Meeting Title: Brainforge x LMNT : Bi Weekly Checkin with Tech Team Date: 2026-01-21 Meeting participants: Awaish Kumar, Jason Wu, Steve Sizer, Shivani Amar, Andy Weist, Uttam Kumaran
WEBVTT
1 00:00:08.390 ⇒ 00:00:09.160 Shivani Amar: Hi!
2 00:00:11.400 ⇒ 00:00:12.340 Jason Wu: Bishop money.
3 00:00:13.060 ⇒ 00:00:14.099 Shivani Amar: How you doing?
4 00:00:16.570 ⇒ 00:00:17.500 Jason Wu: Good.
5 00:00:19.710 ⇒ 00:00:20.589 Shivani Amar: How you doing, Steve?
6 00:00:20.880 ⇒ 00:00:22.679 Shivani Amar: Is it, like, snowing? I’m pretty good.
7 00:00:23.480 ⇒ 00:00:26.829 Steve Sizer: No, no, it’s just… it’s just the sunlight hitting that window, so…
8 00:00:26.830 ⇒ 00:00:30.330 Shivani Amar: The sunlight is making me think it’s snow.
9 00:00:30.330 ⇒ 00:00:34.970 Steve Sizer: Yeah, you can’t… you can’t see, no, it is, it is, bright sunshine, yeah.
10 00:00:35.350 ⇒ 00:00:36.030 Shivani Amar: It’s okay.
11 00:00:36.030 ⇒ 00:00:41.039 Steve Sizer: It’s just the incredibly warm weather that’s in Arizona that’s making.
12 00:00:41.040 ⇒ 00:00:42.050 Jason Wu: I was really interested in practice.
13 00:00:42.050 ⇒ 00:00:43.850 Shivani Amar: going in Phoenix.
14 00:00:43.850 ⇒ 00:00:48.260 Steve Sizer: No, no. Definitely not.
15 00:00:48.260 ⇒ 00:00:49.300 Shivani Amar: Definitely not.
16 00:00:50.050 ⇒ 00:00:51.050 Shivani Amar: Oh, man.
17 00:00:51.050 ⇒ 00:00:51.530 Jason Wu: on details.
18 00:00:51.530 ⇒ 00:00:54.649 Shivani Amar: Hey, Awish, are you there?
19 00:00:54.650 ⇒ 00:00:56.090 Awaish Kumar: I don’t… yep.
20 00:00:56.980 ⇒ 00:00:58.420 Shivani Amar: Are you having internet issues?
21 00:00:58.950 ⇒ 00:01:05.200 Awaish Kumar: Yeah, I’m trying to… Log into it, so… Okay, just a few minutes.
22 00:01:06.060 ⇒ 00:01:07.139 Shivani Amar: Okay, gotcha.
23 00:01:07.610 ⇒ 00:01:11.130 Shivani Amar: Okay, I’ll be right back.
24 00:02:36.090 ⇒ 00:02:37.410 Awaish Kumar: Hi, yeah.
25 00:02:39.750 ⇒ 00:02:42.370 Awaish Kumar: I just, I think, I don’t know if…
26 00:02:44.890 ⇒ 00:02:45.269 Jason Wu: We cheat.
27 00:02:45.270 ⇒ 00:02:48.400 Awaish Kumar: Automoting, but yeah, we can start.
28 00:02:48.400 ⇒ 00:02:49.565 Jason Wu: Yep.
29 00:02:50.730 ⇒ 00:02:53.379 Awaish Kumar: Basically, the agenda…
30 00:02:54.290 ⇒ 00:03:02.749 Awaish Kumar: But today was, like, to show you the… the model we have built, and, like, how the… how we have structured it.
31 00:03:02.900 ⇒ 00:03:07.860 Awaish Kumar: And, like, the… Showing you, basically, the end-to-end flow.
32 00:03:10.150 ⇒ 00:03:10.970 Jason Wu: Okay.
33 00:03:11.560 ⇒ 00:03:16.339 Awaish Kumar: Yeah, I’ll switch to the laptop now, and then I’ll share my screen.
34 00:03:35.220 ⇒ 00:03:36.160 Awaish Kumar: Yep.
35 00:03:38.030 ⇒ 00:03:39.610 Awaish Kumar: Yes.
36 00:04:02.050 ⇒ 00:04:09.160 Awaish Kumar: So yeah, like, I think in the last meeting, I already shared the… Polyatomic?
37 00:04:12.420 ⇒ 00:04:14.219 Awaish Kumar: So, like, this…
38 00:04:15.180 ⇒ 00:04:27.709 Awaish Kumar: So, like, there are a few changes we made is… are, like, mostly… so we already had Shopify, we already had Recharge, then we had, like, for wholesale, then we identified to…
39 00:04:27.820 ⇒ 00:04:33.379 Awaish Kumar: new Google Sheets, they were using for their reporting.
40 00:04:33.710 ⇒ 00:04:45.179 Awaish Kumar: So one is, like, the… the CRM. They are… they are managing, which is in the Google Sheets, and then there are some, manual, pull.
41 00:04:45.440 ⇒ 00:04:57.349 Awaish Kumar: they have been doing, for some time, and they have come up with some transform mapping between customer ID and the company name.
42 00:04:58.000 ⇒ 00:04:59.579 Awaish Kumar: For wholesale partners.
43 00:04:59.970 ⇒ 00:05:02.219 Awaish Kumar: So, we tried to bring it in.
44 00:05:02.600 ⇒ 00:05:09.779 Awaish Kumar: And when we upload these, like, basically, if we go to the history.
45 00:05:10.390 ⇒ 00:05:15.990 Awaish Kumar: The configuration, we can see, like, all of the sheets which are available in this spreadsheet.
46 00:05:16.290 ⇒ 00:05:18.610 Awaish Kumar: And then the ones we need to select.
47 00:05:18.730 ⇒ 00:05:22.320 Awaish Kumar: It will basically just, create some tables in…
48 00:05:22.520 ⇒ 00:05:25.410 Awaish Kumar: and Snowflake in this, output schema.
49 00:05:26.600 ⇒ 00:05:34.890 Awaish Kumar: So, yeah, that’s basically… we… that’s how we ingested this data into Snowflake.
50 00:05:35.020 ⇒ 00:05:38.369 Awaish Kumar: Which lives in the raw database of the Snowflake.
51 00:05:39.090 ⇒ 00:05:44.139 Awaish Kumar: And… Then, if we move to the.
52 00:05:44.610 ⇒ 00:05:53.419 Andy Weist: Can we… just a real quick question on that. I believe you worked with Madison on that CRM data from Google Sheets. I’m just wondering…
53 00:05:54.040 ⇒ 00:06:09.979 Andy Weist: Because it’s in Google Sheets, and I know that format of the CRM has evolved frequently, meaning the sheet changes, the columns change, it hasn’t been very stable for very long, and we’re looking at possibly moving to a more normalized, sort of.
54 00:06:10.600 ⇒ 00:06:15.569 Andy Weist: Process or different system for the CRM.
55 00:06:17.680 ⇒ 00:06:26.280 Andy Weist: do you have guidance on, like, best practice here? I don’t know how your conversations with Madison went, as far as…
56 00:06:26.430 ⇒ 00:06:30.080 Andy Weist: Using that spreadsheet and the stability of it.
57 00:06:30.970 ⇒ 00:06:40.799 Andy Weist: Like, if 2 months from now, that spreadsheet wasn’t being used anymore, or if it changed significantly, would we have issues with
58 00:06:40.950 ⇒ 00:06:44.539 Andy Weist: what data we’re pulling in. Obviously, I would think that, like.
59 00:06:44.740 ⇒ 00:06:47.709 Andy Weist: If the sheet changed, if the… the…
60 00:06:48.270 ⇒ 00:06:53.260 Andy Weist: ID changed, or they moved data somewhere, obviously, like, Polytomic would have to be updated.
61 00:06:53.450 ⇒ 00:06:57.280 Andy Weist: Is there a plan there? Was that a discussion point at all?
62 00:06:58.270 ⇒ 00:07:03.969 Awaish Kumar: like, for CRM specifically, we, we haven’t discussed, like, if
63 00:07:04.130 ⇒ 00:07:08.039 Awaish Kumar: They want to use some kind of different tool, but that was…
64 00:07:08.580 ⇒ 00:07:28.299 Awaish Kumar: That was a discussion, in our early discussions with Shivani, that for CRM, we have to lose some kind of tool, I have to move that, but yeah, for now, we have been mainly focusing on whatever they have, and then they are basically spending, like, Madison kind of spending
65 00:07:29.920 ⇒ 00:07:35.910 Awaish Kumar: like, average, maybe I can, like, estimate 1 to 2 hours.
66 00:07:36.270 ⇒ 00:07:39.690 Awaish Kumar: Just in the Google Sheets for reporting.
67 00:07:39.910 ⇒ 00:07:46.069 Awaish Kumar: To understand, like, what partners have ordered what, and then somebody is
68 00:07:46.280 ⇒ 00:07:51.510 Awaish Kumar: is about to churn, or somebody, which has been…
69 00:07:51.570 ⇒ 00:08:02.229 Awaish Kumar: a custom, like, has been ordering in, like, maybe once a month, and they haven’t ordered for, like, 2 months or 3 months now. Like, these kind of analysis.
70 00:08:02.230 ⇒ 00:08:13.449 Awaish Kumar: And she has been doing that, manually, and for that, she didn’t have the place where to get the data from, so she goes into Shopify, downloads all the CSVs, and then
71 00:08:13.710 ⇒ 00:08:22.010 Awaish Kumar: she’s not able to even do that, because the sheets, like, have… like, one of the examples I can just show you here…
72 00:08:23.550 ⇒ 00:08:30.370 Awaish Kumar: for… that… So… yeah… What sheet was that?
73 00:08:30.600 ⇒ 00:08:33.130 Awaish Kumar: Yeah, so this is one of the sheets, right?
74 00:08:33.280 ⇒ 00:08:36.469 Awaish Kumar: So, you can see, we have here around…
75 00:08:39.000 ⇒ 00:08:47.100 Awaish Kumar: If I just remove the filters… Then we have prone group…
76 00:08:47.320 ⇒ 00:08:53.380 Awaish Kumar: 13,000 rows, right? And that is just for 50 partners in January.
77 00:08:53.920 ⇒ 00:09:11.899 Awaish Kumar: like, so she got 13,000 rows for 50 partners for the month of January, and if she just extends the period or include more partners, this data just grows exponentially, and then she, like, it is, like,
78 00:09:12.310 ⇒ 00:09:19.930 Awaish Kumar: exhausting to do any kind of analysis in Google Sheets. It just becomes slow, you are not able to apply formulas, filtering, all of that.
79 00:09:20.040 ⇒ 00:09:22.490 Awaish Kumar: So, it was not workable, so what…
80 00:09:22.670 ⇒ 00:09:32.200 Awaish Kumar: In this first phase, what we wanted, basically just to help them with the… with the analytical reporting part, so that,
81 00:09:34.480 ⇒ 00:09:45.140 Awaish Kumar: Basically, they are… basically, they… she has more time on making the decisions or identifying the patterns instead of doing these old Google Sheet kind of stuff.
82 00:09:46.300 ⇒ 00:09:51.460 Andy Weist: Okay, yeah, I think I understand the current state of their management of the spreadsheet.
83 00:09:51.890 ⇒ 00:09:59.060 Andy Weist: I guess my question is more towards… If we consider the, like.
84 00:09:59.610 ⇒ 00:10:04.389 Andy Weist: The fact that they will probably… Move away from spreadsheets?
85 00:10:04.730 ⇒ 00:10:12.020 Andy Weist: Or if they do, is it gonna be very hard to, like, reconfigure our setup with Polytomic?
86 00:10:12.290 ⇒ 00:10:13.169 Awaish Kumar: No, not…
87 00:10:13.170 ⇒ 00:10:14.610 Andy Weist: It’s a different data source.
88 00:10:14.610 ⇒ 00:10:34.379 Awaish Kumar: Not at all. Basically, what we are building is on top of the CRM data, and any CRM platform we are going to use in future, if we move away from Google Sheets, it is going to have… for example, if I can just show you what I have built so far.
89 00:10:36.720 ⇒ 00:10:40.650 Awaish Kumar: So… If we go into this,
90 00:10:42.310 ⇒ 00:10:47.839 Awaish Kumar: this sheet, right? So, this is the sheet we have built for her, and if you can see…
91 00:10:48.030 ⇒ 00:10:58.930 Awaish Kumar: basically, this is the information she needs, like, customer ID, email, name, phone number, company name, to identify the partner, and that is, like, every CRM tool is going to
92 00:10:59.120 ⇒ 00:11:01.459 Awaish Kumar: Give us that information.
93 00:11:01.610 ⇒ 00:11:08.940 Awaish Kumar: And then, some of the information is, like, this segmentation. Basically, that’s basically the logic in the Shopify.
94 00:11:09.270 ⇒ 00:11:11.810 Awaish Kumar: For each order, we have a tag.
95 00:11:12.090 ⇒ 00:11:15.529 Awaish Kumar: And through the tags, we are managing,
96 00:11:15.590 ⇒ 00:11:28.180 Awaish Kumar: what segment, in the wholesale it belongs? Like, some are specialty retail, some are online resellers, if someone is trusted health, like, in whatever category they are in.
97 00:11:28.190 ⇒ 00:11:37.850 Awaish Kumar: So that is being identified through tags. Similarly, like, if there is… if somebody has the refrigerator or not, that is going to be identified through Shopify data.
98 00:11:38.170 ⇒ 00:11:40.849 Awaish Kumar: And most of it is, like,
99 00:11:41.090 ⇒ 00:11:46.860 Awaish Kumar: Like, if total orders or, like, the… total revenue…
100 00:11:47.640 ⇒ 00:11:52.259 Awaish Kumar: last order ID, first order date, last order date, tags.
101 00:11:52.870 ⇒ 00:11:58.830 Awaish Kumar: So, our majority of these metrics, which they use to identify,
102 00:11:59.130 ⇒ 00:12:03.669 Awaish Kumar: the patterns are coming… is coming from Shopify.
103 00:12:03.890 ⇒ 00:12:12.550 Awaish Kumar: So, that is basically… Will be some changes on our modeling part, but yeah.
104 00:12:13.180 ⇒ 00:12:14.270 Awaish Kumar: Like, you know.
105 00:12:14.270 ⇒ 00:12:16.819 Andy Weist: Okay, I wasn’t aware of… is…
106 00:12:17.300 ⇒ 00:12:23.410 Andy Weist: Is this data going into Snowflake at all, or are we just pushing more data into Sheets?
107 00:12:23.410 ⇒ 00:12:28.340 Awaish Kumar: Everything is in Snowflake. This sheet is just a view for…
108 00:12:28.560 ⇒ 00:12:40.420 Awaish Kumar: wholesale team, if, like, maybe, like, if, like, Laura and Madison are not able to go into Snowflake and carry themselves, we are just building some sheets for them.
109 00:12:40.700 ⇒ 00:12:45.299 Jason Wu: Got it. And then, in the future, Awish, this would be…
110 00:12:45.950 ⇒ 00:12:47.490 Awaish Kumar: Yeah, in the future, it’s going to be…
111 00:12:47.490 ⇒ 00:12:48.629 Jason Wu: tool, right?
112 00:12:48.630 ⇒ 00:12:49.780 Awaish Kumar: Yeah, Objector.
113 00:12:50.800 ⇒ 00:12:51.260 Jason Wu: Okay.
114 00:12:51.260 ⇒ 00:12:52.550 Awaish Kumar: That’s right, that’s right, like.
115 00:12:52.550 ⇒ 00:13:06.450 Jason Wu: So all we’re doing here for this sheet, at least, is just demonstrating we’ve pulled all the data, aggregated it into Snowflake, and then this is just one result of it, right?
116 00:13:06.690 ⇒ 00:13:09.040 Awaish Kumar: So basically…
117 00:13:09.340 ⇒ 00:13:24.150 Awaish Kumar: like, if we take wholesale as our customer from a big element, a customer, if we just take one department wholesale as our customer, right? So what we are trying to help them is, whatever they are doing manually right now.
118 00:13:24.260 ⇒ 00:13:37.229 Awaish Kumar: If we can just take it and do it and automate it. So they don’t spend any time in doing any kind of reporting. Yeah, they will be doing analysis on top of it, but they will not going into doing it.
119 00:13:37.300 ⇒ 00:13:44.439 Awaish Kumar: And then it is going to end up in… finally, it is going to end up in the Omni, but for now, because wholesale is the…
120 00:13:44.580 ⇒ 00:13:54.729 Awaish Kumar: is the partner we identified. For them, we can actually have, you can say, quick turnaround time on setting up the… setting them up for success.
121 00:13:54.970 ⇒ 00:14:13.619 Awaish Kumar: for anyone else, like, for example, e-commerce. We don’t have Amazon data yet, we don’t have, for example, Walmart from, like, Walmart online data, like, e-commerce data for Walmart. So that is, like, a blocker for us to completely.
122 00:14:13.910 ⇒ 00:14:14.820 Jason Wu: I see.
123 00:14:15.170 ⇒ 00:14:22.250 Awaish Kumar: like, provide any kind of support for e-commerce, but we can do that for wholesale, and… Got it, I see.
124 00:14:23.710 ⇒ 00:14:27.680 Jason Wu: I… I know, I know Utsom’s not… Utsom’s not on this call right now, right?
125 00:14:27.960 ⇒ 00:14:28.630 Awaish Kumar: Bye.
126 00:14:28.630 ⇒ 00:14:36.500 Jason Wu: I think this is… like, to me, this is a good example of… The next step about connecting
127 00:14:36.800 ⇒ 00:14:41.099 Jason Wu: to, like, the UI. You know, so when we last saw the demo for Omni.
128 00:14:41.210 ⇒ 00:14:43.610 Jason Wu: I know kind of the takeaway out of that was.
129 00:14:43.730 ⇒ 00:14:45.990 Jason Wu: What are the next steps to…
130 00:14:46.210 ⇒ 00:14:51.110 Jason Wu: Either get set up with a trial, or start to kind of visualize it with, like, some real data.
131 00:14:51.670 ⇒ 00:15:00.319 Jason Wu: Versus what was shown before, and to me, this is a great example of that, right? So, if you’re saying we actually have, like, everything that we need to generate some
132 00:15:00.460 ⇒ 00:15:08.540 Jason Wu: some reporting for wholesale. Like, in my mind, this is a great opportunity to move them out of the spreadsheet
133 00:15:08.640 ⇒ 00:15:17.190 Jason Wu: Right? And then simply say, hey, look, your sheet, we use it as kind of a data ingest, and then start to, like, use this as the main case for the visualization part as well.
134 00:15:17.660 ⇒ 00:15:37.009 Awaish Kumar: Yeah, like, you’re right, that’s our ultimate plan, but, like, to, like, we have BI tool investigations, like, scheduled in the next month. Like, the setting up the trial version, and then starting with the basic
135 00:15:37.100 ⇒ 00:15:39.330 Awaish Kumar: Wholesale reporting and all of that.
136 00:15:40.530 ⇒ 00:15:55.290 Awaish Kumar: So, in the meantime, like, we still have to ingest some of the data, like, our… what’d you say, like, the… the time is going into all of these things, like ingest… like, solving the ingestion issues which we have, or the modeling…
137 00:15:55.500 ⇒ 00:16:04.640 Awaish Kumar: And then, understanding the metrics. So, this exercise is not just for… like, we identified a few things, like, if you see partner status.
138 00:16:05.200 ⇒ 00:16:13.120 Awaish Kumar: Like, this is the definition we just suggested, right? They didn’t have this kind of metric before.
139 00:16:13.400 ⇒ 00:16:24.820 Awaish Kumar: And we… we include it, but we don’t know, like, if it is going… it is worthwhile for them. So this is also one of the exercises through which we are identifying what are the good metrics for wholesale.
140 00:16:25.020 ⇒ 00:16:32.369 Awaish Kumar: And once we are all set up, it is… I think it is… it will be really quick to move it out of Google Sheets.
141 00:16:34.500 ⇒ 00:16:35.220 Jason Wu: Okay.
142 00:16:36.600 ⇒ 00:16:51.819 Awaish Kumar: Yeah, but we are on it, like, the Omni, I already… based on our conversation, I already started the call with Hutam, that, like, we need trial of the Omni, and I think he’s… he’s working on that, like, the Omni part.
143 00:16:51.820 ⇒ 00:16:57.369 Jason Wu: Right. Yeah, I just… I mean, just as a call-out, like, I think…
144 00:16:57.780 ⇒ 00:17:00.370 Jason Wu: What you’ve shown here is… is good.
145 00:17:00.600 ⇒ 00:17:06.310 Jason Wu: I think to make sure that we… Continue to, like, drive that…
146 00:17:06.500 ⇒ 00:17:08.350 Jason Wu: You know, what is, like, the…
147 00:17:08.750 ⇒ 00:17:19.369 Jason Wu: in the end, kind of the end result solution that we’re looking for, as far as, like, that self-service reporting. Like, to me, this is, like, the perfect use case, you know, for, like, you know, number one for, kind of, the UI.
148 00:17:20.770 ⇒ 00:17:22.260 Awaish Kumar: Yeah, totally, right.
149 00:17:22.569 ⇒ 00:17:29.309 Awaish Kumar: Like, that’s what we’ve built, and the next step is basically that, like.
150 00:17:29.330 ⇒ 00:17:43.619 Awaish Kumar: like, right now, we are just identifying some patterns, like, to… so this is… if you look at this, this is a really big table, and it is… it is going to end up only in Snowflake, like, it is…
151 00:17:43.790 ⇒ 00:17:51.530 Awaish Kumar: like, in the… in the… in the BI tool, I don’t think, like, it is going to be… going to live it… live there like… like this, like…
152 00:17:51.760 ⇒ 00:18:09.269 Awaish Kumar: in there, we might have some kind of these view, right? If you look at this view, this is more like a summary view, showing something for the partners, and we might create some charts out of it, some… it’s going to end up in being, like, some…
153 00:18:09.480 ⇒ 00:18:13.299 Awaish Kumar: Line charts, some, some, like,
154 00:18:14.040 ⇒ 00:18:16.270 Awaish Kumar: Bar charts and stuff like that.
155 00:18:16.490 ⇒ 00:18:20.489 Awaish Kumar: And that is basically, going to end up in Omni.
156 00:18:26.410 ⇒ 00:18:33.310 Awaish Kumar: And, like, these metrics through which they are going to identify the patterns and make some decisions.
157 00:18:36.880 ⇒ 00:18:43.279 Awaish Kumar: Yeah, if I just… yeah, if I go back to the GitHub…
158 00:18:43.840 ⇒ 00:18:54.009 Awaish Kumar: like, what’s going on behind that is… is that, like, if you can see that in the macros, we are… I’m trying to add any
159 00:18:54.890 ⇒ 00:19:05.510 Awaish Kumar: kind of business logic we are writing. It is just going to end up in these… one of these macros, so we standardize the definition across the organization.
160 00:19:05.640 ⇒ 00:19:14.630 Awaish Kumar: So, like, if this is just saying it’s wholesale, like, we are just identifying if a part… if a Shopify customer is a wholesale partner.
161 00:19:14.820 ⇒ 00:19:32.730 Awaish Kumar: And there’s a logic, like, in that, and, like, it is an instant drives one, and easy to manage, easy to change, and also easy to understand, like, for anyone to just go there, figure out, like, how we are assigning segments to different customers, how we are…
162 00:19:34.910 ⇒ 00:19:35.940 Awaish Kumar: signing.
163 00:19:36.230 ⇒ 00:19:41.110 Awaish Kumar: if it is a wholesale partner, and then also, there is some data from CRM, and then
164 00:19:41.220 ⇒ 00:19:52.090 Awaish Kumar: application data, and then there is data from Shopify, and we are not… we don’t have customer IDs link… or any ID linking these two data together, so we are trying to…
165 00:19:52.290 ⇒ 00:20:00.439 Awaish Kumar: Match it… match it via different, fields, like choosing email, or a company name, or addresses, or…
166 00:20:00.850 ⇒ 00:20:04.220 Awaish Kumar: Yeah, or alternative emails, or whatever is info.
167 00:20:04.510 ⇒ 00:20:06.950 Awaish Kumar: Identifiable information is there.
168 00:20:07.080 ⇒ 00:20:14.779 Awaish Kumar: So, that’s… basically, we… Standardize any logic, and put it in here.
169 00:20:15.950 ⇒ 00:20:29.920 Awaish Kumar: And then we have some models. In the models, it’s mainly, you can see I have added, like, wholesale customers as a new, folder, which is basically identifying,
170 00:20:30.230 ⇒ 00:20:37.929 Awaish Kumar: a mod for us. Like, this wholesale customer is a separate mod. If you go in there, we have wholesale DIM customers.
171 00:20:38.530 ⇒ 00:20:55.539 Awaish Kumar: So, what’s going to happen at the end is, right now, like, this is the kind of logic, you can say, the SQL which is running behind building this dream customer table, which I just showed in the Google Sheet. And what is going to happen at the end is that
172 00:20:55.910 ⇒ 00:21:05.039 Awaish Kumar: I have just for wholesale. I’m serving just wholesale right now, so it’s there for wholesale, then maybe we can just take one step
173 00:21:07.350 ⇒ 00:21:18.620 Awaish Kumar: One more step, and then we’re going to go and, like, create a, like, DIM customer for Shopify, which basically includes the e-commerce customers and the wholesale customers.
174 00:21:18.650 ⇒ 00:21:28.500 Awaish Kumar: And that… and then, ultimately, we have this only-channel view, where once we have data from all Amazon or the Walmart, so all…
175 00:21:28.760 ⇒ 00:21:43.429 Awaish Kumar: these all small tables will join into a big one table, a dim customer, which basically have customers from across, different channels for element. And then you can identify it through some flags.
176 00:21:46.530 ⇒ 00:21:58.210 Awaish Kumar: So you can see here are a few models I added, since the last time. So we have DIM Customer, which is basically the detailed view of, all the…
177 00:21:58.340 ⇒ 00:22:06.049 Awaish Kumar: partners we have in wholesale, and if it’s more in there, I… some SQL written there.
178 00:22:08.870 ⇒ 00:22:11.519 Awaish Kumar: And, yeah, basically…
179 00:22:11.840 ⇒ 00:22:27.430 Awaish Kumar: some customer data, some orders data, and joining it together to build this one. Then, on top of it, then we have some summary models, which are basically just a sheet which I’ve showed. So, this is to serve, like, one of…
180 00:22:27.590 ⇒ 00:22:36.879 Awaish Kumar: one of the use cases that Medicine was working on. So we are identifying for each partner what is a weekly summary view of,
181 00:22:37.040 ⇒ 00:22:38.800 Awaish Kumar: Revenue and orders.
182 00:22:38.940 ⇒ 00:22:42.230 Awaish Kumar: And basically, to identify the pattern.
183 00:22:42.880 ⇒ 00:22:52.369 Awaish Kumar: How… they made a change on January 7th that they moved some partners from previous tag to specialty retail.
184 00:22:52.610 ⇒ 00:22:55.310 Awaish Kumar: And they wanted to identify if that
185 00:22:55.650 ⇒ 00:23:02.030 Awaish Kumar: Changed anything in the behavior of the customer in terms of, like, placing the order.
186 00:23:02.500 ⇒ 00:23:13.810 Awaish Kumar: And for supporting that use case, we basically build that view, which can show them, like, how a customer has been ordering in the last few weeks, and…
187 00:23:13.900 ⇒ 00:23:25.159 Awaish Kumar: how they have ordered in the first week of January, before the 7th, January… before the change of 7th January, and after that. So, basically.
188 00:23:25.350 ⇒ 00:23:29.870 Awaish Kumar: Based on that, I, like, they could identify
189 00:23:29.980 ⇒ 00:23:36.430 Awaish Kumar: some patterns, and maybe able to make some… some decisions from there… from there.
190 00:23:36.580 ⇒ 00:23:41.700 Awaish Kumar: And then we have, as shown earlier, some sales.
191 00:23:42.640 ⇒ 00:23:45.040 Awaish Kumar: Martin, yeah, I think that…
192 00:23:45.480 ⇒ 00:23:51.600 Awaish Kumar: in the, basically, the intermediate, this is where I’m doing some normalization.
193 00:23:53.300 ⇒ 00:24:00.320 Awaish Kumar: before making some mods, so in this, if we go into the Shopify, we… I’m just doing some…
194 00:24:00.600 ⇒ 00:24:11.850 Awaish Kumar: cleanups. So, with the name, it is quite clear what it is for. So, we are in Shopify folder, that means we are working with Shopify data.
195 00:24:12.020 ⇒ 00:24:23.460 Awaish Kumar: Then, when it says in Shopify orders, that means we are working on order data, or a user’s data, or basically, specifically for wholesale DIM customer table.
196 00:24:23.610 ⇒ 00:24:27.499 Awaish Kumar: So, yeah, that’s how it goes from… so if I just…
197 00:24:27.820 ⇒ 00:24:34.990 Awaish Kumar: Give an overview of this. We… We have, raw folder, where everything…
198 00:24:35.660 ⇒ 00:24:46.590 Awaish Kumar: From, for example, these are different sources we are pulling in from, and with each source, we have just the one-on-one mapping with whatever is in the route source table.
199 00:24:46.880 ⇒ 00:24:49.210 Awaish Kumar: And once it is in there.
200 00:24:49.240 ⇒ 00:24:54.889 Awaish Kumar: We move on to the intertables, intermediate layer. That is basically…
201 00:24:54.920 ⇒ 00:25:11.429 Awaish Kumar: in this layer where we do some transformation, some normalization, if there is data in the JSON or something, we just do it here. And then in that process, if I identify if there are some logics which can be standardized, then they move into the macros.
202 00:25:11.600 ⇒ 00:25:18.199 Awaish Kumar: And finally, based on intermediate models and macros, we build the Mars.
203 00:25:20.920 ⇒ 00:25:24.769 Awaish Kumar: Yeah, so… Any questions on this one?
204 00:25:28.710 ⇒ 00:25:30.230 Jason Wu: No questions on my side?
205 00:25:32.270 ⇒ 00:25:41.810 Awaish Kumar: Okay yeah, so… That’s, I think,
206 00:25:42.020 ⇒ 00:25:57.590 Awaish Kumar: basically, on the dbt side, that’s the… there are the changes we have made so far, and what we are looking… yeah. In terms of, CICD, if you… I can show you the actions. I have shown it earlier.
207 00:25:59.200 ⇒ 00:26:13.660 Awaish Kumar: But you can see there are two workflows. One is staging, which runs for PR validation, and then there is one called intraday Run, which is now basically running each hour, but basically,
208 00:26:14.010 ⇒ 00:26:20.019 Awaish Kumar: Like, we are going to end up, identifying the patterns for each team.
209 00:26:20.460 ⇒ 00:26:38.300 Awaish Kumar: that, like, for wholesale, maybe we need refreshes once a day. For e-commerce, we might need refreshes, twice a day. So we have to identify these patterns while working with the teams, and based on that, we might generate a few more workflows here for each individual team.
210 00:26:38.360 ⇒ 00:26:41.329 Awaish Kumar: And in those workflows, we will only be
211 00:26:41.480 ⇒ 00:26:45.410 Awaish Kumar: Updating models, which are relevant to that specific team.
212 00:26:45.760 ⇒ 00:26:49.860 Awaish Kumar: And then… Yeah.
213 00:26:49.990 ⇒ 00:27:00.430 Awaish Kumar: Now, what is, what’s going to happen is that these sheets needs to be… basically update it.
214 00:27:03.690 ⇒ 00:27:10.490 Awaish Kumar: the sheets basically needs to be updated daily. So, for now, for example, I have just built this sheet.
215 00:27:10.950 ⇒ 00:27:18.930 Awaish Kumar: By downloading the data, but what we are… will be doing this week is using Polytomic.
216 00:27:19.460 ⇒ 00:27:23.330 Awaish Kumar: We are, we are going to set up the reverse ETL part as well.
217 00:27:23.490 ⇒ 00:27:28.320 Awaish Kumar: So… in the River City Tail Park, basically, in the
218 00:27:29.360 ⇒ 00:27:45.320 Awaish Kumar: In the polyatomic, we have connections. We can basically have, Snowflake, we have… if we have models in the Snowflake, and then we wanted to move this data in Google Sheets, so we just can go to… go into the model sinks.
219 00:27:45.430 ⇒ 00:27:51.829 Awaish Kumar: And basically, we need to add some models, we can select the connection, where’s data going to come from, we can name it.
220 00:27:52.030 ⇒ 00:27:59.040 Awaish Kumar: And select the table, or, like, whatever way we are going to use for that specific,
221 00:28:00.090 ⇒ 00:28:03.369 Awaish Kumar: For example, if I want to build a model out of,
222 00:28:05.690 ⇒ 00:28:14.409 Awaish Kumar: Yeah, this is the connection for RAW, so it’s only showing the… Schemas from raw…
223 00:28:14.580 ⇒ 00:28:20.049 Awaish Kumar: database, but we need to have… create one more Snowflake connection for our prod.
224 00:28:20.150 ⇒ 00:28:38.669 Awaish Kumar: database, so that, like, it has access to the final parts table, and basically using that, we are going to generate a model in Polytomic, and then connect it with the Google Sheet we have, and basically it will sync the data from here to the Google Sheet.
225 00:28:38.750 ⇒ 00:28:41.020 Awaish Kumar: And, it’s going to do…
226 00:28:41.300 ⇒ 00:28:46.040 Awaish Kumar: On a schedule, we can set up… At minimum,
227 00:28:46.420 ⇒ 00:28:52.930 Awaish Kumar: I think 5 minutes or so, so that’s how it’s going to sync, so we can make it…
228 00:28:53.090 ⇒ 00:29:01.890 Awaish Kumar: as quick as possible, or, like, whatever is necessary. If it is… if we don’t need that, we can just keep it daily refreshes, or… or…
229 00:29:02.140 ⇒ 00:29:04.929 Awaish Kumar: Twice a day, refresher, something like that.
230 00:29:08.950 ⇒ 00:29:12.829 Awaish Kumar: Yeah, so that’s basically it.
231 00:29:13.110 ⇒ 00:29:19.599 Awaish Kumar: on… on this. So, yeah. Is there any question?
232 00:29:24.120 ⇒ 00:29:26.780 Jason Wu: No additional questions on my side.
233 00:29:27.270 ⇒ 00:29:31.769 Jason Wu: with Shivani and Utem back on the call, I was making the comment to a wish that
234 00:29:32.460 ⇒ 00:29:38.920 Jason Wu: like, I… like, the spreadsheet that what you’re showing us in terms of kind of, like, the aggregated data that we’re showing for wholesale.
235 00:29:39.050 ⇒ 00:29:44.720 Jason Wu: Based on kind of everything now in Snowflake. I’m just kind of curious how we can use this as, like.
236 00:29:44.890 ⇒ 00:29:48.600 Jason Wu: Kinda like the… The jump-off point for…
237 00:29:49.080 ⇒ 00:29:55.049 Jason Wu: you know, exploring Omni as, like, the UI tool. Just really thinking about…
238 00:29:55.970 ⇒ 00:30:13.669 Jason Wu: like, even though we’ll be able to provide some information right now to wholesale, it’s just, like, how quickly can we kind of close that gap? So, we really see kind of, like, the end-to-end, right? And it’s, I mean, with the dashboards, there’s that visual-ness away from the spreadsheet that I think is just kind of where the power lies.
239 00:30:13.710 ⇒ 00:30:16.489 Jason Wu: So, just kind of thoughts around, kind of, when we last spoke.
240 00:30:16.600 ⇒ 00:30:23.080 Jason Wu: After seeing kind of the Omni, the Omni demo, kind of, you know, where those next steps lie, and kind of the timing with that.
241 00:30:23.750 ⇒ 00:30:34.020 Shivani Amar: And with them, just context for you, I slacked you one-to-one, but I did a call with somebody on our supply side of the business today, and he really likes Domo from his previous world.
242 00:30:34.020 ⇒ 00:30:44.500 Shivani Amar: And so, like, I think, like, before we just, like, select something, especially if we select Omni, then the question becomes, can we test it with our own data to make sure we like it, versus…
243 00:30:45.430 ⇒ 00:30:52.120 Shivani Amar: Versus just going with it, and then, like, getting input from a few people at the company to say, what do you… what do you want us to explore?
244 00:30:52.890 ⇒ 00:31:05.140 Uttam Kumaran: Definitely. So, Jason, in terms of timeline, I mean, you kind of, like, nailed the reason why we actually went after this wholesale example. Really, the timeline around the BI execution is just next month.
245 00:31:05.210 ⇒ 00:31:23.779 Uttam Kumaran: We will be doing an exploration of Omni, Sigma, Domo, and yes, we’ll actually be using this use case as, like, a first and then client. Of course, we… we could always do, like, dummy data and show that, but it’s actually helpful to take this and take that business unit, so…
246 00:31:23.920 ⇒ 00:31:38.349 Uttam Kumaran: net-net, like, we basically are planning this for… for next month. It doesn’t mean… and basically, kind of, like, all… additionally, we will be trying to get a trial for as long as we need. There’s no reason we haven’t kicked that off now, apart from just…
247 00:31:38.440 ⇒ 00:31:50.590 Uttam Kumaran: there’s just a… as soon as we start talking to vendors, they’re gonna just take up a bunch of our brainpower and cycles, so we wanted to close out this month, especially because wholesale has some immediate priorities that
248 00:31:50.650 ⇒ 00:32:04.689 Uttam Kumaran: the Google Sheet, even, that we produced, and that model is, like, already assisting with, and so we felt like, okay, if… let’s just nail that, nail the dbt models for them, and then we’re gonna explore the BI landscape next month.
249 00:32:04.910 ⇒ 00:32:06.740 Uttam Kumaran: And, and, and records.
250 00:32:07.720 ⇒ 00:32:21.239 Jason Wu: Yeah, and I think my comments really kind of stem from, like, making sure that we’re really kind of setting the context right with the business that… I mean, you know, and obviously I think, like, wholesale, you know, they kind of understand this already, but it’s…
251 00:32:21.380 ⇒ 00:32:28.700 Jason Wu: like, just taking their spreadsheet and converting it to another spreadsheet isn’t the end goal. It’s… It’s,
252 00:32:29.130 ⇒ 00:32:39.810 Jason Wu: here’s, like, a stopgap, right? You know, so you can start using, you know, using the data, like, right now off of production data. But making sure that we just are…
253 00:32:40.070 ⇒ 00:32:46.779 Jason Wu: sit in that context as far as, like, kind of, like, the broader, like, strategy, and kind of, like, here’s where it’s gonna land.
254 00:32:47.060 ⇒ 00:32:59.230 Jason Wu: Yeah, that’s… that’s really kind of the main observation there. So, I… and maybe that’s a way of also saying, like, depending on kind of what the reactions are from… have you had a chance to show this to Laura… to Laura yet? I can’t remember if that’s.
255 00:32:59.230 ⇒ 00:33:03.099 Uttam Kumaran: Yeah, we did. We’ve been chatting with them basically all week.
256 00:33:03.100 ⇒ 00:33:03.420 Jason Wu: Okay.
257 00:33:03.420 ⇒ 00:33:06.429 Uttam Kumaran: yesterday, and on a Monday, we shared, we shared this.
258 00:33:06.650 ⇒ 00:33:10.369 Jason Wu: Okay. I mean, I don’t know how that call went, but, like.
259 00:33:10.560 ⇒ 00:33:30.040 Jason Wu: Is there, like, that energy to say, hey, we’d love to see this and this and this, where, like, put it this way, do we pause and say, you know, we’re gonna get there? This is where kind of the UI tool comes in, you know, versus investing, you know, any sort of additional time that might be throwaway on the sheet if the next step’s already kind of the UI?
260 00:33:30.260 ⇒ 00:33:37.549 Shivani Amar: Well, no, I think that part of the… oh, sorry, I’m answering that question just to say that, like, there’s still some, like.
261 00:33:37.550 ⇒ 00:33:57.889 Shivani Amar: do we have the main fields that you would even want in your day-to-day, and are we defining them correctly? And hey, they’re also duplicates, like, there’s just, like, something from, like, seeing the table before we just say, let’s, like, layer a BI on top. Like, already we can see there are duplicate rows, and so it’s like, what do Madison and Laura need to do now to go back and, like, clean up some of their own data?
262 00:33:57.890 ⇒ 00:34:00.459 Shivani Amar: So that the things we’re ingesting are clean.
263 00:34:00.630 ⇒ 00:34:15.510 Shivani Amar: then there’s, like, an example of, like, what do we consider an at-risk wholesale partner, which is, like, a definition that, like, once we have alignment on it, it will be good as an input into a BI tool. But, like, the first thing is we want their reaction to, like.
264 00:34:15.510 ⇒ 00:34:23.239 Shivani Amar: would somebody be at risk if they haven’t placed an order in 90 days, or is that too short of a window? Would we call somebody at risk if they haven’t placed an order in, like.
265 00:34:23.239 ⇒ 00:34:44.610 Shivani Amar: 180 days, like, what is that limit that we want to call a wholesale partner potentially at risk? And then once we’ve, like, come up with some of those definitions, they can say, I want to see all my, like, trusted health partners that are at risk, and, like, whatever, and then have Omni do something like that. But I think it’s nice to start with alignment on the clean table to be like, is this even, like, the right method to do this?
266 00:34:45.969 ⇒ 00:34:47.169 Shivani Amar: And, like, are the right titles.
267 00:34:47.170 ⇒ 00:34:48.059 Uttam Kumaran: Yeah, so.
268 00:34:48.239 ⇒ 00:34:58.129 Shivani Amar: They had, like, revenue and then subtotal revenue, and it’s like, what is the difference between those, right? So before I feed anything to a BI tool, I want to just make sure that the language we’re all using is, like, consistent.
269 00:34:59.260 ⇒ 00:35:08.550 Uttam Kumaran: Yeah, and really, like, the world they were coming from is, like, they’re hitting export in Shopify, like, every week or so, right? So it’s so… it’s super, super…
270 00:35:08.550 ⇒ 00:35:23.979 Uttam Kumaran: prehistoric, and so this is already, like, a huge leap forward. And I think to your point, like, it’s actually… we would have to do this work to actually enable the BI solution, anyways. Like, we weren’t gonna be able… like, we can’t…
271 00:35:23.980 ⇒ 00:35:31.450 Uttam Kumaran: we wouldn’t… what we… what the step we didn’t skip is, like, some companies will just put BI right on the raw data and say, like, go for it, right? So…
272 00:35:31.450 ⇒ 00:35:45.610 Uttam Kumaran: we kind of need to create this, like, governance layer in between. So, like, net-net, it’s not, you know, there is an order of operations. I think kind of the discussion, Shivani, I want to bring up is… is there’s going to be multiple stakeholders of data and multiple ways
273 00:35:45.610 ⇒ 00:36:04.210 Uttam Kumaran: that they use it. To give you a couple examples, there will be some people in the company that only consume the data from, like, fixed dashboards. There will be some folks that are creating dashboards and are exploring the data. There also will continue to be some folks that want exports, in order to do some more sophisticated analysis
274 00:36:04.210 ⇒ 00:36:20.249 Uttam Kumaran: you know, in Excel or Google Sheets. So, I think it’s to make sure that as a data platform, we’re confident that we can support all of those. In addition, there’s gonna be some customers that are, like, you know, MHI or NetSuite is, like, gonna be a customer, potentially a customer of the data platform. So.
275 00:36:20.360 ⇒ 00:36:29.459 Uttam Kumaran: there’s both a lot of inputs, modeling, like, a governed layer, and then it kind of gets shipped out. So, certainly, I think we need to agree on, like, what
276 00:36:29.460 ⇒ 00:36:45.649 Uttam Kumaran: what are we supporting? I think, Shivani, we’ve discussed that there are still going to be a wide variety of folks. Like, some people are going to need this in a tabular format in order to do, you know, Excel-based modeling. There’s also going to be some folks that just want to build dashboards and scheduled reports.
277 00:36:47.860 ⇒ 00:37:00.099 Uttam Kumaran: and that there is a, like, there’s a business use case for… for both, you know? Because even working with you, Shivani, I could tell your brain is much more tabular, but there’s, of course, some people that are like, I just need to see the dashboard visually, right? Those are both, like.
278 00:37:00.100 ⇒ 00:37:01.799 Shivani Amar: other tabular brain.
279 00:37:01.800 ⇒ 00:37:03.799 Uttam Kumaran: That’s my fault.
280 00:37:03.800 ⇒ 00:37:04.230 Shivani Amar: Did they get close?
281 00:37:04.230 ⇒ 00:37:14.839 Uttam Kumaran: stakeholder history. And we want to support both. Some people, it’s going to be because of their time, some people it’s just gonna be their level of sophistication with the data.
282 00:37:14.880 ⇒ 00:37:28.720 Uttam Kumaran: And… but all of that, again, we don’t want them to go straight… right now, they’re… think about them going straight to the source. They’re just pulling orders from Shopify, not able to layer on business logic, really manual, weekly, monthly reconciliations.
283 00:37:28.810 ⇒ 00:37:30.319 Uttam Kumaran: That’s what we want to avoid.
284 00:37:33.650 ⇒ 00:37:36.099 Jason Wu: Got it. Yeah, rodered that. Thanks for the clarity, Adam.
285 00:37:36.810 ⇒ 00:37:56.509 Uttam Kumaran: Yeah, and so for us, for all the BI tools, I’m gonna be, you know, pushing them to give us, you know, typically most of them will do 4-week trials, like, I’ll push to get as much, as further as we can, so that we can actually take this use case and actually see it across several systems, and, you know, make the decision that’s best.
286 00:37:58.590 ⇒ 00:37:59.610 Uttam Kumaran: So, yeah.
287 00:38:03.020 ⇒ 00:38:11.980 Jason Wu: While we’re on the topic of trials, so I know we’re kind of nearing our trial end for Snowflake,
288 00:38:11.980 ⇒ 00:38:12.580 Shivani Amar: Over snow.
289 00:38:12.580 ⇒ 00:38:13.569 Jason Wu: As far as, like.
290 00:38:13.830 ⇒ 00:38:26.620 Jason Wu: as far as, like, next steps for that, I mean, obviously the easiest thing is just, you know, add my credit card and just kind of start charging for it, but is there a process where we would get connected with sales, just kind of understand, like, what does, like, an annual, like, an enterprise-level contract look like?
291 00:38:27.170 ⇒ 00:38:29.359 Jason Wu: Or, what’s the best approach there?
292 00:38:30.010 ⇒ 00:38:44.230 Uttam Kumaran: Yeah, so what we’ll do, I did see that yesterday when I was in Snowflake. So one, we have some more free credits, so I may… I may go ahead and ask them to see, like, how long they can extend it, but yes, like, they’ll… they’ll assign us a…
293 00:38:44.260 ⇒ 00:39:02.569 Uttam Kumaran: someone on sales to kind of explain, like, what a capacity contract would look like. Because we don’t… right… what they will guide us towards doing, and this is from, you know, our experience, we’ve probably done Snowflake, like, 30 times, they’re gonna guide you towards trying to buy a fixed amount of credits, over the year. Typically, their contracts start at
294 00:39:02.570 ⇒ 00:39:07.009 Uttam Kumaran: Like, 15K annually, around there, and that will…
295 00:39:07.140 ⇒ 00:39:24.359 Uttam Kumaran: And you’ll get basically a 20% to 40% discount on credits by signing up for annual. What we’ll look at now is look at, like, what is our run rate expense. It’s not going to be anywhere near that right now, so we have kind of a decision to make. One is, like, we kind of purchased that
296 00:39:24.370 ⇒ 00:39:41.480 Uttam Kumaran: in advance of, like, us ramping usage, or we just wait until the point at which that equation sort of makes sense. From the vendor perspective, they’re just gonna push us towards the annual contract. So, I guess I’m curious, Jason, on, like, how you think about things. Would you prefer that we, like.
297 00:39:41.600 ⇒ 00:39:48.699 Uttam Kumaran: you know, continue to… I mean, again, from our earlier conversation, it’s like, see if we can run month-to-month, and then maybe if the economics make sense.
298 00:39:49.240 ⇒ 00:39:57.770 Uttam Kumaran: think about, like, okay, like, an annual contract would work here. Of course, like, the annual contract comes with some better support and a few other bells and whistles.
299 00:39:58.070 ⇒ 00:40:03.170 Uttam Kumaran: So, like, yeah, I mean, we can… we can… We can discuss that.
300 00:40:03.500 ⇒ 00:40:12.540 Jason Wu: Yeah. I mean, tools like this, like, I’m also realistic when it comes, like, month-to-month for, like, these larger tools, like, sometimes that’s just not… not possible, and I don’t.
301 00:40:12.540 ⇒ 00:40:18.469 Uttam Kumaran: Like, in the BI layer, in the BI layer, it won’t be… it won’t be possible, for the most part. Yeah, like… Yeah.
302 00:40:18.470 ⇒ 00:40:20.759 Jason Wu: like, where I… where I tend to go is…
303 00:40:20.830 ⇒ 00:40:40.820 Jason Wu: you know, like, we work with some vendors where they’re like, oh, in order to get, kind of, sort of, any kind of discount, you know, you need to lock in on, like, a 2- or 3-year deal, and I’m like, that’s a non-starter for me, you know, and they’ll usually budge and say, okay, one year. And that’s kind of typically where it just gives us enough time to play with it, digest it, ultimately, and, like, feel like this is going to be a long-term tool or not. So…
304 00:40:40.900 ⇒ 00:40:52.899 Jason Wu: you know, I mean, we haven’t had any issues with Snowflake, and I think we’re already kind of, you know, saying this just logically makes sense. So, for me, it’s just really trying to understand, like, well, what’s the best deal? You know, and like, how do we work towards that?
305 00:40:53.740 ⇒ 00:40:57.359 Uttam Kumaran: Yeah, so let me… I’ll, it’s, again, like, they’re…
306 00:40:57.470 ⇒ 00:41:16.989 Uttam Kumaran: it’s really funny, we talk to a lot of Snowflake reps, and we typically, like, I know their discount schedules and a lot inside out, so I can share with you a lot of, how they typically price and, like, what those look like. Typically for our clients, we’re like, hey, once you start to hit the usage rate, where if you scroll that out 12 months.
307 00:41:16.990 ⇒ 00:41:24.470 Uttam Kumaran: you would be… you’d be well over, like, what they would give you, as an annual, like, we should go for it. We’re probably within…
308 00:41:24.670 ⇒ 00:41:44.610 Uttam Kumaran: two or three months of that right now, because as you think about tenants of Snowflake, right, we’ll have the BI tool, we’ll have dbt running models, we have Polyatomic that’s landing stuff in there. We may also have, MHI on the NetSuite side, and so those will all kind of be over the next
309 00:41:45.020 ⇒ 00:41:53.680 Uttam Kumaran: 3 months, where, like, all the core tenants are connected, and then probably we’ll be able to have a pretty good sense of how usage is going to change over time.
310 00:41:54.000 ⇒ 00:42:06.509 Uttam Kumaran: So if we’re not in a rush, I would probably vote to just… like, I can… I’ll start the… I can connect us with them, so we have a… put a face to a name on who our rep is. I would say, just, let’s wait until, like, we hit that.
311 00:42:06.510 ⇒ 00:42:07.050 Jason Wu: Okay.
312 00:42:07.050 ⇒ 00:42:08.829 Uttam Kumaran: Point at which… it’s, like.
313 00:42:08.830 ⇒ 00:42:09.190 Jason Wu: Yeah.
314 00:42:09.190 ⇒ 00:42:10.280 Uttam Kumaran: Money-wise.
315 00:42:10.280 ⇒ 00:42:16.880 Jason Wu: Totally fine for that, not a rush, more of a call-out that we’re starting to get the trials up, you know, kind of notices.
316 00:42:16.880 ⇒ 00:42:17.609 Uttam Kumaran: Yeah, okay.
317 00:42:17.610 ⇒ 00:42:18.080 Jason Wu: But…
318 00:42:18.080 ⇒ 00:42:30.430 Uttam Kumaran: So yeah, maybe I’ll… I can ping you today. I’ll go see if we can get it extended, and then I’ll see if we can… I’ll let you know if we want to put a card down. And then on the BI side, though, yes, they will…
319 00:42:30.430 ⇒ 00:42:39.699 Uttam Kumaran: like, 95% will be, annual deals. But there, you have a lot more leverage, like, you can really, really negotiate hard.
320 00:42:39.870 ⇒ 00:42:48.610 Uttam Kumaran: And so that’s what we’ll… we’ll help you guys do on, like, seat counts, and cost per seat, and annual discounts, and term, and things like that, so…
321 00:42:49.360 ⇒ 00:42:50.780 Jason Wu: Great. Bye.
322 00:42:55.920 ⇒ 00:42:56.250 Uttam Kumaran: Great.
323 00:42:56.250 ⇒ 00:42:59.330 Andy Weist: In our work with…
324 00:42:59.470 ⇒ 00:43:09.100 Andy Weist: Tom, you brought up MHI earlier. You know, that ball is getting rolling. Our NetSuite integration project is getting rolling.
325 00:43:10.300 ⇒ 00:43:17.539 Andy Weist: where do you see we are as far as possibly being able to use the BI platform as the ingest point for NetSuite?
326 00:43:17.690 ⇒ 00:43:19.410 Andy Weist: I think, like.
327 00:43:20.550 ⇒ 00:43:29.699 Andy Weist: at the start, we’ll mostly want Shopify data, and probably payments things, so… I guess I’m… I haven’t,
328 00:43:30.400 ⇒ 00:43:41.400 Andy Weist: I haven’t understood yet if we’re all the way with Shopify integration, if it’s just a subset test. I know we were backfilling. I think Steve also was curious about that.
329 00:43:42.240 ⇒ 00:43:51.469 Andy Weist: Do you think we’re in a place where we could start building the MHI, the NetSuite integration, off of Snowflake, or is that a red flag to you? Just what’s your opinion there?
330 00:43:52.770 ⇒ 00:44:01.370 Uttam Kumaran: Yeah, so, I think we talked a little bit briefly last week about this. From my understanding, this is, like, an… basically an alternative to them
331 00:44:01.510 ⇒ 00:44:09.299 Uttam Kumaran: ingesting the Shopify data into their analytics warehouse. Right now, we have the schema
332 00:44:09.310 ⇒ 00:44:26.000 Uttam Kumaran: prepared, and basically data is getting backfilled historically. So in terms of driving towards setting up an integration, there’s nothing on the shape of the data that’s gonna change, and we already have all the core objects landed. There’s just still some backfills occurring.
333 00:44:26.040 ⇒ 00:44:34.869 Uttam Kumaran: So for me, it’s sort of… I would be interested to hear, one is, like, what are their milestones for that integration? Like.
334 00:44:34.980 ⇒ 00:44:43.220 Uttam Kumaran: Do they just need to hook it up and sort of start to drive some data so they can build on top of it? Or do they require a certain subset?
335 00:44:43.260 ⇒ 00:44:58.489 Uttam Kumaran: you know, is it just, like, the most recent data? So I think there’s some questions for me. In terms of being able to handle that, like, I don’t think there’s any problem. We have it all segregated into a raw schema. All of our data is kind of done in that way, where
336 00:44:58.490 ⇒ 00:45:03.730 Uttam Kumaran: We have it all in raw, and our modeling for reporting needs happens in a different database.
337 00:45:03.950 ⇒ 00:45:12.940 Uttam Kumaran: So getting that hooked up, that’s no problem. Would just be interested to hear, like, what’s their timeline, and, like, what objects they need.
338 00:45:13.310 ⇒ 00:45:18.759 Uttam Kumaran: So, less on, like, if it’s possible, more on, like, what they need, and…
339 00:45:18.870 ⇒ 00:45:25.459 Uttam Kumaran: And, like, how to facilitate getting that to them. I’m also interested, like, okay, are they looking for…
340 00:45:25.630 ⇒ 00:45:32.599 Uttam Kumaran: like, just a, basically, service credentials to access? Are they looking for, like, an export to a bucket?
341 00:45:32.870 ⇒ 00:45:35.059 Uttam Kumaran: Or is this something more complicated?
342 00:45:36.180 ⇒ 00:45:42.370 Andy Weist: Okay, yeah, we have a meeting with them tomorrow that will be pertinent on all those points. I can help drive, like.
343 00:45:42.370 ⇒ 00:45:47.390 Uttam Kumaran: I can draft… Yeah, I can draft some of those questions, even to just… so you have them on hand.
344 00:45:47.670 ⇒ 00:45:48.540 Uttam Kumaran: Or you can…
345 00:45:48.540 ⇒ 00:46:02.449 Andy Weist: Basically, if we decide to push the NetSuite integration toward, like, onto the BI layer, I’m personally in favor of not having multiple systems pulling data from Shopify and, like, all these different systems. Agreed.
346 00:46:02.450 ⇒ 00:46:02.980 Uttam Kumaran: Yeah.
347 00:46:02.980 ⇒ 00:46:09.110 Andy Weist: independently, and instead running everything through BI, and then going to NetSuite, so you kind of have a single ingest point.
348 00:46:10.510 ⇒ 00:46:20.780 Andy Weist: I can help… we’re still talking through some of those things. They have, like, this mutation service that will summarize data that we’re trying to figure out if
349 00:46:21.090 ⇒ 00:46:22.929 Andy Weist: If we’re definitely going with that.
350 00:46:23.310 ⇒ 00:46:28.219 Andy Weist: But at some point, it might be worthwhile just having both teams talk as well.
351 00:46:28.390 ⇒ 00:46:33.679 Andy Weist: I can try to get some of those answers and just shape that up, but it sounds like
352 00:46:33.780 ⇒ 00:46:46.990 Andy Weist: we could at least start pushing Shopify out of Snowflake into NetSuite if we wanted to set up a quick, like, proof of concept or something soon. Also, do not know timeframe on them yet, but we may get more information tomorrow.
353 00:46:47.790 ⇒ 00:46:51.570 Uttam Kumaran: Yeah, I tend to agree, this is… yeah, yeah, go ahead, Jason.
354 00:46:52.460 ⇒ 00:46:54.539 Jason Wu: I’m sorry to complete your thought.
355 00:46:55.460 ⇒ 00:47:12.719 Uttam Kumaran: No, I said I tend to agree, like, we’re already ingesting this, we’re already paying someone to… we’re already paying a vendor to do that. There’s gonna be several tenants like NetSuite that will pull from this data in the future, so keeping one way of doing that is probably… is probably fair.
356 00:47:12.940 ⇒ 00:47:23.039 Uttam Kumaran: you know, just as a… they just become another customer of, like, sort of the data platform to do so, but it seems like they’re just gonna be grabbing Raw, and they’re gonna do mapping and things on their side, so…
357 00:47:25.250 ⇒ 00:47:37.459 Uttam Kumaran: I mean, to be honest, we’re having similar thoughts when we think about, like, an atomic, or a cube, or these more purpose-driven analytical tools that certain business units are using, that, like, our team will… may have to support
358 00:47:37.580 ⇒ 00:47:44.619 Uttam Kumaran: sending data, or making data available for them. For example, we’re considering, like, a pretty advanced forecasting tool.
359 00:47:44.670 ⇒ 00:48:03.560 Uttam Kumaran: Yes, like, could we go build a really robust forecasting tool in our typical BI? Will it take a long time and compete with priorities? Yes. So, like, we’re discussing, okay, if a FP&A team adopts this tool, how do… how does that team also leverage what the BI team has already developed in terms of landing raw data?
360 00:48:03.680 ⇒ 00:48:15.760 Uttam Kumaran: Instead of, like, having whatever that tool go on their own, connecting directly, or them duplicating work, or basically limiting features. So we’re thinking about supporting, like, multiple tenants in this way.
361 00:48:15.980 ⇒ 00:48:25.290 Andy Weist: Yeah, I’m really just concerned about redundancy here, because the two projects are so similar in terms of integrations and stuff, so…
362 00:48:25.430 ⇒ 00:48:25.870 Uttam Kumaran: Yeah.
363 00:48:25.870 ⇒ 00:48:41.050 Andy Weist: I just want to make sure we, especially Element Team, is aware of making sure, like, we’re kind of streamlining this in some way that makes sense, and we’re not just building some two redundant systems that do all the same things and then become a hassle to maintain long-term, so…
364 00:48:41.530 ⇒ 00:48:44.089 Andy Weist: Okay, thanks, thanks for that update.
365 00:48:48.970 ⇒ 00:48:51.170 Uttam Kumaran: Jason, I don’t know if you had anything else.
366 00:48:51.170 ⇒ 00:48:54.640 Jason Wu: No, it was just echoing the same component,
367 00:48:54.880 ⇒ 00:48:56.980 Jason Wu: I mean, the project’s been kicked off.
368 00:48:56.990 ⇒ 00:49:09.610 Jason Wu: Originally, we were supposed to have that ACE conversation, that’s the mutation kind of tool that, Andy was referring to. We were supposed to have that on Tuesday, that’ll happen on Thursday now, so we’ll get some updates there.
369 00:49:09.610 ⇒ 00:49:17.990 Jason Wu: Regardless, what MHI has told us is they don’t have Snowflake resources, you know, they’re kind of a NetSuite shop, you know, so…
370 00:49:18.110 ⇒ 00:49:34.209 Jason Wu: you know, my expectation is, after they go through, kind of, their requirements kind of phase, they’ll put together of, like, what would they want the data to look like to be sent to NetSuite, and then we’d lean on BrainForge, and we’d have that discussion with you to say what’s the best way to kind of get it over there.
371 00:49:34.980 ⇒ 00:49:47.980 Uttam Kumaran: Yeah, so in terms of supporting that, yeah, I don’t have any worry or fear about that. I think more, it’s, like, I want to understand their SLAs, and then I think Shivani, for our context, is, like, they may become our, like.
372 00:49:48.210 ⇒ 00:49:57.359 Uttam Kumaran: first, like, kind of customer, right? Even if it’s just getting them the raw data, they are downstream of our work, and so I just want to make sure that’s
373 00:49:57.580 ⇒ 00:50:01.009 Uttam Kumaran: Maintaining that and supporting that is just on our roadmap.
374 00:50:01.350 ⇒ 00:50:05.169 Uttam Kumaran: And then, like, yeah, kind of, like, what their SLAs are in case…
375 00:50:05.660 ⇒ 00:50:08.830 Uttam Kumaran: You know, things break, and so we have a direct line to them.
376 00:50:09.230 ⇒ 00:50:15.040 Shivani Amar: Yeah, go ahead. I think tomorrow we’ll be telling, after Jason has this conversation with them, and then we’ll, like.
377 00:50:15.040 ⇒ 00:50:15.550 Uttam Kumaran: Yeah.
378 00:50:15.550 ⇒ 00:50:19.469 Shivani Amar: We’ll just have more intel, and if it makes sense for you guys to get to know each other, but…
379 00:50:19.760 ⇒ 00:50:33.200 Shivani Amar: Which will probably make sense at some point. But yeah, it’s a funny thing, I just had a conversation with somebody on the supply chain side of our business who’s, like, eager to, like, like, oh, when is the Discover gonna start with me and Brainforge? And, like, you know, so…
380 00:50:33.200 ⇒ 00:50:34.189 Uttam Kumaran: He’s kidding.
381 00:50:34.190 ⇒ 00:50:39.540 Shivani Amar: when is my data gonna get ingested? So… so I think we will, we’ll just…
382 00:50:39.540 ⇒ 00:50:43.339 Uttam Kumaran: We just want to nail it, I think, for everybody. Yeah, we want to be thoughtful, and we want to nail it for everybody.
383 00:50:43.340 ⇒ 00:51:02.049 Shivani Amar: I think, like, I just need to… I was just educating him, and I was like, we started with the commercial side of the business, and he’s like, oh, that’s helpful for me to, like, understand contextually. And he’s like, if I had a data warehouse, I think it would save me 3 hours a day in what I’m doing right now. And I was like, oh, wow! Okay!
384 00:51:02.050 ⇒ 00:51:04.950 Uttam Kumaran: Gonna become famous, or infamous.
385 00:51:04.950 ⇒ 00:51:22.470 Shivani Amar: Yeah, yeah, so I was like, we’re here to serve in time, so I think, Jason, like, you, me, and Utham can jam on this either next week or, like, early Feb, like, what, like, the timeline beyond the just commercial side of the business.
386 00:51:23.290 ⇒ 00:51:24.030 Uttam Kumaran: Yeah.
387 00:51:24.030 ⇒ 00:51:29.630 Shivani Amar: just, like, a refresh of the roadmap, because the Gantt chart at the beginning was more just, like, for that one piece of the business.
388 00:51:30.690 ⇒ 00:51:31.050 Uttam Kumaran: Yeah.
389 00:51:31.050 ⇒ 00:51:31.630 Shivani Amar: Cool.
390 00:51:32.630 ⇒ 00:51:33.690 Shivani Amar: Okay.
391 00:51:36.220 ⇒ 00:51:36.810 Uttam Kumaran: Okay.
392 00:51:37.110 ⇒ 00:51:54.249 Uttam Kumaran: Great, so I think by the next time we chat, yeah, even if this MHI conversation gets further tomorrow, and, like, we want to chat next week, we can. But, you know, I think probably by the next time us and the tech team meet, we will be a bit further on wholesale.
393 00:51:54.280 ⇒ 00:51:58.630 Uttam Kumaran: And, like, I guess, or I guess technically that… that may be within REST and Assess, right?
394 00:51:58.990 ⇒ 00:52:01.360 Shivani Amar: Yeah, I don’t think there is… In the two weeks, so…
395 00:52:01.510 ⇒ 00:52:03.560 Shivani Amar: Yeah, I think the next one.
396 00:52:04.150 ⇒ 00:52:04.660 Uttam Kumaran: It’s a fab.
397 00:52:04.660 ⇒ 00:52:05.260 Shivani Amar: slated.
398 00:52:05.260 ⇒ 00:52:06.330 Uttam Kumaran: First week.
399 00:52:07.360 ⇒ 00:52:11.939 Shivani Amar: Like, I think… did we cancel the series and only do it for the sprint? It looks like…
400 00:52:11.940 ⇒ 00:52:17.250 Uttam Kumaran: I didn’t… I don’t think I… yeah, I’m not sure if I was smart and did that.
401 00:52:18.050 ⇒ 00:52:19.059 Shivani Amar: So why don’t we do the next.
402 00:52:19.060 ⇒ 00:52:21.640 Uttam Kumaran: We’ll do every… Yeah.
403 00:52:22.780 ⇒ 00:52:23.800 Shivani Amar: Feb 11th.
404 00:52:24.620 ⇒ 00:52:25.190 Uttam Kumaran: Okay.
405 00:52:25.600 ⇒ 00:52:32.750 Shivani Amar: But then there’s a chance that me, you, and Jason will do a sync, like, even earlier next week than our… than our regular call, just to talk.
406 00:52:32.750 ⇒ 00:52:33.110 Uttam Kumaran: Yeah.
407 00:52:33.110 ⇒ 00:52:34.379 Shivani Amar: It’s like an AI thing.
408 00:52:35.210 ⇒ 00:52:44.459 Uttam Kumaran: Yeah, and I think Feb 11th, you know, ideally we will… we can discuss as a group, like, how we’re… we’re attacking BI, and, like, I think we’ll be on our way into that world, so…
409 00:52:44.820 ⇒ 00:52:45.670 Shivani Amar: Perfect.
410 00:52:47.620 ⇒ 00:52:48.700 Uttam Kumaran: Great, okay, so I have some stoked.
411 00:52:48.700 ⇒ 00:52:49.530 Shivani Amar: Domo?
412 00:52:50.500 ⇒ 00:52:53.229 Uttam Kumaran: I’ve used Domo in the past,
413 00:52:53.460 ⇒ 00:52:59.040 Uttam Kumaran: they’re just… that’s… they were… they were good in… I don’t know, was the stakeholder in marketing?
414 00:52:59.230 ⇒ 00:53:00.040 Uttam Kumaran: Or finance?
415 00:53:00.040 ⇒ 00:53:01.029 Shivani Amar: supply chain.
416 00:53:01.670 ⇒ 00:53:02.219 Shivani Amar: But he…
417 00:53:02.220 ⇒ 00:53:02.859 Uttam Kumaran: Okay, like, I’m working.
418 00:53:02.860 ⇒ 00:53:04.870 Shivani Amar: previous world, so maybe he used it, like, 5 years.
419 00:53:04.870 ⇒ 00:53:07.000 Uttam Kumaran: Where… do you know where… do you know where he worked?
420 00:53:07.250 ⇒ 00:53:08.120 Shivani Amar: -
421 00:53:08.310 ⇒ 00:53:13.079 Uttam Kumaran: Okay, alright, that’s okay. Yeah, I have some friends that used Domo when they were working in…
422 00:53:13.400 ⇒ 00:53:23.949 Uttam Kumaran: And, like, this other e-commerce retailer business. It’s good, I mean, there’s just better… it’s not… that was, like, yeah, they were bigger 5 years ago. Them mode were just bigger 5 years ago, so…
423 00:53:23.950 ⇒ 00:53:25.819 Shivani Amar: Oh, he used to work at Where To Go.
424 00:53:27.140 ⇒ 00:53:28.320 Uttam Kumaran: Oh, nice.
425 00:53:28.670 ⇒ 00:53:39.949 Uttam Kumaran: Yeah. That’s funny. Wait, maybe he can help us if we need help with integration stuff. But, yeah, I mean.
426 00:53:40.770 ⇒ 00:53:47.889 Uttam Kumaran: they’ve… it’s… again, it’s gonna be just… it’s so subjective, it’s tough. Like, Domo is a good tool for a bunch of things.
427 00:53:48.040 ⇒ 00:54:06.360 Uttam Kumaran: I… I feel like I like some of them. I like Sigma and Omni, like, just much more robust. And the AI features are, like… I don’t… I haven’t been keeping up with Domo, but Domo, Mode, and a few tools, they were really big 5-6 years ago, and they sort of are no longer as big, which maybe they got bought, or they just… yeah.
428 00:54:06.710 ⇒ 00:54:12.619 Uttam Kumaran: But we’ll go through a bunch, again, like, I think we’ll have a month to sort of show everything side by side, and see, like, what the flavors…
429 00:54:13.340 ⇒ 00:54:15.250 Uttam Kumaran: What we like, you know?
430 00:54:15.810 ⇒ 00:54:16.630 Shivani Amar: Okay.
431 00:54:17.410 ⇒ 00:54:18.479 Shivani Amar: Thank you.
432 00:54:19.890 ⇒ 00:54:23.959 Uttam Kumaran: Perfect. Okay, thanks everyone. Jason, I’ll follow up on some of the Snowflake pieces with you.
433 00:54:25.780 ⇒ 00:54:26.850 Jason Wu: Great, thanks, Hotel.
434 00:54:27.740 ⇒ 00:54:28.310 Uttam Kumaran: Okay.
435 00:54:28.480 ⇒ 00:54:30.049 Uttam Kumaran: Alright, thank you, everyone.
436 00:54:30.050 ⇒ 00:54:30.580 Shivani Amar: Thank you.
437 00:54:30.580 ⇒ 00:54:31.319 Uttam Kumaran: Appreciate it.
438 00:54:31.810 ⇒ 00:54:32.170 Steve Sizer: Thank you.
439 00:54:32.170 ⇒ 00:54:32.680 Uttam Kumaran: I…