam Kumaran: But I wanna give them like, 80% of the Fuckin response. So they can just copy paste that in.
241 00:31:37.930 ⇒ 00:31:44.260 Brian Yang: Yeah. So like what you’ll need for that is, you’ll need the clean data from Zendesk in in the Snowflake accounts you already have.
242 00:31:44.480 ⇒ 00:32:04.470 Brian Yang: You’ll need some functions to to call open AI which which you you have to write, and then you need to add some good sequel queries that like at, you know, new ticket comes in. Run the queries on your on your database to figure out the context for for this person and for this type of ticket, right? And then feed that into open eye. So yeah, it’s doable.
243 00:32:06.330 ⇒ 00:32:12.330 Uttam Kumaran: Okay, cool. I wanna try that I have. Now that I have, like someone who can help me write like python stuff. I I’m
244 00:32:12.630 ⇒ 00:32:23.199 Uttam Kumaran: I’m gonna have him help me write that function, do. The other thing is so I rewrote Walmart Api, which 5 train doesn’t support, and I tried to have them build it, and they, fucking
245 00:32:23.310 ⇒ 00:32:34.570 Uttam Kumaran: took like a more. They dare goats pretty much ghosted me, and I’m actually going to respond to them and say, Hey. you know haven’t heard back from you. But we actually went and built it ourselves. And we’re gonna list it on the marketplace. So like. get fucked.
246 00:32:34.620 ⇒ 00:32:35.970 Uttam Kumaran: But
247 00:32:36.180 ⇒ 00:32:40.919 Brian Yang: II wanna I wanna list it as a native app dude.
248 00:32:42.370 ⇒ 00:32:51.820 Uttam Kumaran: Wait, one of your employees is work building like who’s who who built the Walmart thing? Well, it’s like dude. We were using this company called Nexla, which is like a 5 train alternative
249 00:32:52.020 ⇒ 00:32:54.620 Uttam Kumaran: for pulling Walmart data.
250 00:32:54.670 ⇒ 00:33:04.459 Uttam Kumaran: And basically it. Next, you can write your own rest. Api. So we were just calling the Walmart Api and pulling down orders data. I just took that, rewrote it in Snowflake.
251 00:33:04.860 ⇒ 00:33:10.630 Uttam Kumaran: in Snow park at the store procedure take like, just the way we did fema stuff. And it works.
252 00:33:11.330 ⇒ 00:33:15.040 Uttam Kumaran: Oh, yeah. But I wanna I wanna be able to list that
253 00:33:15.200 ⇒ 00:33:23.460 Uttam Kumaran: for other people to use like in in Snowflake Market Place there’s a place where you can look up connectors.
254 00:33:23.900 ⇒ 00:33:36.290 Uttam Kumaran: or like they call it connectors, or fucking something like that. I wanna list like Walmart orders, Api and I just wanna take in your Api key, and then I’ll just spit out your orders thing.
255 00:33:39.610 ⇒ 00:33:43.779 Uttam Kumaran: and then I’ll charge you for, like per pole, or
256 00:33:43.990 ⇒ 00:33:46.260 Uttam Kumaran: I don’t know I have to think about it. But
257 00:33:46.980 ⇒ 00:33:49.900 Brian Yang: oh, yeah, yeah, that that that could work.
258 00:33:50.430 ⇒ 00:33:51.630 You see what I mean.
259 00:33:52.050 ⇒ 00:33:55.680 Brian Yang: How are you gonna get them to trust you to pass off their Api key, though.
260 00:33:55.960 ⇒ 00:34:00.830 Uttam Kumaran: Well, dude, I looked, there’s already people that are doing like Hubspot connectors.
261 00:34:00.910 ⇒ 00:34:05.220 all types of connectors already on there. I’ll send you a couple
262 00:34:06.840 ⇒ 00:34:15.380 Uttam Kumaran: I mean, like, II think the trust thing is like, I just have to like, I mean, one like I have. There’ll be some legal guarantees, basically
263 00:34:15.639 ⇒ 00:34:24.009 Uttam Kumaran: and then I’ll make sure that we don’t like. There’s what I’ll I’ll have to outline exactly this to show that, hey? We actually don’t see your Api key.
264 00:34:24.120 ⇒ 00:34:27.050 Uttam Kumaran: You know. I’ll make that super clear.
265 00:34:27.460 ⇒ 00:34:37.649 Uttam Kumaran: But I also it’s again, it’s a fuck. II don’t know what the mechanism is, but it’s basically a function. So I’m probably just giving them act the native app. And it runs on their instance. So I don’t see anything.
266 00:34:39.190 ⇒ 00:34:51.170 Uttam Kumaran: Oh, yeah, yeah, that’s easy. Then, yeah, that’s what I mean, like, II just, I’m basically just writing the python code listing it. And then the native app runs on their thing. So they pu they put in their Api key.
267 00:34:51.239 ⇒ 00:34:58.549 Uttam Kumaran: and then I probably can run both. I could probably create both the task and run the store procedure. And then that’s kind of theirs.
268 00:34:58.710 ⇒ 00:35:00.070 or something like that.
269 00:35:00.500 ⇒ 00:35:02.050 Brian Yang: Yeah,
270 00:35:02.280 ⇒ 00:35:23.459 Brian Yang: I don’t. How does Snowflake lock down your application code? Because it’s using all snowflake mechanisms right? Like I for for this project I had to download some Cyberson data. So I just like made a share. And then I cloned all the tables and then I dropped. I dropped to share to cybers in, because I just wanted one permanent copy. Right? What’s the stop there from just stealing your application?
271 00:35:24.280 ⇒ 00:35:37.710 Uttam Kumaran: I don’t think you can actually see the app code. The cybers and data is different, because now that you cloned it, you don’t get historical, right? You don’t get. You don’t get no, you’re not historicals. You don’t get new data like you won’t get the updates.
272 00:35:37.820 ⇒ 00:35:52.180 Uttam Kumaran: So that’s like, basically their only way of preventing the clone or like kind of being like, you know, you can clone it, but that’s why. But it’s still because mechanism, for example, I can allow you to view the data, but I won’t allow you to get new updates
273 00:35:52.460 ⇒ 00:36:10.079 Uttam Kumaran: so or like, I can give you a subset before you buy it. For example, cybers should say we’re only going to give you 3 months a date until you buy it, and then, after you buy it with the for a flat fee. It’s then something for updates, right. They could do something like that. They, these guys, are sponsored by snowflake. So they they just fucking do whatever they want like. They don’t make any money.
274 00:36:10.220 ⇒ 00:36:17.450 Uttam Kumaran: but I don’t think that the user actually sees the app code. They just see the app.
275 00:36:17.680 ⇒ 00:36:25.819 Uttam Kumaran: I’ll I’ll send you one. I’ll send you one of the. I’ll send you a loom of me walking through one of the Hubspot like connectors that I got from
276 00:36:26.160 ⇒ 00:36:32.920 Uttam Kumaran: a marketplace, because, literally, it’s just a Ui for you to input your Api keys, input what tables you want. And then you run it.
277 00:36:34.290 ⇒ 00:36:38.830 Uttam Kumaran: It’s sort of like they just. It’s like 5 train out of the box all on. So flake.
278 00:36:39.360 ⇒ 00:36:47.259 Brian Yang: yeah, it’s okay. So okay, that makes sense. It’s not like protecting your application code. It’s basically like they’re giving you a platform to write a Sas application.
279 00:36:47.430 ⇒ 00:36:53.180 Uttam Kumaran: Yeah? And they want you to basically sell like Etl as a service.
280 00:36:53.430 ⇒ 00:36:54.200 Brian Yang: Yeah.
281 00:36:54.480 ⇒ 00:36:59.939 Uttam Kumaran: it’s like every flavor of that. It’s like data sets. It’s like the actual python code.
282 00:37:00.430 ⇒ 00:37:04.539 Uttam Kumaran: And then I mean for me, it’s like dude. What I’m gonna do is basically
283 00:37:04.760 ⇒ 00:37:11.129 Uttam Kumaran: I’m going to rewrite the stuff that I’m going to rewrite the stuff. That’s not that complicated
284 00:37:11.250 ⇒ 00:37:17.610 Uttam Kumaran: right? Like pulling Walmart orders per day is fucking easy. I don’t. I don’t need that to run through.
285 00:37:18.120 ⇒ 00:37:20.980 Uttam Kumaran: I don’t need that to run through
286 00:37:21.480 ⇒ 00:37:24.079 Uttam Kumaran: 5 trained. Because I just need like 2 tables.
287 00:37:24.890 ⇒ 00:37:25.700 Brian Yang: Yeah.
288 00:37:26.990 ⇒ 00:37:28.979 Brian Yang: yeah, I think,
289 00:37:29.510 ⇒ 00:37:57.640 Brian Yang: I think there’s definitely a market for like basically hitting rest Apis as a service, because Api’s are so easy for us now. But like, when you’re just starting, you’re like, Oh, shit! I need. All I know how to do is like download. Csv is a rest. Api. What does that even mean like? And if you’re just like writing a wrapper for a rest. Api. Then, like, I’m I’m sure there’s people that got off Snowflake that need to hit rest Apis, that that people are just like terrified of them. So you’re just basically selling that as a service. Right? Yeah.
290 00:37:58.300 ⇒ 00:37:59.810 Brian Yang: yeah, I like it.
291 00:38:00.790 ⇒ 00:38:12.790 Uttam Kumaran: So let me. I’ll I’ll send it to you when it’s when it’s done. I’m trying to hopefully get that done in the next like few weeks. But I’ll send you a video that Hubspot Connector, because I was like holy shit. I didn’t realize you could kind of do this.
292 00:38:13.050 ⇒ 00:38:20.649 Uttam Kumaran: And then dude. I went to a snowflake like user group thing the other night and I saw them use all their new AI functions.
293 00:38:20.770 ⇒ 00:38:26.379 Uttam Kumaran: It’s pretty thick. basically, you can put in like you want to use mistrole, or you want to use
294 00:38:26.570 ⇒ 00:38:35.429 Uttam Kumaran: like Facebook llama. And then you can run summarize. You can ask it questions. You can embed text. So you won’t need to call open AI.
295 00:38:35.610 ⇒ 00:38:38.260 Uttam Kumaran: You can run it. You’ll be able to run it all on Snowflake.
296 00:38:38.750 ⇒ 00:38:45.450 Brian Yang: Oh, so yeah, that’s that’s what I think there’s a like local Lms, I think, is the next next big thing, because.
297 00:38:45.490 ⇒ 00:38:47.330 Brian Yang: like what I’m doing in like
298 00:38:47.520 ⇒ 00:38:57.949 Uttam Kumaran: sending data to Openai is like, that’s not feasible like, yeah, so that’s exactly it. These guys are like, we’re gonna run it right on your instance. And we don’t see any of it.
299 00:38:58.490 ⇒ 00:39:01.510 Brian Yang: Yeah, yeah, I think there’s a big market for that, too.
300 00:39:02.030 ⇒ 00:39:06.799 Uttam Kumaran: Yeah, it was sick. I was like holy fuck. They they wouldn’t give me access, though, unless I’m on a
301 00:39:07.180 ⇒ 00:39:10.619 Uttam Kumaran: annual contract, but dude, if you call them, they’ll give you access. Bro.
302 00:39:11.070 ⇒ 00:39:19.730 Brian Yang: We we were trying to get access to like their new like Llmvot. I don’t remember what’s called like. I have a preview on this on the We work account right now.
303 00:39:19.980 ⇒ 00:39:26.530 Uttam Kumaran: They told me they would give me access if I sign someone for a 12 K contract. You guys can easily get access.
304 00:39:27.340 ⇒ 00:39:28.729 Brian Yang: Oh, shit. Yeah.
305 00:39:29.420 ⇒ 00:39:35.260 Uttam Kumaran: Tell whoever’s talking to like? Ask again. Cause? I asked once, and they were like, All right.
306 00:39:36.420 ⇒ 00:39:43.320 Brian Yang: Well, well, basically, Steve Rollin is the he’s not gonna ask.
307 00:39:43.440 ⇒ 00:39:51.970 Uttam Kumaran: No, he’s he’s just incompetent at everything. So take take over that because you guys can. You don’t have to write all this bullshit right now.
308 00:39:52.990 ⇒ 00:40:06.619 Uttam Kumaran: Yeah. But then I’ll be out of a job. So I I’d rather just say this. I keep forgetting. I keep forgetting the real game like y’all you remember y’all, Lou?
309 00:40:07.180 ⇒ 00:40:24.119 Brian Yang: Yeah. Well, y’all, Bill, all of this like terrible R. Back stuff, and me as not knowing, thought he was just incompetent. I think he’s just made it overly complicated for job security, and now I have it, and no one understands the R. Back that he don’t accept me. And now I’m just like, Oh, yeah, thanks yong.
310 00:40:24.520 ⇒ 00:40:29.979 Uttam Kumaran: I can’t believe it, dude. I’m mistake like years ago, is like still living.
311 00:40:30.110 ⇒ 00:40:37.010 Brian Yang: You want to say that. How many central developers central deployer still granting
312 00:40:37.270 ⇒ 00:40:42.199 Uttam Kumaran: it’s like the same roles, but the same permissions. So maybe one thing different. And you’re like.
313 00:40:42.360 ⇒ 00:40:44.459 Uttam Kumaran: we’re super secure. You’re like, both.
314 00:40:44.840 ⇒ 00:40:49.360 Uttam Kumaran: Nobody cares about the room booking data. Anyways.
315 00:40:50.990 ⇒ 00:40:52.300 Brian Yang: Data. Yeah.
316 00:40:52.350 ⇒ 00:41:09.099 Uttam Kumaran: I actually, I actually got a call from this company called Property Vista. They’re a property management like software company, and they need a bunch of snowflake help. I’m actually planning on having a follow up call with them. But it’s really similar to we work like they have like
317 00:41:09.150 ⇒ 00:41:22.020 Uttam Kumaran: occupancy, maintenance tickets, surveys. I was like, yeah, I pretty much know, exactly like what you guys need. They’re like, we just have so many objects. And like, we don’t know all the ids. And we’re trying to build, build an activity stream. And I’m like.
318 00:41:22.220 ⇒ 00:41:24.470 Uttam Kumaran: Oh, easy sauce.
319 00:41:24.950 ⇒ 00:41:26.160 Brian Yang: nice.
320 00:41:26.820 ⇒ 00:41:36.769 Uttam Kumaran: nice. It’s a lot of data engineering. That’s a lot of.
321 00:41:37.190 ⇒ 00:41:43.639 Uttam Kumaran: And so we can hopefully, just like. really flex how much we did it. We work and try to get that contract.
322 00:41:43.950 ⇒ 00:41:44.770 Brian Yang: Yeah.
323 00:41:45.620 ⇒ 00:41:57.069 Brian Yang: when I when I get this like data like running your scheduler on your like running airflow or Dvt on your on your snowflake instance, like Ui thing up like
324 00:41:57.630 ⇒ 00:42:07.910 Uttam Kumaran: would love to have you guys try it on on some works. And if it’s actually any use, imagine if you could do all this like without using Dvt, you know, all synthetic native. Right?
325 00:42:08.340 ⇒ 00:42:11.319 Uttam Kumaran: No, I mean it would. It would be great. I
326 00:42:11.580 ⇒ 00:42:22.150 Uttam Kumaran: I basically, I’m using Dvc, so bare bones like, I have a couple of macros. but like I only used it because the orchestration is really nice. I don’t do anything else on there.
327 00:42:23.390 ⇒ 00:42:28.120 Brian Yang: The only thing is like, if you can have it version control through Github.
328 00:42:28.350 ⇒ 00:42:41.020 Uttam Kumaran: And then somehow, like it has like runners where it could take to pull it down the Github code and do and do deployments and things like that like that’s that’s it. But again, like I don’t, I don’t really care either way.
329 00:42:41.240 ⇒ 00:42:46.930 Uttam Kumaran: I would again. I want everything on snowflake as possible, because I want to sell snowflake contracts.
330 00:42:46.980 ⇒ 00:42:52.719 and I wanna only buy them one software I can. I don’t need DBT. For anything. I don’t use DBT. Cloud at all.
331 00:42:53.530 ⇒ 00:42:56.460 Brian Yang: Yeah. And I’m I’m basically like trying to build
332 00:42:56.690 ⇒ 00:43:12.260 Brian Yang: your your documentation and your scheduling and your test like Dvt plus like Allen or their Dvc Docs, basically on Snowflake native, like the the data store on Snowflake. And I’m just giving you an interface to make it look good. So it looks like an actual like data platform.
333 00:43:12.450 ⇒ 00:43:15.410 Check out this company called SQL Mash. Have you heard of these guys?
334 00:43:16.400 ⇒ 00:43:19.689 Brian Yang: There was some chatter about it.
335 00:43:19.900 ⇒ 00:43:21.820 Brian Yang: There’s some stuff.
336 00:43:21.840 ⇒ 00:43:25.229 Uttam Kumaran: It’s kind of like an open source. Dbt, I think. But
337 00:43:25.380 ⇒ 00:43:27.509 Uttam Kumaran: maybe it might be helpful.
338 00:43:31.560 ⇒ 00:43:32.360 Brian Yang: Hmm.
339 00:43:35.510 ⇒ 00:43:37.149 Brian Yang: data transformation.
340 00:43:44.200 ⇒ 00:43:51.459 Oh, yeah, this definitely, this has like transformation type stuff, too. Yeah, I think it’s like a, it’s open source, right? Is it?
341 00:43:54.110 ⇒ 00:44:06.050 Uttam Kumaran: I was talking to Mike about it. And he was like, these guys pretty much. You can just use use them as a Dvt. But
342 00:44:06.350 ⇒ 00:44:08.659 Uttam Kumaran: I don’t have fucking time for that right now. But
343 00:44:10.280 ⇒ 00:44:11.350 Brian Yang: yeah.
344 00:44:20.310 ⇒ 00:44:23.310 Uttam Kumaran: okay, I gotta run. I just got to this.
345 00:44:26.370 ⇒ 00:44:31.749 Brian Yang: alright sounds good. I’ll talk to you later. Then.
346 00:44:32.190 ⇒ 00:44:34.239 Uttam Kumaran: alright, dude! I’ll send you a loom with this stuff.
347 00:44:34.340 ⇒ 00:44:36.279 Let me know. You think I’ll send you on slack.
348 00:44:36.980 ⇒ 00:44:40.700 Brian Yang: Yeah, sounds good. Alright. I’ll see you later.