Meeting Title: Lilo Data Warehouse Discussion Date: 2026-01-21 Meeting participants: Casie Aviles, Pranav Narahari, Samuel Roberts, Awaish Kumar
WEBVTT
1 00:04:17.480 ⇒ 00:04:19.259 Samuel Roberts: Hey, guys, sorry.
2 00:04:22.390 ⇒ 00:04:24.540 Samuel Roberts: Bouncing from meeting to meeting just now.
3 00:04:26.040 ⇒ 00:04:26.960 Pranav Narahari: Hey, all good.
4 00:04:28.520 ⇒ 00:04:29.869 Samuel Roberts: How are we doing today?
5 00:04:31.480 ⇒ 00:04:32.789 Pranav Narahari: Pretty good.
6 00:04:36.850 ⇒ 00:04:39.959 Samuel Roberts: Did… awake… did not accept this meeting.
7 00:04:40.400 ⇒ 00:04:41.500 Samuel Roberts: Okay.
8 00:04:48.410 ⇒ 00:04:49.570 Samuel Roberts: Absolutely.
9 00:04:57.390 ⇒ 00:05:00.660 Samuel Roberts: Did he even acknowledge when I said this before? No, he didn’t. Okay.
10 00:05:01.470 ⇒ 00:05:07.120 Samuel Roberts: Alright, well, I’ll see if he joins, let me get the link.
11 00:05:08.070 ⇒ 00:05:13.570 Samuel Roberts: Is there a way to… Yeah, there.
12 00:05:26.910 ⇒ 00:05:36.739 Samuel Roberts: Okay. While it’s the three of us, I guess let’s talk about just general, current stuff.
13 00:05:37.090 ⇒ 00:05:42.330 Samuel Roberts: So, I think, Pranav, how is the reports?
14 00:05:43.910 ⇒ 00:05:49.310 Pranav Narahari: Yeah, so, haven’t been making progress on reports today, because I’ve just been waiting for Bobby, so I’ve just been working on.
15 00:05:49.310 ⇒ 00:05:50.440 Samuel Roberts: Oh, okay.
16 00:05:50.440 ⇒ 00:05:51.010 Pranav Narahari: Uff.
17 00:05:51.780 ⇒ 00:05:52.879 Samuel Roberts: Hello, what’s up? Sorry.
18 00:05:53.480 ⇒ 00:05:56.739 Pranav Narahari: the… just the UI stuff for, Phase 2.
19 00:05:56.950 ⇒ 00:05:57.640 Pranav Narahari: Okay.
20 00:05:57.640 ⇒ 00:05:58.480 Samuel Roberts: Cool, cool.
21 00:05:59.150 ⇒ 00:06:00.070 Pranav Narahari: So…
22 00:06:00.450 ⇒ 00:06:09.490 Pranav Narahari: That’s fine, you know? We still have a lot of other stuff to do. It seems like that Shopify data ingestion thing you already got, right?
23 00:06:10.330 ⇒ 00:06:12.399 Samuel Roberts: Yeah, so that was pretty…
24 00:06:12.650 ⇒ 00:06:23.649 Samuel Roberts: clean once I tweaked it with cursor and got that file out. I was able to load it… I had some issues running a second Postgres database on my machine, so it wouldn’t let me connect.
25 00:06:23.850 ⇒ 00:06:30.730 Samuel Roberts: So I just dumped it into a SQLite database, and ran some SQL on it, and got some cohort analysis, and it all seemed pretty good.
26 00:06:30.910 ⇒ 00:06:31.560 Samuel Roberts: Nice.
27 00:06:32.960 ⇒ 00:06:46.300 Samuel Roberts: I actually had a whole conversation with Cursor about Data Warehouse, and, like, explained, like, we’re gonna be ingesting more things, what do we want to do, kind of, you know, what we were hoping to talk with Awish about here.
28 00:06:46.670 ⇒ 00:06:51.680 Samuel Roberts: Let me see if I can find what it said, now that I’ve jumped around to, like, 3 other conversations later.
29 00:06:52.090 ⇒ 00:06:55.920 Samuel Roberts: Yeah, so…
30 00:06:59.990 ⇒ 00:07:04.370 Samuel Roberts: Part of the issue here is that I just don’t know how much data we’re gonna be getting.
31 00:07:07.090 ⇒ 00:07:10.280 Pranav Narahari: With, like, the last two… like, is there not two years’ worth of data?
32 00:07:10.840 ⇒ 00:07:15.109 Samuel Roberts: No, no, that is 2 years’ worth of data. I’m talking about, like, storage-wise. Like, are we talking…
33 00:07:15.300 ⇒ 00:07:19.869 Samuel Roberts: How many, you know what I mean? Like, what are we talking in terms of, like, gigabytes of storage?
34 00:07:20.830 ⇒ 00:07:21.629 Samuel Roberts: How many connections?
35 00:07:21.630 ⇒ 00:07:22.340 Pranav Narahari: Oh, I agree.
36 00:07:22.340 ⇒ 00:07:22.830 Samuel Roberts: The other…
37 00:07:22.830 ⇒ 00:07:24.440 Pranav Narahari: For the data warehouse.
38 00:07:24.840 ⇒ 00:07:27.780 Samuel Roberts: Yes, yes, yes, sorry, so I had a whole conversation about, like.
39 00:07:27.900 ⇒ 00:07:30.449 Samuel Roberts: after I got the Shopify thing working, I was like.
40 00:07:30.950 ⇒ 00:07:34.400 Samuel Roberts: the cursor… I wish… can I share these conversations on you?
41 00:07:34.600 ⇒ 00:07:36.990 Samuel Roberts: Probably export it, but…
42 00:07:38.750 ⇒ 00:07:52.939 Samuel Roberts: What did I say here? Yeah, I was trying to set up the warehouse. It went through, like, BigQuery, Snowflake, just using Postgres, DuckDB, Mother Duck for DuckDB, and then whether or not loading it on our own, or doing something like AirByte.
43 00:07:53.090 ⇒ 00:07:56.080 Samuel Roberts: Which is what they’ve talked about.
44 00:07:56.560 ⇒ 00:07:57.650 Pranav Narahari: Huh.
45 00:07:59.010 ⇒ 00:08:09.180 Samuel Roberts: Its suggestions weren’t bad. I mean, part of it was just, like, you’re using Postgres already, just set up another Postgres database and dump the data there, and when you need a bigger warehouse, move to a bigger warehouse.
46 00:08:09.730 ⇒ 00:08:15.339 Samuel Roberts: But… I would say…
47 00:08:15.500 ⇒ 00:08:22.760 Samuel Roberts: when it came to the ETL stuff, it was also like, yeah, we can do that, and I’m like, yeah, you’re always saying we can do whatever, and then sometimes things don’t work, so, like…
48 00:08:23.020 ⇒ 00:08:29.819 Samuel Roberts: But it did specifically say, like, when there’s more than, like, 10 connections, it’s definitely worth going with something else. So I think…
49 00:08:31.390 ⇒ 00:08:35.340 Samuel Roberts: There’s an argument to be made for just dumping this data
50 00:08:35.710 ⇒ 00:08:39.619 Samuel Roberts: I just don’t know, scaling-wise, if that’s gonna make the most sense.
51 00:08:40.150 ⇒ 00:08:44.130 Samuel Roberts: Thumbs up there, Casey.
52 00:08:46.010 ⇒ 00:08:48.480 Casie Aviles: Oh, yeah, yeah, I was just listening,
53 00:08:49.480 ⇒ 00:08:53.259 Casie Aviles: I think, yeah, if there’s not gonna be a lot of connections, then I think…
54 00:08:53.360 ⇒ 00:08:57.089 Casie Aviles: some… a tool like Polyatomic won’t make sense.
55 00:08:57.270 ⇒ 00:09:00.140 Casie Aviles: And… Yeah.
56 00:09:00.960 ⇒ 00:09:03.620 Samuel Roberts: So I think the path forward…
57 00:09:03.980 ⇒ 00:09:09.679 Samuel Roberts: In the immediate future, would be just, like, You know, we’re hitting…
58 00:09:10.500 ⇒ 00:09:15.490 Samuel Roberts: potentially 4 APIs to dump a little bit of data. Okay. Awish, welcome.
59 00:09:16.540 ⇒ 00:09:17.409 Samuel Roberts: How are you?
60 00:09:18.010 ⇒ 00:09:19.359 Awaish Kumar: I’m good, how about you?
61 00:09:19.670 ⇒ 00:09:26.999 Samuel Roberts: Good, good. We were just talking through some of this stuff. We sort of wanted to get your input on some of this.
62 00:09:27.270 ⇒ 00:09:32.860 Samuel Roberts: Before we kind of make some decisions moving forward for Leela.
63 00:09:33.600 ⇒ 00:09:34.790 Awaish Kumar: Okay.
64 00:09:35.450 ⇒ 00:09:42.029 Samuel Roberts: So, I don’t know if you saw the message, pronounced sent, but it was basically, like, right now we’re looking at Shopify data.
65 00:09:42.410 ⇒ 00:09:47.410 Samuel Roberts: To build this forecasting dashboard that they want.
66 00:09:48.100 ⇒ 00:09:56.930 Samuel Roberts: Eventually, we’ll be adding… meta ads, we’ll be adding Google Ads and, Flavio for email.
67 00:09:56.930 ⇒ 00:09:57.880 Awaish Kumar: Okay, yup.
68 00:09:58.250 ⇒ 00:10:09.680 Samuel Roberts: And so, those are, like, the main four things. We’re gonna be doing this for… right now, for one brand, but the Lilo supports a bunch of different brands. I don’t remember how many. Do you guys remember how many they mentioned would be in?
69 00:10:09.680 ⇒ 00:10:13.520 Pranav Narahari: 70 plus right now, but that’s gonna scale up this year.
70 00:10:14.290 ⇒ 00:10:16.310 Awaish Kumar: Definitely, yeah, that’s definitely their plan.
71 00:10:16.580 ⇒ 00:10:19.560 Awaish Kumar: Okay, do we know the volume of the data?
72 00:10:19.860 ⇒ 00:10:31.200 Samuel Roberts: Well, that’s what we’re… that’s what we were just discussing, like, I don’t necessarily know if I have a… like, we’re… they were looking for… for this proof of concept, it was a two-year… two years back in Shopify for all these… for this… this brand.
73 00:10:31.900 ⇒ 00:10:37.320 Awaish Kumar: Are they selling… okay, if they’re selling on Shopify, do you know the…
74 00:10:38.230 ⇒ 00:10:40.360 Awaish Kumar: Kind of revenue they are generating.
75 00:10:41.590 ⇒ 00:10:48.619 Samuel Roberts: I’m sure it’s different brand to brand, so that’s the other side of this, is it could be some small, some bigger, I don’t have a great sense yet.
76 00:10:48.620 ⇒ 00:10:49.410 Awaish Kumar: This…
77 00:10:49.410 ⇒ 00:10:53.520 Pranav Narahari: Adds any additional context, like, there are, like, you know…
78 00:10:53.630 ⇒ 00:10:59.859 Pranav Narahari: thousands of dollars happening per day, so, like, it’s not… it’s, like, something that we probably need to be updating, like.
79 00:11:00.010 ⇒ 00:11:03.160 Pranav Narahari: Either on, like, an hourly basis, or at least on a daily basis.
80 00:11:03.160 ⇒ 00:11:05.239 Samuel Roberts: I think, yeah, a daily basis is probably what we’re looking at.
81 00:11:05.240 ⇒ 00:11:06.660 Awaish Kumar: Okay, this might,
82 00:11:07.230 ⇒ 00:11:16.279 Awaish Kumar: Okay, but what I… okay, so Lilo is actually an e-commerce platform selling multiple brands, or…
83 00:11:16.750 ⇒ 00:11:23.849 Samuel Roberts: No, no, we… they are a, like an ad agency, basically. They run ads, and so…
84 00:11:23.950 ⇒ 00:11:29.990 Samuel Roberts: For all these different brands. So, these brands are on Shopify, and then they’re advertising on Google and Meta.
85 00:11:30.120 ⇒ 00:11:30.800 Samuel Roberts: And then…
86 00:11:30.800 ⇒ 00:11:32.459 Awaish Kumar: Okay. Not at that.
87 00:11:33.140 ⇒ 00:11:39.550 Awaish Kumar: So we will be ingesting data for Lido’s grinds, for example, right?
88 00:11:39.550 ⇒ 00:11:40.540 Samuel Roberts: Correct, yep.
89 00:11:41.230 ⇒ 00:11:47.140 Awaish Kumar: Okay, if Nero has 10 customers, we might have to ingest data for all those
90 00:11:47.510 ⇒ 00:11:50.530 Awaish Kumar: 10 customers, okay, got it.
91 00:11:50.630 ⇒ 00:12:05.380 Awaish Kumar: That… okay, that means that… that we need a data warehouse which can support the existing data load, and also can continue to support if they add multiple brands into their portfolio.
92 00:12:05.750 ⇒ 00:12:07.899 Samuel Roberts: Yes, up to 70, potentially.
93 00:12:08.700 ⇒ 00:12:10.040 Awaish Kumar: Okay.
94 00:12:10.040 ⇒ 00:12:10.610 Samuel Roberts: drawing.
95 00:12:11.500 ⇒ 00:12:19.870 Awaish Kumar: Okay, and how… how they are with the… selecting the tool? For example, like…
96 00:12:20.520 ⇒ 00:12:25.819 Awaish Kumar: They are tight on budgets, or very open with budgets? How that is.
97 00:12:27.660 ⇒ 00:12:37.309 Samuel Roberts: They don’t seem super tight with the budget. You know, they’re paying us a decent amount to build this, so I think they want probably the right thing, but I also don’t want something that’s, like.
98 00:12:37.560 ⇒ 00:12:40.410 Samuel Roberts: Overkill for what we’re doing.
99 00:12:40.580 ⇒ 00:12:47.800 Samuel Roberts: But I don’t know, they’ve talked about AirByte a bunch for the ETL, I don’t know your experience with that.
100 00:12:48.040 ⇒ 00:12:51.570 Awaish Kumar: to be honest, like, if I talk about data warehouse.
101 00:12:51.800 ⇒ 00:13:09.880 Awaish Kumar: if I… like, there… like, all the data warehouses kind of can support same things, right? Yeah. Like, all the big thing, big data warehouses, like Snowflake, Google, BigQuery, Redshift, Mother Duck, they… they all can support,
102 00:13:10.020 ⇒ 00:13:23.059 Awaish Kumar: and all have same… same features, but it just depends on pricing. Some more… some are flexible with month-to-month, some are more like con… like, ERB contracts.
103 00:13:24.150 ⇒ 00:13:33.380 Awaish Kumar: Some are, like, expensive, some are inexpensive, but having said that, Snowflake is…
104 00:13:33.700 ⇒ 00:13:36.960 Awaish Kumar: It’s hot in the market, and is the leading…
105 00:13:37.780 ⇒ 00:13:42.250 Awaish Kumar: Is leading this area for data warehousing.
106 00:13:42.510 ⇒ 00:13:43.180 Awaish Kumar: Michigan.
107 00:13:43.180 ⇒ 00:13:48.369 Pranav Narahari: Another thing that I think we should factor in is that they have, like.
108 00:13:48.810 ⇒ 00:13:59.420 Pranav Narahari: To kind of set, like, a pretty tight timeline, and so probably a system that won’t take a long time to implement, based on your experience, would be what we should probably maximize for.
109 00:14:00.300 ⇒ 00:14:18.140 Awaish Kumar: Yeah, yeah, that’s… that’s completely okay, like, we already are working with… we have worked with, I think, more than 30 clients, in… in… in Brainforge, and doing, like, setting up Snowflake for them. So, that’s… that’s not a problem.
110 00:14:18.470 ⇒ 00:14:29.220 Awaish Kumar: In any ways, we… for data warehouses, Snowflake would be the best choice, because in terms… like, I’m talking about multiple
111 00:14:29.730 ⇒ 00:14:32.210 Awaish Kumar: Things, like, in terms of scalability.
112 00:14:32.360 ⇒ 00:14:52.330 Awaish Kumar: it’s really easy to scale, right? If we… if they are… they acquire more brands, we need more capacity, it will just… we can just spin up more warehouses, and it will work. In terms of, segregation between the brands, like, if we more need security, we can have, like.
113 00:14:52.330 ⇒ 00:14:58.910 Awaish Kumar: Different roles, which can spin different warehouses and different warehouses for different brands.
114 00:14:58.960 ⇒ 00:15:02.750 Awaish Kumar: In such a way that those don’t interact with each other.
115 00:15:02.750 ⇒ 00:15:06.730 Samuel Roberts: Maybe another question of mine. Yeah, like, as they add brands, will we
116 00:15:07.070 ⇒ 00:15:10.989 Samuel Roberts: To automate that process of, like, the new ingestion, or is that something that…
117 00:15:11.410 ⇒ 00:15:14.949 Samuel Roberts: you know, we’d have to set up a new Snowflake every time, or something.
118 00:15:14.950 ⇒ 00:15:16.810 Awaish Kumar: Hello,
119 00:15:17.490 ⇒ 00:15:26.040 Awaish Kumar: We don’t. Basically, we… we can do the same Snowflake instance, right? We can… but, like, what, like, we…
120 00:15:26.590 ⇒ 00:15:44.839 Awaish Kumar: Polytomic can load it for you using same warehouse, like, okay, in the Snowflake warehouse, there’s a concept of compute, and actually, that is also called virtual warehouse. I’m talking about that. So, for example… for example, if they want to monitor their
121 00:15:44.980 ⇒ 00:15:56.230 Awaish Kumar: warehouse pricing for individual brand. We can separate the compute for each individual brand, like…
122 00:15:56.340 ⇒ 00:16:03.260 Awaish Kumar: The data we are processing for 10 brands, and how much we are spending on each brand for processing data.
123 00:16:03.380 ⇒ 00:16:10.680 Awaish Kumar: the… the data of each brand. So basically… and also in terms of
124 00:16:10.920 ⇒ 00:16:26.520 Awaish Kumar: in terms of, like, security, like, brands also need, like, okay, we don’t want our data to be in the same place as others, right? So, like, that ensures that, like, their compute is different, Snowflake handles storage differently, we can put it in a different database.
125 00:16:26.600 ⇒ 00:16:31.000 Awaish Kumar: Which is named with their own line name, or whatever, brand name.
126 00:16:31.220 ⇒ 00:16:37.000 Awaish Kumar: That can easily happen, we don’t need more snowflake instances, just one.
127 00:16:37.310 ⇒ 00:16:42.619 Awaish Kumar: We do need to… we do need to set up different connectors in ETL, either.
128 00:16:42.620 ⇒ 00:16:43.140 Samuel Roberts: Like.
129 00:16:43.140 ⇒ 00:16:51.600 Awaish Kumar: invite, or whatever. I… I, like, we… we are a good fan of… Polyatomic, as a… Right.
130 00:16:51.600 ⇒ 00:16:52.080 Samuel Roberts: Right.
131 00:16:52.080 ⇒ 00:17:07.740 Awaish Kumar: As a company, everybody uses… most of our clients use Polytomic, and there are a few reasons for that. Number one, we have direct, Slack comms with them, right? Easy to get them their support if something breaks. Number two.
132 00:17:07.880 ⇒ 00:17:18.139 Awaish Kumar: they are… they are extremely, you can say, inexpensive. Then, for example, Fivetrend, right? Sure.
133 00:17:18.579 ⇒ 00:17:23.639 Awaish Kumar: For that reason, also, they qualify for it. And…
134 00:17:24.209 ⇒ 00:17:33.429 Awaish Kumar: we do need to, like, set up different connectors to different Shopify accounts, but that will be in Paratomic, and basically…
135 00:17:33.539 ⇒ 00:17:48.809 Awaish Kumar: We can… the destination could be the same Snowflake account, that’s okay. Having said that, we are having some issues with Polyatomic right now, and for that reason, I don’t, like…
136 00:17:49.019 ⇒ 00:17:54.989 Awaish Kumar: Right now, I can’t tell, like, directly recommend that, like, I’m in talks with them, like, Annutham are both.
137 00:17:55.269 ⇒ 00:17:59.089 Awaish Kumar: Talking to them on that issue, and that issue is, like, syncing the data.
138 00:17:59.249 ⇒ 00:18:12.339 Awaish Kumar: So, basically, I don’t know the data volume right now for individual brands, but if, for example, they are working for a $30 million business, which is having… which is going, for example.
139 00:18:14.639 ⇒ 00:18:25.279 Awaish Kumar: maybe 5,000, 10,000 orders per day, then, like, and if Qualtomic is not able to sync it every day, it can cause delays.
140 00:18:25.549 ⇒ 00:18:33.709 Awaish Kumar: And then it can cause delays in reporting, and hence, maybe we don’t have the recent, latest data for
141 00:18:33.989 ⇒ 00:18:36.329 Awaish Kumar: For client, right? For, for reporting.
142 00:18:36.330 ⇒ 00:18:37.230 Samuel Roberts: Right.
143 00:18:37.400 ⇒ 00:18:45.020 Awaish Kumar: Right? And we have been facing this issue for a few clients, like, like, our data is, like,
144 00:18:45.220 ⇒ 00:18:52.299 Awaish Kumar: 15… one to two weeks behind, or sometimes we don’t have historical data.
145 00:18:54.780 ⇒ 00:19:05.269 Awaish Kumar: But, like, we are hopeful that they will listen to us and resolve it, because I and Nutam are basically working on that, like, and all of our clients are using their platform.
146 00:19:05.410 ⇒ 00:19:25.259 Awaish Kumar: So, we don’t want to lose that relationship as well. So, yeah, hopefully that… I will have an answer for that, but if things go well there, then Polyatomic would be the best choice, because we all know how to just set it up. Otherwise, the other options are, like.
147 00:19:25.760 ⇒ 00:19:38.859 Awaish Kumar: AirByte is one of them, because it’s an open source version, we can use that, but I’m not sure if you want to use open source, or… and deploy it ourselves, or do we want to consider any managed
148 00:19:39.270 ⇒ 00:19:40.620 Awaish Kumar: Virgin Self.
149 00:19:41.230 ⇒ 00:19:42.300 Awaish Kumar: Invite.
150 00:19:43.080 ⇒ 00:19:43.810 Samuel Roberts: Okay.
151 00:19:44.330 ⇒ 00:19:48.130 Samuel Roberts: So then if we… so say Polytomic works out.
152 00:19:48.390 ⇒ 00:19:51.519 Samuel Roberts: Are we able to add those connectors.
153 00:19:52.560 ⇒ 00:19:54.139 Awaish Kumar: Yeah, we are ready to come.
154 00:19:55.410 ⇒ 00:19:59.559 Awaish Kumar: And why do we need to do that programmatically, or…
155 00:19:59.970 ⇒ 00:20:02.879 Samuel Roberts: If they add new brands, my thought is, like, how did.
156 00:20:02.880 ⇒ 00:20:04.350 Awaish Kumar: Who’s working.
157 00:20:04.940 ⇒ 00:20:06.280 Awaish Kumar: How frequent is that?
158 00:20:06.620 ⇒ 00:20:11.440 Samuel Roberts: Well, I mean, they’re gonna spin up 70, and I imagine they’re gonna want more pretty, like, as they sign an alliance.
159 00:20:12.380 ⇒ 00:20:17.900 Awaish Kumar: I understand that, but my question is, like… like, number one, there’s no…
160 00:20:18.180 ⇒ 00:20:21.549 Awaish Kumar: Polyatomic has RESTful APIs. You can basically do.
161 00:20:21.550 ⇒ 00:20:23.329 Samuel Roberts: Okay, that’s what I’m wondering, that’s what I’m hoping, okay.
162 00:20:24.010 ⇒ 00:20:39.610 Awaish Kumar: That’s one other thing, but even if that works or not, like, if it is just one brand added in a week, like, we can just go in and do 5… in 5 minutes, we can add it there. So, in both cases, it should just work fine.
163 00:20:40.820 ⇒ 00:20:41.390 Samuel Roberts: Okay.
164 00:20:43.000 ⇒ 00:20:44.480 Samuel Roberts: Okay, that’s good to know.
165 00:20:44.770 ⇒ 00:20:47.060 Awaish Kumar: But they have RESTful APIs, right?
166 00:20:47.410 ⇒ 00:21:05.319 Awaish Kumar: And, in terms of cost for retail tool, for Polyatomic, I think they have to sign a contract, right? I think they can do both, month-to-month and yearly.
167 00:21:05.580 ⇒ 00:21:24.880 Awaish Kumar: And, it’s cheaper than 510, and it is around, I think… but roughly, it is around, like, $500, for around 6 million rows, in a month. That’s the… that’s the minimum, like, package they sell.
168 00:21:25.940 ⇒ 00:21:26.600 Samuel Roberts: Okay.
169 00:21:27.200 ⇒ 00:21:28.420 Awaish Kumar: Yep. Okay.
170 00:21:30.000 ⇒ 00:21:30.730 Samuel Roberts: September.
171 00:21:33.510 ⇒ 00:21:46.270 Samuel Roberts: Okay, yeah, so before you joined, I was mentioning that I had chatted with Cursor a bunch about trying to figure some of this stuff out, and giving me some back and forths, and I had been able to pull the Shopify data with a script, just to get it out. I think I posted it there, but…
172 00:21:46.480 ⇒ 00:21:48.390 Samuel Roberts: in the channel, I think…
173 00:21:48.980 ⇒ 00:21:55.979 Samuel Roberts: I’m… we’re trying to figure out, like, the difference between, you know, getting some stuff quickly, and then long-term as well.
174 00:21:57.160 ⇒ 00:22:03.880 Awaish Kumar: Yeah, even if you want it quickly, like, you can obviously use polyatomic right there. That’s the best way…
175 00:22:04.770 ⇒ 00:22:10.500 Awaish Kumar: That’s really quick, right? So, we have our internal instance. If you need, like.
176 00:22:10.850 ⇒ 00:22:15.210 Awaish Kumar: One-time dump of something, you can use that.
177 00:22:15.730 ⇒ 00:22:26.289 Awaish Kumar: Right? Okay. We have our internal instance, we… if you want to… yeah. You can use that for one-time dumps, but yeah, we don’t, schedule anything, all of our internal instances.
178 00:22:27.110 ⇒ 00:22:28.220 Samuel Roberts: Got it, okay.
179 00:22:29.160 ⇒ 00:22:30.299 Samuel Roberts: Okay, this is how…
180 00:22:30.300 ⇒ 00:22:43.639 Awaish Kumar: Once they have… once you show them, we can get this data through Polyatomic, if the client is okay, we move on to trial, right? Trial is kind of free, like, for a month or so.
181 00:22:43.820 ⇒ 00:22:44.790 Awaish Kumar: Also.
182 00:22:47.130 ⇒ 00:22:47.890 Samuel Roberts: Okay.
183 00:22:50.100 ⇒ 00:22:54.449 Samuel Roberts: Okay. Do you guys have any other questions and stuff while we’re… while we’ve got away?
184 00:22:57.370 ⇒ 00:23:00.400 Pranav Narahari: No, I think that was… Most of what I had.
185 00:23:01.780 ⇒ 00:23:03.670 Pranav Narahari: No other questions.
186 00:23:04.280 ⇒ 00:23:09.360 Awaish Kumar: I don’t know, why… why are you considering why? Like,
187 00:23:10.410 ⇒ 00:23:13.210 Awaish Kumar: Is that because it’s open source, or…
188 00:23:13.440 ⇒ 00:23:14.070 Awaish Kumar: Right.
189 00:23:14.070 ⇒ 00:23:15.479 Samuel Roberts: For everybody, is that what you’re saying?
190 00:23:16.330 ⇒ 00:23:17.460 Awaish Kumar: Yep.
191 00:23:17.740 ⇒ 00:23:20.649 Samuel Roberts: Yeah, I mean, the client is a little,
192 00:23:21.680 ⇒ 00:23:27.630 Samuel Roberts: like, they, they specifically asked about AirByte, And so I wasn’t sure…
193 00:23:27.950 ⇒ 00:23:41.379 Samuel Roberts: I’m not super well-versed in the ETL space, so I was just trying to figure out, is that, you know, is it something they want and we’re just gonna have to use it, or is it a good thing to use, or is it… I think they know other people using it, and so I wanted to make sure.
194 00:23:42.250 ⇒ 00:23:45.559 Awaish Kumar: Yeah, like, Airbuyed is kind of a…
195 00:23:46.090 ⇒ 00:24:03.470 Awaish Kumar: I don’t know if AirByte introduced Early Cloud recently, but, like, it’s an open source version. Yeah. We can add new connectors if we want, like, if they are not able to build it, we can just build it ourselves, right? There are a few flexibilities of having open source.
196 00:24:03.470 ⇒ 00:24:15.019 Awaish Kumar: But then you have to manage it, right? Right, right. You have to have an instance up and running for… for your air byte, and if it fails, you have to maintain it, so these are a few…
197 00:24:15.020 ⇒ 00:24:15.920 Samuel Roberts: Very fair.
198 00:24:15.920 ⇒ 00:24:17.379 Awaish Kumar: Pros and cons of that.
199 00:24:18.000 ⇒ 00:24:34.770 Samuel Roberts: Okay. The other question I guess I’m wondering is, if I am just gonna get some data, like Shopify right now, like, and put it into something like Postgres, just so I can… we can start building this… this UI with the data underneath it,
200 00:24:34.970 ⇒ 00:24:39.540 Samuel Roberts: How… you know, quickly do we need to move to something like Snowflake, you think?
201 00:24:41.010 ⇒ 00:24:46.569 Samuel Roberts: like, what… I guess my question is, like, what is Postgres, like, where’s the limit of that, kinda?
202 00:24:47.620 ⇒ 00:24:59.140 Awaish Kumar: Okay, so, which worst case you want to use, like, the one you’re already using, like, is it an operational, like, system, or it will be a net new
203 00:24:59.370 ⇒ 00:25:00.150 Awaish Kumar: Sarah.
204 00:25:00.150 ⇒ 00:25:11.149 Samuel Roberts: I would set up a new server on… I mean, we’re using the Postgres database on Railway right now, but my thought was to just set up a new one, just to have the data separate from the app data.
205 00:25:12.550 ⇒ 00:25:15.160 Awaish Kumar: Okay, so…
206 00:25:15.420 ⇒ 00:25:28.250 Awaish Kumar: with that, there’s only one, like, there’s one problem when the data volume grows. Like, for example, you might not feed it until you have maybe more than…
207 00:25:28.980 ⇒ 00:25:46.730 Awaish Kumar: 10, 20 terabytes of data, and only… and then the bigger problem comes in, when there are a lot of, users, right? Also, like, if you… if you connect, like, if you have, like.
208 00:25:47.090 ⇒ 00:26:07.060 Awaish Kumar: 10, 20 active sessions, everybody’s running queries, there are systems connected to it, right? Which are… some systems are reading, some are writing, and then it locks tables, right? Then you feel the delay in things, right? A user can see, okay, I’m querying it, it takes me, like, 20 minutes of…
209 00:26:07.060 ⇒ 00:26:09.050 Awaish Kumar: And I’m not able to get my…
210 00:26:09.100 ⇒ 00:26:17.150 Awaish Kumar: data or things like that. You will notice that if, like, there will be a lot of different sessions and users.
211 00:26:19.020 ⇒ 00:26:20.070 Awaish Kumar: Okay.
212 00:26:20.690 ⇒ 00:26:24.310 Awaish Kumar: But I think Snowflake is also, like, a really…
213 00:26:25.270 ⇒ 00:26:28.939 Awaish Kumar: like, it’s really quick to set up, right? We don’t need…
214 00:26:28.940 ⇒ 00:26:34.780 Samuel Roberts: That’s the other thing I was gonna say, is it even worth it? But, okay. Yeah, so just Snowflake to start is probably the way to go.
215 00:26:35.580 ⇒ 00:26:44.700 Awaish Kumar: Yeah, like… Like, we can set it up in one day, so… Right, whatever you need.
216 00:26:45.420 ⇒ 00:26:50.769 Samuel Roberts: Sounds good. Alright, yeah, I think… I think that definitely helps clarify some things here, and definitely gives us some,
217 00:26:51.600 ⇒ 00:26:52.630 Samuel Roberts: Coming back to them.
218 00:26:52.630 ⇒ 00:27:00.949 Awaish Kumar: Also, in terms of pricing, I think cloud… cloud databases are more expensive than data warehouses.
219 00:27:02.030 ⇒ 00:27:02.740 Samuel Roberts: Okay.
220 00:27:02.970 ⇒ 00:27:05.579 Samuel Roberts: No, like, on a per day-to-life kind of basis.
221 00:27:05.750 ⇒ 00:27:07.029 Samuel Roberts: Yeah, okay.
222 00:27:10.010 ⇒ 00:27:15.700 Samuel Roberts: Okay, cool. I think… I feel pretty good about maybe…
223 00:27:16.260 ⇒ 00:27:26.000 Samuel Roberts: maybe looking at airbutt a little bit, looking at polyatomic a little bit, but I am a little nervous about the freshness issue you’re talking about with polyatomic, because they’re definitely going to want this, like, on a daily basis.
224 00:27:26.000 ⇒ 00:27:37.910 Awaish Kumar: Yeah, we are… we are already, like, we are ourselves, like, really, concerned about that, and we are on top of it, right? Like, we don’t want to introduce it
225 00:27:38.040 ⇒ 00:27:51.450 Awaish Kumar: again, to, like, see the client calling for it, like, every day, okay, why is our not… our data is not there. So, yeah, we are looking into it. Hopefully, it will be resolved. Otherwise.
226 00:27:51.570 ⇒ 00:27:55.970 Awaish Kumar: We can have a, like, decision on, like… but,
227 00:27:56.420 ⇒ 00:28:02.320 Awaish Kumar: Like, you can start with digging data, like, it will be… like, trial versions are really…
228 00:28:02.320 ⇒ 00:28:03.120 Samuel Roberts: Yeah.
229 00:28:03.120 ⇒ 00:28:08.889 Awaish Kumar: kind of free, right? If it doesn’t work out, we can just say, okay, it’s not working for us, that’s all.
230 00:28:09.560 ⇒ 00:28:15.170 Samuel Roberts: Sure. Okay, that sounds good. Yeah, so I guess we can bring this back to them and… Sorry?
231 00:28:15.930 ⇒ 00:28:27.229 Pranav Narahari: Sorry, yeah, I was just gonna say one thing on the data freshness front. Definitely, like, we want… we don’t want to have some of those issues, but I’m wondering how,
232 00:28:27.300 ⇒ 00:28:42.029 Pranav Narahari: important it is, and how, like… I think the reporting tool that we create will still be pretty useful, because they’re going to be looking on a timeframe of 6 months to 1 year of, like, data, and so if you’re not having the data for the last 2 days, it should be…
233 00:28:42.130 ⇒ 00:28:55.980 Pranav Narahari: okay. Let’s probably just be transparent about, like, okay, with Polytonic, like, this is an issue that, you know, one of our team members has had, like, for another client, and then…
234 00:28:56.090 ⇒ 00:28:56.970 Pranav Narahari: We can just bring.
235 00:28:56.970 ⇒ 00:29:05.500 Awaish Kumar: Yeah, we don’t… we don’t… no, no, we don’t want to… we don’t want to, like, call that out in a… in front of a client, right? I’m just.
236 00:29:05.500 ⇒ 00:29:06.540 Samuel Roberts: Yeah, no, I would…
237 00:29:06.700 ⇒ 00:29:12.380 Samuel Roberts: Yeah, I wouldn’t mention that yet. We’re looking at those tools and stuff, but I wouldn’t say…
238 00:29:12.650 ⇒ 00:29:13.880 Samuel Roberts: do that yet.
239 00:29:14.000 ⇒ 00:29:16.479 Pranav Narahari: Yeah, but, yeah, maybe we can just…
240 00:29:16.940 ⇒ 00:29:24.790 Pranav Narahari: still factor in polytonic, if… I mean, it’s up to you, but, I don’t think it’ll be as big of an issue, for us, at least.
241 00:29:25.270 ⇒ 00:29:27.899 Samuel Roberts: Well, for that part, but there’s also the, like, daily…
242 00:29:28.580 ⇒ 00:29:30.710 Samuel Roberts: Forecasting tracker and stuff that they’re gonna want.
243 00:29:32.140 ⇒ 00:29:37.990 Pranav Narahari: So, are we just not going to have, like, a regular database anymore? Everything’s just going to be in the data warehouse?
244 00:29:41.210 ⇒ 00:29:45.700 Samuel Roberts: Well, the… the… this data would be in the data warehouse. The…
245 00:29:46.330 ⇒ 00:29:48.830 Samuel Roberts: the Postgres, like, the app data would be.
246 00:29:49.320 ⇒ 00:29:50.490 Samuel Roberts: Postgres still.
247 00:29:52.470 ⇒ 00:30:02.929 Pranav Narahari: Okay, right. So, reports, I guess, would still… would come from the data warehouse. So I guess, yeah, then we definitely need very fresh data. Okay, so then, yeah.
248 00:30:02.930 ⇒ 00:30:16.369 Samuel Roberts: Yeah, no, I think for now, that’s what Uten was mentioning before, like, doing the calculations on the fly for the reports, I think, is kind of fine for now, but eventually we’ll be able to do more powerful stuff when we have, like, the data, we’re not hitting the API for Shopify and Klaviyo and Meta every time.
249 00:30:16.900 ⇒ 00:30:21.490 Pranav Narahari: Okay, yeah, then I totally agree, yeah. Data freshness is actually super important then, because, like.
250 00:30:21.490 ⇒ 00:30:21.940 Samuel Roberts: Yeah.
251 00:30:22.490 ⇒ 00:30:24.540 Pranav Narahari: Gonna be a daily thing, and needs to have…
252 00:30:24.540 ⇒ 00:30:26.000 Awaish Kumar: Yesterday’s data.
253 00:30:26.240 ⇒ 00:30:27.700 Samuel Roberts: Exactly, exactly.
254 00:30:28.550 ⇒ 00:30:44.819 Awaish Kumar: So, ideally… ideally, I can recommend Polyatomic Snowflake as a tech stack, right, for building a data warehouse, and then dbt on top of it for doing your transformation, right? And…
255 00:30:44.820 ⇒ 00:30:52.269 Awaish Kumar: Then, finally, the… you can use polyatomic as well for reverse ATL. So, if you want to put data somewhere
256 00:30:52.670 ⇒ 00:31:01.269 Awaish Kumar: again, in a sheet, Google Sheets, or into the app, or if you want to have a BI tool on top of it for reporting. But…
257 00:31:01.640 ⇒ 00:31:16.280 Awaish Kumar: Having said that, as a caveat, I just mentioned we are having… currently having some talks with Polyatomic, and hopefully… I’m not sure when do you need a decision, but we should have something by end of week.
258 00:31:17.650 ⇒ 00:31:25.730 Samuel Roberts: Okay, yeah, I think that’s probably all right. We can… we can wait till then, because we’re making progress just with the, like, little bits of data we’re getting to start for this proof of concept.
259 00:31:25.880 ⇒ 00:31:27.870 Samuel Roberts: I’m not too re- worried about that.
260 00:31:28.480 ⇒ 00:31:34.799 Awaish Kumar: Yeah, and once, like, once we discuss that with the polyatomic, we are going to see how
261 00:31:35.270 ⇒ 00:31:43.049 Awaish Kumar: things are going to work with Vatomic, otherwise we… we can explore other options, like, in the next week.
262 00:31:44.180 ⇒ 00:31:45.480 Samuel Roberts: Okay, this one’s good.
263 00:31:45.480 ⇒ 00:31:46.010 Awaish Kumar: cap.
264 00:31:46.500 ⇒ 00:31:47.290 Awaish Kumar: Okay.
265 00:31:47.930 ⇒ 00:31:48.470 Samuel Roberts: Great.
266 00:31:48.930 ⇒ 00:31:50.969 Samuel Roberts: Any other thoughts, guys?
267 00:31:51.170 ⇒ 00:31:52.610 Samuel Roberts: Any other questions?
268 00:31:55.270 ⇒ 00:31:57.450 Casie Aviles: I think that was pretty clear, you know.
269 00:31:57.450 ⇒ 00:32:00.110 Samuel Roberts: Yeah. Alright, great. Thank you very much, Awish.
270 00:32:00.690 ⇒ 00:32:02.850 Awaish Kumar: Yep, thanks a lot.
271 00:32:03.270 ⇒ 00:32:03.950 Awaish Kumar: Right.
272 00:32:04.690 ⇒ 00:32:05.260 Samuel Roberts: Bye.
273 00:32:05.790 ⇒ 00:32:13.019 Samuel Roberts: Do you guys want to hang on real quick and just digest that? Just, so I’m… I’m thinking, yeah, the polyatomic thing definitely…
274 00:32:13.250 ⇒ 00:32:23.979 Samuel Roberts: we… we can… yeah, we’re not gonna make that decision immediately. Wait and see if they sort that out, because they could, I suppose, go either way. The air bite thing, though, they… they definitely seem kind of on the…
275 00:32:24.690 ⇒ 00:32:26.980 Samuel Roberts: side of AirByte,
276 00:32:28.520 ⇒ 00:32:32.959 Samuel Roberts: So it might be worth just investigating that a little bit. I don’t know… Casey, have you used AirByte at all?
277 00:32:33.660 ⇒ 00:32:42.189 Casie Aviles: No, but it’s one of the things I actually considered when we were, you know, doing some similar work for AI.
278 00:32:42.410 ⇒ 00:32:45.299 Casie Aviles: Where we wanted to, like, consolidate data.
279 00:32:45.550 ⇒ 00:32:49.180 Casie Aviles: For knowledge, but yeah, I haven’t really gone too much into it.
280 00:32:49.770 ⇒ 00:32:57.839 Samuel Roberts: Okay. Yeah, because they do have a cloud version, I didn’t… I don’t know if I remember the pricing on hand, but there’s also a self-hosted version.
281 00:32:58.520 ⇒ 00:33:05.230 Samuel Roberts: Which, you know, is not… You know, you have to manage the infrastructure, but the app itself.
282 00:33:05.230 ⇒ 00:33:05.740 Casie Aviles: Oh, this is still.
283 00:33:05.740 ⇒ 00:33:08.730 Samuel Roberts: Open source connectors and stuff, so…
284 00:33:09.170 ⇒ 00:33:13.380 Samuel Roberts: Okay, trying to think what else, then. So, Snowflake…
285 00:33:13.560 ⇒ 00:33:16.210 Samuel Roberts: Might be what we want to recommend, I don’t know.
286 00:33:16.680 ⇒ 00:33:20.679 Pranav Narahari: I just don’t have a great sense of the volume of data we’re really talking here over time.
287 00:33:25.220 ⇒ 00:33:28.710 Samuel Roberts: That’s what I’m.
288 00:33:28.930 ⇒ 00:33:34.620 Pranav Narahari: I think, since we don’t have, like, we kind of have a… We don’t have enough…
289 00:33:34.990 ⇒ 00:33:39.139 Pranav Narahari: information on, like, the… the scale of the data.
290 00:33:39.760 ⇒ 00:33:42.320 Pranav Narahari: I don’t think that’s really even a question we could, like.
291 00:33:42.480 ⇒ 00:33:45.140 Pranav Narahari: ask Bobby and Zach either, because, like…
292 00:33:45.140 ⇒ 00:33:49.540 Samuel Roberts: Yeah, I think I’ll probably have to do a little… a dry run on some of this exporting, see?
293 00:33:50.070 ⇒ 00:33:59.420 Pranav Narahari: Yeah, but also, just based on what Awish said, like, we could spin up, like, these things, like, pretty quickly. I think we kind of maximize for the thing that we can spin up quickly, and then also is it gonna…
294 00:33:59.490 ⇒ 00:34:10.970 Pranav Narahari: like, have any issues for us. And then in the future, if we need to, like, pivot to, like, another tool, it sounds like it wouldn’t be a big deal.
295 00:34:11.940 ⇒ 00:34:20.599 Pranav Narahari: But I think for right now, like, Snowflake sounds like it’ll do the trick. It’ll probably do the trick even for, like, the foreseeable future.
296 00:34:20.600 ⇒ 00:34:32.540 Pranav Narahari: But I don’t think it’s, like, worth doing, like, the research right now. I think it’ll slow us down a lot, and I don’t even think we’ll feel super confident even after that about, like, did we choose the right tool? I think it’s kind of like…
297 00:34:32.540 ⇒ 00:34:33.079 Samuel Roberts: Yeah, you’re right.
298 00:34:33.080 ⇒ 00:34:43.930 Pranav Narahari: Kind of what, like, Utam said about, like, another thing that we were working with, like, let’s just kind of, like, choose a tool. I think when we were choosing between, like, Cloud Code, Cursor, or… and,
299 00:34:44.239 ⇒ 00:34:50.140 Pranav Narahari: in, like, codecs. I think these are all, like, great tools, so, like, let’s just choose one, and then…
300 00:34:50.400 ⇒ 00:34:53.789 Pranav Narahari: That we feel confident with, and then just, like, ride it out.
301 00:34:55.400 ⇒ 00:34:58.019 Samuel Roberts: Yeah, I think that’s probably fair.
302 00:34:58.450 ⇒ 00:35:01.659 Pranav Narahari: Especially for this, like, circumstance, where it’s like…
303 00:35:01.830 ⇒ 00:35:08.249 Pranav Narahari: you know, there are gonna be some unknowns, I just don’t want that to, like, block us for a long time, us ask them a ton.
304 00:35:08.250 ⇒ 00:35:15.240 Samuel Roberts: No, I think you’re right. I just wanted to make sure we weren’t going, like, crazy overkill for anything here, you know, like… Oh, okay. Some of these tools are for, like, you know.
305 00:35:15.890 ⇒ 00:35:19.679 Samuel Roberts: kind of intense, like, data analysis, and what we’re doing is not…
306 00:35:19.920 ⇒ 00:35:30.080 Samuel Roberts: You know, we’re getting all the data out and doing a little bit of stuff on top of it, but really, it’s not like… they’re not going to be writing to the database, they’re not going to be making queries, we’re kind of doing that all for them in the middle, you know?
307 00:35:30.590 ⇒ 00:35:31.080 Pranav Narahari: Yeah.
308 00:35:31.580 ⇒ 00:35:39.439 Samuel Roberts: So I don’t think it’s… it’s… I just want to make sure we’re not, like, going, like, yeah, pay, you know, a few thousand dollars for this tool that we’re gonna use, like, a tiny portion of.
309 00:35:40.430 ⇒ 00:35:45.760 Pranav Narahari: Okay, yeah. I guess with that time… with the money part of it, too, yeah.
310 00:35:46.040 ⇒ 00:35:56.450 Samuel Roberts: Yeah, I mean, I don’t know exactly what their budget is in terms of, like, the warehousing side. That might be something Tom may have a better sense of, but I mean, they’re dropping a decent amount of money for us to do this quickly, so I think, like.
311 00:35:56.980 ⇒ 00:35:59.269 Samuel Roberts: They’re gonna want it in a way that is…
312 00:35:59.790 ⇒ 00:36:05.660 Samuel Roberts: the right way, but obviously you don’t want to end up giving them on the hook for, like, a giant bill ongoing for…
313 00:36:05.800 ⇒ 00:36:08.200 Samuel Roberts: Stuff they’re not using fully.
314 00:36:08.550 ⇒ 00:36:11.790 Samuel Roberts: But I’m just looking at Snowflake’s pricing calculator here.
315 00:36:11.980 ⇒ 00:36:15.669 Samuel Roberts: I think we’re probably pretty good.
316 00:36:15.780 ⇒ 00:36:19.670 Samuel Roberts: I just don’t know what their… Like…
317 00:36:21.580 ⇒ 00:36:24.309 Samuel Roberts: Is it monthly? Is it, you know, all that sort of stuff?
318 00:36:26.360 ⇒ 00:36:26.980 Pranav Narahari: Yep.
319 00:36:27.200 ⇒ 00:36:29.019 Samuel Roberts: I think it’s probably fine.
320 00:36:31.480 ⇒ 00:36:37.259 Samuel Roberts: But, yeah, the ETL, I think we gotta kind of wait on that then and see, but it might be worth even just trying
321 00:36:37.770 ⇒ 00:36:39.739 Samuel Roberts: I don’t know what Airbite Cloud is.
322 00:36:39.950 ⇒ 00:36:41.220 Samuel Roberts: Price-wise.
323 00:36:42.330 ⇒ 00:36:44.770 Samuel Roberts: Prices, plans…
324 00:36:46.850 ⇒ 00:36:51.710 Samuel Roberts: they don’t even have prices on there. Of course they don’t. I hate websites with a pricing page that’s just like…
325 00:36:52.550 ⇒ 00:36:54.940 Samuel Roberts: Call us to find out.
326 00:36:59.400 ⇒ 00:37:00.280 Pranav Narahari: Oh, really?
327 00:37:00.730 ⇒ 00:37:03.850 Samuel Roberts: Yeah, it just says, like, contact us, or get started, or…
328 00:37:06.080 ⇒ 00:37:09.459 Samuel Roberts: I guess I could log in and sign. Alright, maybe I’ll play around with that.
329 00:37:15.610 ⇒ 00:37:21.670 Samuel Roberts: But I think more importantly is getting some of the, like, Google Concept stuff rolling more.
330 00:37:23.380 ⇒ 00:37:27.600 Samuel Roberts: I’m just trying to think of the best way to…
331 00:37:27.910 ⇒ 00:37:30.300 Samuel Roberts: to this now, so I think…
332 00:37:32.790 ⇒ 00:37:34.960 Samuel Roberts: You were working on the UI, you said, a little bit?
333 00:37:35.510 ⇒ 00:37:36.190 Pranav Narahari: Yeah.
334 00:37:36.800 ⇒ 00:37:37.550 Samuel Roberts: Okay.
335 00:37:38.420 ⇒ 00:37:40.609 Samuel Roberts: I’m trying to think of the best way to, like… Go ahead.
336 00:37:41.180 ⇒ 00:37:46.209 Pranav Narahari: I’ve just been working on that in the meantime, so I was planning on finalizing the reporting stuff today, but…
337 00:37:46.390 ⇒ 00:37:51.169 Pranav Narahari: Since, well, actually, no, it looks like Bobby just responded a little bit ago.
338 00:37:51.910 ⇒ 00:37:52.590 Samuel Roberts: Okay.
339 00:37:52.880 ⇒ 00:37:53.600 Pranav Narahari: Yeah.
340 00:37:54.640 ⇒ 00:38:00.330 Pranav Narahari: Okay, but yeah, I was planning on working on that. I’ll probably… since he responded, I’ll probably start working on the reporting stuff again.
341 00:38:00.860 ⇒ 00:38:02.929 Samuel Roberts: Okay, cool. Is there anything, like.
342 00:38:03.230 ⇒ 00:38:07.330 Samuel Roberts: worth sharing there, since I have the data, is there anything to, like, plug in?
343 00:38:08.570 ⇒ 00:38:10.279 Samuel Roberts: Or is it still… For reporting?
344 00:38:10.650 ⇒ 00:38:14.340 Samuel Roberts: No, no, no, for the, forecasting.
345 00:38:16.560 ⇒ 00:38:18.480 Pranav Narahari: for the forecasting UI.
346 00:38:18.890 ⇒ 00:38:20.829 Samuel Roberts: Yeah, is there anything worth pushing it, or…
347 00:38:21.560 ⇒ 00:38:26.920 Samuel Roberts: Okay, that’s… I was making sure. So I was like, if I have this data, I’m wondering… I’m trying to figure out the best way to, like.
348 00:38:28.260 ⇒ 00:38:35.630 Samuel Roberts: connect them real quick, but I don’t think it’s necessary right now, so let’s not stress about it. Okay. Yeah, focus on the reports, let’s get that out. Oh, go ahead.
349 00:38:36.160 ⇒ 00:38:41.560 Pranav Narahari: they had that, like, collab notebook, right? So I can… honestly, from there, I can probably…
350 00:38:41.920 ⇒ 00:38:46.259 Pranav Narahari: Handle, like, running that script against, like, the data, and then, like…
351 00:38:46.260 ⇒ 00:38:50.020 Samuel Roberts: Yeah, yeah, I posted the data, right? So you’re good to go with that to start.
352 00:38:50.930 ⇒ 00:38:52.090 Samuel Roberts: Okay. Perfect.
353 00:38:52.660 ⇒ 00:38:55.770 Samuel Roberts: Yeah, yeah, run with that. That’s good.
354 00:38:57.470 ⇒ 00:39:04.889 Samuel Roberts: Cool. And then, yeah, but finish the report so we can get that all into prod. I think they posted there’s, like, another bug with Klaviyo we gotta dig into just now, so…
355 00:39:06.010 ⇒ 00:39:06.750 Pranav Narahari: Gotcha.
356 00:39:07.260 ⇒ 00:39:08.559 Samuel Roberts: I’ll…
357 00:39:11.390 ⇒ 00:39:15.890 Samuel Roberts: wait for this, I get to get ingested into the platform and just get some quick notes from it.
358 00:39:16.100 ⇒ 00:39:21.249 Samuel Roberts: And just run some stuff by UTM in terms of, like, Snowflake, airbyte, polyatomic stuff.
359 00:39:23.860 ⇒ 00:39:26.559 Samuel Roberts: And then we can update them, too.
360 00:39:27.690 ⇒ 00:39:31.360 Samuel Roberts: I don’t know what we update them with, but,
361 00:39:31.530 ⇒ 00:39:33.829 Samuel Roberts: Let them know we got some options.
362 00:39:34.210 ⇒ 00:39:38.480 Samuel Roberts: Maybe get some nice sense of their pricing… Sensitivity.
363 00:39:39.400 ⇒ 00:39:40.659 Pranav Narahari: Yeah, I think that’s a good tip.
364 00:39:40.660 ⇒ 00:39:41.629 Samuel Roberts: I like it.
365 00:39:41.630 ⇒ 00:39:43.630 Pranav Narahari: To the end of day.
366 00:39:43.630 ⇒ 00:39:43.960 Samuel Roberts: Yeah.
367 00:39:43.960 ⇒ 00:39:46.090 Casie Aviles: We can say, you know, we’ve kind of, like.
368 00:39:46.190 ⇒ 00:39:53.399 Pranav Narahari: done a ton of research on the options for data warehousing, based on, like, our requirements. We’ve talked to, like.
369 00:39:53.740 ⇒ 00:40:02.970 Pranav Narahari: like, you know, other people within our company that have built similar solutions. And then, yeah, you can ask them about, like, pricing sensitivity,
370 00:40:02.970 ⇒ 00:40:03.580 Samuel Roberts: Yeah.
371 00:40:03.890 ⇒ 00:40:10.600 Pranav Narahari: Yeah, I think that’s the main question. And then also, we can just have some, like, high-level, like, figures about, like, how many…
372 00:40:10.730 ⇒ 00:40:17.359 Pranav Narahari: I think we read somewhere that there’s 70 brands, is that still accurate as of right now? Do you see that, like.
373 00:40:17.610 ⇒ 00:40:20.549 Pranav Narahari: Changing in the near future.
374 00:40:20.550 ⇒ 00:40:21.190 Samuel Roberts: Yeah.
375 00:40:22.900 ⇒ 00:40:29.639 Pranav Narahari: And, yeah, we basically are just trying to, like, assess from a high level, like, what are going to be, like, the data…
376 00:40:30.020 ⇒ 00:40:31.959 Pranav Narahari: the data volume, right? So…
377 00:40:31.960 ⇒ 00:40:32.630 Samuel Roberts: Yep.
378 00:40:32.800 ⇒ 00:40:35.949 Pranav Narahari: Yeah, any metrics that they can provide on that would be helpful.
379 00:40:36.940 ⇒ 00:40:37.560 Samuel Roberts: Okay.
380 00:40:39.050 ⇒ 00:40:40.439 Samuel Roberts: Alright, I think that sounds good.
381 00:40:42.890 ⇒ 00:40:44.170 Samuel Roberts: Any other thoughts?
382 00:40:46.560 ⇒ 00:40:47.720 Samuel Roberts: Alright, cool.
383 00:40:50.560 ⇒ 00:40:51.629 Samuel Roberts: I guess that’s it.
384 00:40:52.900 ⇒ 00:40:53.390 Pranav Narahari: Yep.
385 00:40:53.390 ⇒ 00:40:56.240 Samuel Roberts: Alright, so yeah, let me know when the reports are good,
386 00:40:56.780 ⇒ 00:40:59.210 Samuel Roberts: And then, hopefully we can merge in tomorrow morning.
387 00:41:00.060 ⇒ 00:41:00.620 Pranav Narahari: Yeah.
388 00:41:00.870 ⇒ 00:41:02.139 Pranav Narahari: Cool. Sounds good.
389 00:41:02.510 ⇒ 00:41:03.010 Samuel Roberts: Alrighty.
390 00:41:03.010 ⇒ 00:41:03.399 Pranav Narahari: Alright guys.
391 00:41:03.400 ⇒ 00:41:03.979 Samuel Roberts: Thank you, guys.
392 00:41:03.980 ⇒ 00:41:04.530 Pranav Narahari: Alexa.
393 00:41:05.880 ⇒ 00:41:06.279 Casie Aviles: Thank you.
394 00:41:06.280 ⇒ 00:41:07.750 Samuel Roberts: I… bye.