Meeting Title: Magic Spoon — Brainforge sync Date: 2026-01-23 Meeting participants: Demilade Agboola, Mary Burke, Uttam Kumaran, Ashwini Sharma, Justin Tabarini, Joshua Levy, Michael Thorson
WEBVTT
1 00:00:08.160 ⇒ 00:00:09.099 Mary Burke: Hey guys!
2 00:00:11.640 ⇒ 00:00:12.310 Ashwini Sharma: Hello.
3 00:00:12.310 ⇒ 00:00:15.049 Demilade Agboola: I… How’s everyone doing?
4 00:00:16.560 ⇒ 00:00:17.829 Mary Burke: Good, how about you?
5 00:00:18.440 ⇒ 00:00:20.309 Demilade Agboola: Oh, pretty good, pretty good.
6 00:00:20.950 ⇒ 00:00:21.650 Demilade Agboola: -
7 00:00:21.860 ⇒ 00:00:22.790 Demilade Agboola: Okay.
8 00:00:25.060 ⇒ 00:00:27.759 Demilade Agboola: I… are we waiting for Michael?
9 00:00:28.750 ⇒ 00:00:33.419 Mary Burke: Yes, he should be joining, but I think we can… oh, there he is. Yeah, we can jump in and get started.
10 00:00:34.140 ⇒ 00:00:36.150 Demilade Agboola: Alright, let’s do this.
11 00:00:40.200 ⇒ 00:00:44.210 Demilade Agboola: There we go. Second… Bare my screen…
12 00:00:50.280 ⇒ 00:00:51.240 Demilade Agboola: Okay.
13 00:00:54.690 ⇒ 00:01:02.380 Demilade Agboola: So… This week, this is still our…
14 00:01:02.490 ⇒ 00:01:07.040 Demilade Agboola: North Star, and where we’re going towards, and so this, you know, keeps us…
15 00:01:07.460 ⇒ 00:01:09.630 Demilade Agboola: Focused on what our end goal is.
16 00:01:10.500 ⇒ 00:01:16.529 Demilade Agboola: So this week we’re just gonna talk about, like, the updates on the Spain’s API, progress.
17 00:01:16.990 ⇒ 00:01:23.299 Demilade Agboola: The long-running models and any, like, potential roadblocks and just future plans around that as well.
18 00:01:23.780 ⇒ 00:01:31.619 Demilade Agboola: So this week, we’ve largely been on this spot, where we’ve been trying to look at the data uploaded into Redshift.
19 00:01:31.770 ⇒ 00:01:35.220 Demilade Agboola: So, we have loaded data
20 00:01:35.420 ⇒ 00:01:40.719 Demilade Agboola: we got Q… we got some QA data into the Redshift.
21 00:01:41.240 ⇒ 00:01:55.000 Demilade Agboola: And that was what we used to allow us to, you know, load more data, and I know Ashwini has met with the team during the week, just to be able to figure out what columns to load into.
22 00:01:55.330 ⇒ 00:01:59.619 Demilade Agboola: To load, basically, from the API, and how to use that to QA.
23 00:01:59.980 ⇒ 00:02:04.819 Demilade Agboola: So right now, I know Ashwini has basically been able to load that data and…
24 00:02:05.190 ⇒ 00:02:15.520 Demilade Agboola: right now, we’re waiting for the Magic Spoon team to give us feedback on, like, if we’re good to go ahead. And once we get that feedback, we will be able to start, like, loading the data.
25 00:02:15.760 ⇒ 00:02:19.129 Demilade Agboola: Into, the red, into the redshift.
26 00:02:19.970 ⇒ 00:02:21.170 Demilade Agboola: going forward.
27 00:02:22.420 ⇒ 00:02:27.639 Demilade Agboola: Does anyone have any questions or anything around that?
28 00:02:30.440 ⇒ 00:02:43.449 Michael Thorson: Would love to know, so now that we have some data loaded into Warehouse, that looks good, would love to know kind of what the next steps are, in terms of actually validating that what’s in Warehouse is…
29 00:02:43.630 ⇒ 00:02:45.920 Michael Thorson: Accurate with that in-platform data.
30 00:02:46.110 ⇒ 00:02:47.890 Michael Thorson: Like, what do you need from us, I guess?
31 00:02:49.810 ⇒ 00:02:51.369 Demilade Agboola: So I can hand this over to Ashwini.
32 00:02:52.730 ⇒ 00:02:56.820 Ashwini Sharma: Okay, so, once you can confirm that we have what we need.
33 00:02:57.130 ⇒ 00:03:02.509 Ashwini Sharma: I can run a quick check against the production data, platform data that you have loaded.
34 00:03:02.830 ⇒ 00:03:05.589 Ashwini Sharma: And then verify that, you know,
35 00:03:06.000 ⇒ 00:03:08.449 Ashwini Sharma: The numbers that we are extracting, and…
36 00:03:08.980 ⇒ 00:03:12.470 Ashwini Sharma: And generating, based on whatever data is there.
37 00:03:12.630 ⇒ 00:03:15.710 Ashwini Sharma: Matches with what’s there in the platform data.
38 00:03:19.760 ⇒ 00:03:33.769 Michael Thorson: Sweet, and that makes sense versus, like, production and in-plat… or, like, API and, like, in-platform data. Can you speak a little bit to the validation of, like, the different levels of aggregation? Because I know we pulled out a few different
39 00:03:33.900 ⇒ 00:03:37.310 Michael Thorson: kind of layers and API call as well.
40 00:03:37.980 ⇒ 00:03:42.749 Ashwini Sharma: Right, so basically, like, what we did was, 52 weeks of data for…
41 00:03:42.860 ⇒ 00:03:46.410 Ashwini Sharma: Magic Spoon only as a brand, and
42 00:03:47.150 ⇒ 00:03:52.390 Ashwini Sharma: So we reduce the volume of data, right? Instead of 208 weeks or 228 weeks.
43 00:03:52.600 ⇒ 00:03:58.340 Ashwini Sharma: Did it only for 52 weeks. Now, what we can do is, we can aggregate For the current week.
44 00:03:58.670 ⇒ 00:04:01.410 Ashwini Sharma: And it’s, 52-week.
45 00:04:01.960 ⇒ 00:04:05.140 Ashwini Sharma: 24 weeks, 12 weeks, 4 weeks, data.
46 00:04:05.310 ⇒ 00:04:10.919 Ashwini Sharma: I think that model is already ready, so for any attribute, any measures that… that
47 00:04:11.610 ⇒ 00:04:13.490 Ashwini Sharma: You think should be there.
48 00:04:13.650 ⇒ 00:04:23.850 Ashwini Sharma: I can run the aggregation, as long as it is aggregatable, right? There are certain measures which were percentages, and probably that won’t,
49 00:04:24.770 ⇒ 00:04:27.549 Ashwini Sharma: The aggregation over that will not work.
50 00:04:27.890 ⇒ 00:04:31.440 Ashwini Sharma: We’ll have to figure out, how we can derive those things.
51 00:04:31.630 ⇒ 00:04:35.080 Ashwini Sharma: But if it is a simple number, then we can definitely aggregate it.
52 00:04:37.270 ⇒ 00:04:38.400 Justin Tabarini: One,
53 00:04:38.510 ⇒ 00:04:46.570 Justin Tabarini: like, I’d say, like, the practical use of this is to understand, like, what was our target sales in…
54 00:04:46.980 ⇒ 00:04:49.320 Justin Tabarini: A 4-week period or something?
55 00:04:49.490 ⇒ 00:05:00.369 Justin Tabarini: I think, like, I don’t know if you guys are treating that as part of the data modeling side, or on the data validation side. Like, that’s a… like,
56 00:05:00.580 ⇒ 00:05:14.680 Justin Tabarini: those… we view those as, like, part of the QA. It’s like, okay, we’re able to pull target sales by region for these last, whatever, 4 weeks, whatever… whatever data points you want to do, and then we’re also able to do that
57 00:05:15.150 ⇒ 00:05:18.729 Justin Tabarini: On, like, a 52-week, or, like, a one-week, like.
58 00:05:19.010 ⇒ 00:05:25.000 Justin Tabarini: We can do it by different ones, we can look at the total… total MULO, look at the levels of sales, so it’s like…
59 00:05:25.150 ⇒ 00:05:29.699 Justin Tabarini: I think there’s, like, two parts, just, like, yes… Like, I guess, like, yeah.
60 00:05:30.790 ⇒ 00:05:31.860 Ashwini Sharma: So, yeah, I’ll be important.
61 00:05:31.860 ⇒ 00:05:32.600 Justin Tabarini: things.
62 00:05:32.900 ⇒ 00:05:48.410 Ashwini Sharma: So, I had done a data modeling some time back for the initial data that I had extracted into Redshift, and then we added more attributes. But, for the initial data, we can always aggregate at different week intervals, if they are numbers.
63 00:05:49.600 ⇒ 00:05:51.320 Ashwini Sharma: So, for example, the…
64 00:05:51.680 ⇒ 00:05:59.520 Ashwini Sharma: let’s say it’s a volume, right, or a dollar amount. That can be aggregated, but as you had pointed out during last call.
65 00:05:59.770 ⇒ 00:06:12.149 Ashwini Sharma: probably last to last call, that if it is a percentage, then that won’t be aggregated. So if you can tell me, right, you know, I mean, just by looking at the attribute, I won’t be able to tell if it is a numeric or it’s a derived,
66 00:06:12.410 ⇒ 00:06:13.810 Ashwini Sharma: Like, percentage or things.
67 00:06:13.810 ⇒ 00:06:19.400 Justin Tabarini: percentage… an example of percentage ACV, the weighting factor for ACV is market ACV.
68 00:06:20.180 ⇒ 00:06:21.500 Justin Tabarini: So, if you could use that…
69 00:06:21.500 ⇒ 00:06:22.440 Ashwini Sharma: I’ll look…
70 00:06:22.440 ⇒ 00:06:25.200 Justin Tabarini: So, like, aggregate income to values there.
71 00:06:25.750 ⇒ 00:06:28.119 Ashwini Sharma: One second, let me take a note here.
72 00:06:28.880 ⇒ 00:06:31.570 Justin Tabarini: just generally, I guess, like, right now, I feel like…
73 00:06:32.820 ⇒ 00:06:35.440 Justin Tabarini: What do you need from us to be able to do this QA?
74 00:06:36.840 ⇒ 00:06:44.150 Justin Tabarini: kind of what we want to make sure is correct, and what, like, the key outputs we expect from this model? Like, what do you need from us for you to get going?
75 00:06:44.580 ⇒ 00:06:48.239 Ashwini Sharma: Just give me some attributes which are,
76 00:06:49.310 ⇒ 00:06:53.870 Ashwini Sharma: For which we need to do the QA. Obviously, like, against each attribute might be difficult.
77 00:06:54.060 ⇒ 00:06:56.449 Ashwini Sharma: to do the QA, right? There’s a lot.
78 00:06:56.450 ⇒ 00:06:57.940 Justin Tabarini: More data? Yeah.
79 00:06:59.310 ⇒ 00:07:06.000 Justin Tabarini: We can speak to that real quick. Like, I think we’d like revenue TDP.
80 00:07:06.170 ⇒ 00:07:11.090 Justin Tabarini: ACV. Those are, like, 3 good attributes, like, measures.
81 00:07:11.470 ⇒ 00:07:18.800 Justin Tabarini: to use on the level of aggregation. I think different levels of aggregation we would like is by retailer… oh, so yeah, there we go.
82 00:07:18.800 ⇒ 00:07:21.640 Michael Thorson: Yeah, sorry, I had this noted somewhere.
83 00:07:21.890 ⇒ 00:07:25.379 Justin Tabarini: Yeah, so those, the ones in the chat, are the metrics.
84 00:07:25.380 ⇒ 00:07:25.980 Ashwini Sharma: Sorry.
85 00:07:26.250 ⇒ 00:07:31.819 Demilade Agboola: And if I remember correctly, TDP and ACV are derived, so it’s a function of…
86 00:07:32.180 ⇒ 00:07:34.950 Demilade Agboola: Like, it’s calculated.
87 00:07:36.800 ⇒ 00:07:39.550 Justin Tabarini: So technically, we think you can sum TDP?
88 00:07:40.330 ⇒ 00:07:44.069 Justin Tabarini: But that would also be something to help validate for us that that doesn’t work.
89 00:07:44.070 ⇒ 00:07:45.190 Michael Thorson: Yeah. Okay.
90 00:07:45.190 ⇒ 00:07:47.240 Justin Tabarini: you can definitely sum revenue, so, like.
91 00:07:47.240 ⇒ 00:07:47.950 Demilade Agboola: Yeah, revenue.
92 00:07:47.950 ⇒ 00:07:59.550 Justin Tabarini: if you’re able to sum… like, if you’re able to do two different levels of aggregation and sum revenue, and the revenues match, and then you sum the TDPs and they don’t match, that’s… that’s good to call out, just like, hey, TDP’s not working as we thought.
93 00:07:59.770 ⇒ 00:08:00.950 Ashwini Sharma: Yeah, and…
94 00:08:01.030 ⇒ 00:08:18.790 Michael Thorson: In this initial phase two, it’s like, if the answer is TDP does not sum correctly, just having some kind of, like, guidance on what you need to… to potentially, aggregate that. So, happy to, like, like, do that workshop and, like, dig into the documentation together, like.
95 00:08:18.820 ⇒ 00:08:21.779 Michael Thorson: This first round is really just, like, is the data looking?
96 00:08:21.900 ⇒ 00:08:23.130 Michael Thorson: Accurate, yeah.
97 00:08:23.130 ⇒ 00:08:25.060 Justin Tabarini: Where is it looking inaccurate?
98 00:08:25.110 ⇒ 00:08:26.740 Michael Thorson: Also.
99 00:08:27.350 ⇒ 00:08:31.249 Justin Tabarini: In terms of, like, aggregation, we’d love to know, like, a retailer.
100 00:08:31.710 ⇒ 00:08:36.419 Justin Tabarini: region, and then, like, total MULO, so, like, total businesses.
101 00:08:36.590 ⇒ 00:08:39.810 Justin Tabarini: And so, like, filtering for Magic Spoon’s a good way to do that.
102 00:08:40.169 ⇒ 00:08:47.050 Justin Tabarini: And then also by… is it called product category, or is it called subcategory, Michael? The, like, granola, cereal.
103 00:08:47.430 ⇒ 00:09:07.220 Michael Thorson: I think we’re, like, yeah, just to kind of summarize it back, it’s like, we want to aggregate… I think you already spokeish, we need to, like, timeframe aggregation. We also want to investigate, the… it’s called reporting level aggregations, like, does… do, like, UPCs sum up to the brand for Magic Spoon, and kind of, like, up that ladder?
104 00:09:08.830 ⇒ 00:09:16.659 Michael Thorson: Since you’re pulling that at different levels. And then… I think we also want to understand, like, are there any breakpoints in geography?
105 00:09:16.770 ⇒ 00:09:24.820 Michael Thorson: Yeah, just take a look at the geographies and, like, geo- like, account types to understand, like, can we roll up
106 00:09:24.930 ⇒ 00:09:31.640 Michael Thorson: From, like, the regional geographies up to, like, the total like, US geography.
107 00:09:31.840 ⇒ 00:09:40.829 Michael Thorson: Because it is multiple levels of geographic data, there’s a couple lines that should be, like, all sales of cereal in the United States, etc.
108 00:09:42.160 ⇒ 00:09:54.069 Ashwini Sharma: So, geography we filtered out, because geography is a part of filter that… mandatory filters, and in the sheet that you had earlier provided, a list of geographies to filter out.
109 00:09:54.430 ⇒ 00:10:03.879 Ashwini Sharma: So the data is extracted only for that set of geographies. I can… I can drop the list of geographies in… in Slack.
110 00:10:05.680 ⇒ 00:10:06.880 Ashwini Sharma: a second…
111 00:10:09.060 ⇒ 00:10:09.770 Justin Tabarini: Yeah.
112 00:10:14.540 ⇒ 00:10:22.029 Demilade Agboola: Also, would it be helpful to organize, like, a QA session on Monday, where we can, like, just, like, thrash through, like, a bunch of these things?
113 00:10:23.260 ⇒ 00:10:38.439 Justin Tabarini: Yeah, I think it’d be helpful if you guys, like, came with questions. Like, if you guys had the time to actually take a look at the data, start QAing, and then being like, oh, this isn’t matching, and then we kind of debug together. I think that works, but I’d like to have, like, a first…
114 00:10:39.530 ⇒ 00:10:40.970 Joshua Levy: Pass of what…
115 00:10:41.150 ⇒ 00:10:41.989 Justin Tabarini: That’s working now.
116 00:10:42.330 ⇒ 00:10:43.209 Demilade Agboola: That’s really cool.
117 00:10:43.210 ⇒ 00:10:50.670 Joshua Levy: working doc, where we have, like, all these outlined areas. You do, like, the initial, and then just brawl in the doc sort of thing, and can comment back and forth.
118 00:10:51.030 ⇒ 00:10:55.119 Justin Tabarini: Alright, sounds good, sounds good. Also happy to do it live, like, that’s not…
119 00:10:55.240 ⇒ 00:10:59.650 Justin Tabarini: Because it might be quicker, but I do think we want some pre-work.
120 00:10:59.940 ⇒ 00:11:15.329 Demilade Agboola: Yeah, potentially we could do, like, a combination of both, you know, get the document across to you by, say, Monday, Tuesday, and then maybe midweek on that Tuesday or Wednesday, we can have a working session where we’re just like, okay, so this is what we’ve seen so far.
121 00:11:15.440 ⇒ 00:11:22.569 Demilade Agboola: These are the disparities, and then we can just resolve things right there, so it’s faster to get the feedback that we’ll need.
122 00:11:23.330 ⇒ 00:11:23.950 Justin Tabarini: Perfect.
123 00:11:24.170 ⇒ 00:11:25.060 Michael Thorson: Okay. Alright.
124 00:11:26.730 ⇒ 00:11:29.959 Michael Thorson: Great. Yep, thanks for digging into that during today’s meeting, though.
125 00:11:30.080 ⇒ 00:11:31.749 Michael Thorson: Excited to see how it’s looking.
126 00:11:33.210 ⇒ 00:11:41.210 Demilade Agboola: Okay, alright, also, so in terms of, like, the long-running models and general, like, dbt.
127 00:11:41.580 ⇒ 00:12:00.710 Demilade Agboola: Inspection, so we’re doing internal reviews on the document that we’re going to send over to the team, so that should be done today. And we’ll send that over, like, noting everything we’ve seen, from tests to, like, documentation to, long-running models that we’ve talked about.
128 00:12:00.810 ⇒ 00:12:05.150 Demilade Agboola: And then in terms of… The actual, like.
129 00:12:05.420 ⇒ 00:12:18.470 Demilade Agboola: implementation, so that’s a work in progress, and the idea is by early next week, I’ll start to share, like, what I have done, what I’ve tested, what’s worked, what’s not been working, and that would also be the full
130 00:12:18.570 ⇒ 00:12:22.869 Demilade Agboola: Picture of both the auditing as well as the implementation.
131 00:12:23.570 ⇒ 00:12:36.549 Uttam Kumaran: Yeah, the… the blocker here is me. I… we got this done yesterday. Typically we… when we produce things like this, we kind of almost try to do a little bit of, like, narrative up front, so that this doc kind of, like.
132 00:12:36.670 ⇒ 00:12:44.680 Uttam Kumaran: can be reviewed by anyone, kind of, in the future that wants to look at, like, how we conducted the audit, and then it kind of goes into, like, all the things we found.
133 00:12:44.810 ⇒ 00:12:52.420 Uttam Kumaran: There’s certainly, like, some really, really ripe areas for optimization, so… should be out, like, today.
134 00:12:53.100 ⇒ 00:12:55.609 Mary Burke: Can we get any previews of where those areas are?
135 00:12:56.260 ⇒ 00:12:57.030 Mary Burke: Sneak peek?
136 00:12:57.030 ⇒ 00:13:03.099 Uttam Kumaran: Yeah, Demi, I don’t know if you want to just, like, talk through a couple of things that you saw, or… yeah, I can, whatever.
137 00:13:03.670 ⇒ 00:13:10.699 Demilade Agboola: Well, I mean, to be fair, I… like, I have done some of the sneak peeks with the, long-running models specifically, because.
138 00:13:11.170 ⇒ 00:13:23.609 Demilade Agboola: We did have a session where we talked about the things, and we looked into different models as to what we could do to optimize that. Potentially other areas of improvement would be, things around, like.
139 00:13:23.880 ⇒ 00:13:33.580 Demilade Agboola: add in sources to, like, sources.yaml files to the current infrastructure. That way, we can start to add, like, source freshness tests, so in case, like.
140 00:13:33.950 ⇒ 00:13:42.810 Demilade Agboola: the sources are not refreshing, we will be able to see from the DVT end quickly that, like, the data that is going into all of our transformations is stale.
141 00:13:42.850 ⇒ 00:13:56.589 Demilade Agboola: Also things around, like, documentation, so just being able to, like, name the different model, like, give details to different models, why things are named the way they are, or what’s going on in certain, like, business logic.
142 00:13:56.860 ⇒ 00:13:59.949 Demilade Agboola: Also things around, like, testing.
143 00:14:00.010 ⇒ 00:14:18.799 Demilade Agboola: So the idea is, right now, there’s no, like… even if it’s just basic things around primary keys, restrictions. So right now, if you have a primary key that has a… that is… that feels uniqueness, or, there are null values in your primary key, you would have no idea right now.
144 00:14:18.970 ⇒ 00:14:32.750 Demilade Agboola: And obviously, these are, like, no-no’s in, like, just the general sense of data. So those are, like, the high-level things that were noted in the documentation. Like I said, the document is basically…
145 00:14:33.310 ⇒ 00:14:41.319 Demilade Agboola: done. It’s just, like, final internal review things before we send out, because we try to ensure that whatever goes out is of high quality.
146 00:14:42.010 ⇒ 00:14:47.770 Uttam Kumaran: Yeah, I think you’re gonna see that, like, any model that’s running… I mean, for me, usually my rule of thumb is, like.
147 00:14:47.960 ⇒ 00:14:53.220 Uttam Kumaran: Anything that’s more than, like, 5 minutes, that’s usually, like, something we can do.
148 00:14:53.260 ⇒ 00:15:08.899 Uttam Kumaran: There’s also probably models, in between, like, the raw data and, like, the final tables that may not need to be materialized, like, they can be moved to ephemeral. But, like, we saw models that are running for, like, 15, 20, 40 minutes, like.
149 00:15:08.900 ⇒ 00:15:13.969 Uttam Kumaran: Huge opportunity to just, like, immediately turn those incremental and… and get those…
150 00:15:13.970 ⇒ 00:15:28.169 Uttam Kumaran: like, streamlined, and then, yeah, I think a lot of, like, what we’ll recommend is more, like, some stuff that’s, like, really preventative, so you have to… basically, like, how are… how do you know that the data coming in is not adding duplicates, basically?
151 00:15:28.440 ⇒ 00:15:33.299 Uttam Kumaran: And then in order to do the incremental as well, you’ll need to have those primary keys.
152 00:15:33.760 ⇒ 00:15:38.939 Uttam Kumaran: But I feel like the overall, like, repro structure was fine, you know?
153 00:15:41.320 ⇒ 00:15:43.789 Mary Burke: Okay, great. I’m excited to, excited to see the doc.
154 00:15:43.890 ⇒ 00:15:59.300 Mary Burke: And I think, too, when, like, moving forward, thinking through the kind of improvements we can make to our transformation layer, making sure that that knowledge transfer is really, really clear, too, so I think there could be questions that come up with that, so we’re always happy to help out there.
155 00:16:01.190 ⇒ 00:16:02.710 Demilade Agboola: Yeah,
156 00:16:02.810 ⇒ 00:16:15.169 Demilade Agboola: We’re all huge advocates of, like, knowledge transfer, and in fact, it’s part of… we also have some stuff around that here as well, in this. So yeah, we will always, like, find ways in which we can close whatever gaps.
157 00:16:15.250 ⇒ 00:16:23.329 Demilade Agboola: Between, the knowledge that’s generated and knowledge that would need to be transferred to whoever, will be hopping on any project or hopping on the team in the future.
158 00:16:24.280 ⇒ 00:16:32.209 Uttam Kumaran: Yeah, and I know, Ashwini, you also worked on adding a lot of the documentation to how we did the spins pipeline directly in the repo.
159 00:16:32.360 ⇒ 00:16:51.720 Uttam Kumaran: I know, like, we kind of walked into Prefect sort of being something the previous team set out, so I was like, hey, as we… as we go through the journey, like, write everything down so it sits in the repo, basically, so anytime you need to write another pipeline, like, a lot of stuff is clear there, so… and we’re committing that, like, right to the repo.
160 00:16:51.810 ⇒ 00:16:55.100 Uttam Kumaran: So that… You know, whoever checks that out next can see that.
161 00:16:58.400 ⇒ 00:17:05.580 Demilade Agboola: Yeah, so, like, this part, that’s all of this here. So the link is here, but it’s also in the repo directly, so you can always see it.
162 00:17:05.829 ⇒ 00:17:11.380 Demilade Agboola: But yeah, so the idea was just so we can have some transparency and just easy knowledge transfer.
163 00:17:11.670 ⇒ 00:17:15.960 Demilade Agboola: We’ve been able to document, like, what we’ve been doing so far, and just, like.
164 00:17:16.069 ⇒ 00:17:32.609 Demilade Agboola: the knowledge that we’ve gained around the, Spins Data API. So, again, in the future, anyone who wants to work on this or pick up anything can quickly see what’s worked, what’s not worked, and just how we’ve gone about the entire process. So, here when we started from zero.
165 00:17:32.770 ⇒ 00:17:34.780 Demilade Agboola: But from, you know, a much higher level.
166 00:17:38.310 ⇒ 00:17:39.410 Mary Burke: Great, thank you.
167 00:17:40.030 ⇒ 00:17:44.390 Demilade Agboola: You’re welcome. Also, just, like, just general steps.
168 00:17:44.800 ⇒ 00:18:00.490 Demilade Agboola: I guess, like, in terms of the transition, ultimately, we’ll need to just have an idea of, like, what priorities are. Obviously the… we’ll be trying to, like, tie up the Spins API and the Data Mart
169 00:18:00.660 ⇒ 00:18:09.430 Demilade Agboola: As well, based off of that. But ultimately, just kind of knowing, like, the direction in which you want us to be able to maintain and take over.
170 00:18:09.720 ⇒ 00:18:16.880 Demilade Agboola: Will be very important, and obviously the… the MMM data model would obviously be very high priority.
171 00:18:17.480 ⇒ 00:18:22.780 Justin Tabarini: Yeah, can I, can I share my screen and… Show what we have ready?
172 00:18:26.580 ⇒ 00:18:30.150 Justin Tabarini: So we have everything… We need there.
173 00:18:30.800 ⇒ 00:18:33.840 Justin Tabarini: So I think I shared this last week, but it has,
174 00:18:34.400 ⇒ 00:18:47.979 Justin Tabarini: So here, we only want to include the yeses on what we’re including in the data marked. And here I have the marked source. I leave left spins there as, like, generically, we’re gonna want these variables.
175 00:18:48.050 ⇒ 00:18:54.900 Justin Tabarini: don’t know what the data mart has been finalized, the name. I just didn’t leave any data notes either, just because
176 00:18:55.250 ⇒ 00:18:58.509 Justin Tabarini: You guys will know exactly where to pull that data from once you have it.
177 00:19:01.340 ⇒ 00:19:11.239 Justin Tabarini: The rest of it, I have this… the mark that exists now. I kind of call out, like, oh, maybe it’s in a CSV already, so here…
178 00:19:11.430 ⇒ 00:19:13.610 Justin Tabarini: If you, like, click on the links here.
179 00:19:13.780 ⇒ 00:19:32.409 Justin Tabarini: Oh, so not all these links are working. It’s the exact same, I shared this file with you guys, it’s called the data inputs file. This has all the CSVs we might need to upload to the data warehouse, and then include in the MART, so I have, like, there’s no MART because you’re gonna upload the CSV, but this describes the columns and how you would categorize it.
180 00:19:33.160 ⇒ 00:19:35.550 Justin Tabarini: To match the different metric.
181 00:19:36.880 ⇒ 00:19:45.499 Demilade Agboola: Also, just to clarify, just the CSVs, are they, like, the CSV files, are they being updated, or are they static?
182 00:19:45.500 ⇒ 00:19:46.280 Justin Tabarini: Static.
183 00:19:46.300 ⇒ 00:19:50.619 Demilade Agboola: We still need… right now, the plan for this is, like, we’re in a rush.
184 00:19:50.810 ⇒ 00:20:08.070 Justin Tabarini: We want to get this out to our partner ASAP. So, we’re doing everything via CSEs, just getting the one-time polls, seeing which variables they end up using, and then figuring out how to automate those pipelines. So, those are tasks that are kind of coming.
185 00:20:08.560 ⇒ 00:20:09.259 Demilade Agboola: Okay.
186 00:20:09.530 ⇒ 00:20:12.039 Justin Tabarini: But… Right now, just kind of…
187 00:20:13.480 ⇒ 00:20:19.780 Justin Tabarini: Have all the details you should need, kind of listed out here, of what to do, how to categorize them.
188 00:20:20.050 ⇒ 00:20:22.479 Justin Tabarini: Where the data mart is, if it already exists.
189 00:20:23.860 ⇒ 00:20:27.679 Justin Tabarini: CSV, sorry, I miss… mistyped that one.
190 00:20:28.160 ⇒ 00:20:34.619 Justin Tabarini: But yeah, just Slack me if you have questions. I have, like, really good detail on all of this, but it should be super clear.
191 00:20:34.890 ⇒ 00:20:35.530 Demilade Agboola: Okay.
192 00:20:35.780 ⇒ 00:20:41.950 Justin Tabarini: where everything is, there’s a lot of CSVs you need to upload, and then categorize, but they should have
193 00:20:43.130 ⇒ 00:20:44.290 Justin Tabarini: the details.
194 00:20:44.540 ⇒ 00:20:53.209 Justin Tabarini: like, of which ones to use for spends, like, it should be self-explanatory. So using JT data notes, the data mart, and the metric will be…
195 00:20:53.520 ⇒ 00:20:54.530 Justin Tabarini: Pretty helpful.
196 00:20:57.460 ⇒ 00:21:00.370 Justin Tabarini: So I just shared both docs with you, so feel free to…
197 00:21:00.530 ⇒ 00:21:03.059 Justin Tabarini: I shared the file and the doc with you, so…
198 00:21:04.110 ⇒ 00:21:05.660 Demilade Agboola: Feel free to roll.
199 00:21:05.970 ⇒ 00:21:10.249 Demilade Agboola: I have seen it. Sounds great. So this is stuff we’ll start looking at for them.
200 00:21:10.430 ⇒ 00:21:12.060 Demilade Agboola: Next week,
201 00:21:12.630 ⇒ 00:21:18.610 Demilade Agboola: We will, like, internally just, like, figure out how we want to go about it, and just, like.
202 00:21:18.800 ⇒ 00:21:21.999 Demilade Agboola: Create tickets around it and start to, like, chip away at it.
203 00:21:22.470 ⇒ 00:21:25.599 Justin Tabarini: Yeah, what would be estimated timeline on this?
204 00:21:26.920 ⇒ 00:21:29.650 Demilade Agboola: The total mart, or just in…
205 00:21:30.070 ⇒ 00:21:32.730 Demilade Agboola: like, because the Spins API is still part of it.
206 00:21:32.730 ⇒ 00:21:44.920 Justin Tabarini: Yep, yes, exactly. The one thing we were talking about is if we could get the spend data, just all the, like, the stuff that I have tagged below, that’s available and ready to be dealt with.
207 00:21:45.230 ⇒ 00:21:51.040 Justin Tabarini: then that would be… Very helpful to already have that.
208 00:21:51.210 ⇒ 00:21:53.029 Uttam Kumaran: All the CSV stuff.
209 00:21:53.390 ⇒ 00:21:58.619 Justin Tabarini: Like, all the, like, basically, everything I’ve listed in that doc, if we could build a mart without revenue.
210 00:21:58.700 ⇒ 00:22:14.109 Justin Tabarini: That would be helpful, because then we can have a second conversation with our partners on, like, okay, which ones do we want to exclude, which ones do we want to include? Like, the revenue stuff will come, and we will need it for the MMM, and I know that Pipelines, like, has a longer lead time.
211 00:22:14.170 ⇒ 00:22:18.880 Justin Tabarini: The spend stuff would be helpful, because we can continue conversations.
212 00:22:19.180 ⇒ 00:22:21.779 Justin Tabarini: While we’re still working on getting the revenue in.
213 00:22:22.000 ⇒ 00:22:31.340 Demilade Agboola: I mean, for the spend stuff, I mean, unless, like, I will start looking at it, and unless there’s any, like, unforeseen blocker, I believe there should be stuff we can get done by the end of next week.
214 00:22:31.340 ⇒ 00:22:31.680 Justin Tabarini: Okay.
215 00:22:31.680 ⇒ 00:22:36.200 Demilade Agboola: unless it’s an unforeseen blocker, then I’m like, oh, I did not see that coming, but…
216 00:22:36.200 ⇒ 00:22:38.430 Uttam Kumaran: Most of the filters look pretty…
217 00:22:38.430 ⇒ 00:22:39.210 Justin Tabarini: Yes.
218 00:22:39.620 ⇒ 00:22:41.079 Demilade Agboola: That one’s pretty clear, yeah.
219 00:22:41.280 ⇒ 00:22:42.360 Uttam Kumaran: Yeah.
220 00:22:42.360 ⇒ 00:22:45.170 Demilade Agboola: And then it will just be uploading the CSVs.
221 00:22:45.580 ⇒ 00:22:48.900 Demilade Agboola: To the warehouse, and then just kind of modeling away.
222 00:22:49.660 ⇒ 00:22:58.220 Justin Tabarini: Yeah, I think I… I think it is pretty clear. If there’s any uncertainty, just Slack. Like, I’ve been… I built that doc, so it’s like, I… I know what we want to do.
223 00:22:58.220 ⇒ 00:22:59.730 Demilade Agboola: Okay, alright, sounds good.
224 00:23:00.190 ⇒ 00:23:03.769 Justin Tabarini: So, questions I can answer quickly. Thanks!
225 00:23:03.790 ⇒ 00:23:05.120 Demilade Agboola: Alright then, sounds good.
226 00:23:06.370 ⇒ 00:23:06.930 Demilade Agboola: Okay.
227 00:23:06.930 ⇒ 00:23:10.559 Uttam Kumaran: Yeah, so maybe, Demi, let’s see where we are by, like, Wednesday next week.
228 00:23:11.090 ⇒ 00:23:12.780 Uttam Kumaran: And we can give…
229 00:23:13.470 ⇒ 00:23:22.769 Uttam Kumaran: like, we can give team a little bit of, like, okay, how we expect to get to Friday, and then if we need to, like, prioritize certain pieces, we can… we can get their feedback on that.
230 00:23:23.390 ⇒ 00:23:24.720 Demilade Agboola: Okay, alright, sounds good.
231 00:23:27.000 ⇒ 00:23:34.029 Mary Burke: Yeah, so it might be worthwhile to have a few, a few touchpoints next week, too, on the SPINS QA, and then also on the Mart
232 00:23:34.180 ⇒ 00:23:35.829 Mary Burke: Prior to the SPIN’s data.
233 00:23:36.490 ⇒ 00:23:54.479 Uttam Kumaran: Yeah, I think Spin’s QA will probably end up, like, Monday. We’ll probably have a version of that that we’ll walk through, and, like, do an R pass of, like, what looks weird, and then we’ll send it to y’all for, like, does everything look okay? And then we’ll pass it. And then again, if, like, if that is able to be done by Tuesday, then we’ll
234 00:23:55.040 ⇒ 00:24:02.680 Uttam Kumaran: try to wrap the whole thing up, like, that whole piece is part of this, but… heard you, Justin, on, like, what’s the priority as part of that, so…
235 00:24:08.000 ⇒ 00:24:13.520 Demilade Agboola: Okay, does anyone have any other, like, questions or feedback or anything of that nature?
236 00:24:17.850 ⇒ 00:24:23.589 Michael Thorson: just… I’m not sure if we close this loop. Are you all good to, like.
237 00:24:23.810 ⇒ 00:24:26.329 Michael Thorson: with your DBT connection, Debbie?
238 00:24:26.330 ⇒ 00:24:27.440 Demilade Agboola: Yeah, yeah, so…
239 00:24:27.440 ⇒ 00:24:27.900 Michael Thorson: Beautiful.
240 00:24:27.900 ⇒ 00:24:29.190 Demilade Agboola: Fine, everything’s all good.
241 00:24:29.600 ⇒ 00:24:31.660 Michael Thorson: And you’re, yeah, you’re able to write tables and everything.
242 00:24:31.660 ⇒ 00:24:32.360 Demilade Agboola: Yeah.
243 00:24:32.520 ⇒ 00:24:36.359 Michael Thorson: Great. Cool. So now, that’s… that’s it, just tying up that end.
244 00:24:39.490 ⇒ 00:24:43.119 Uttam Kumaran: Mary, do you want to stay on for a sec, and then we can talk about contract stuff?
245 00:24:43.670 ⇒ 00:24:44.600 Mary Burke: Sure, yeah.
246 00:24:44.900 ⇒ 00:24:45.540 Uttam Kumaran: Okay.
247 00:24:46.880 ⇒ 00:24:52.799 Demilade Agboola: Alright then, so if we don’t necessarily have any, like, questions or feedback, I guess we can call it a day.
248 00:24:54.640 ⇒ 00:24:56.210 Mary Burke: Great, thanks, team.
249 00:24:56.490 ⇒ 00:24:57.140 Uttam Kumaran: Thank you.
250 00:24:57.140 ⇒ 00:24:57.790 Demilade Agboola: Yeah.
251 00:25:02.490 ⇒ 00:25:03.310 Uttam Kumaran: Awesome.
252 00:25:03.740 ⇒ 00:25:06.199 Uttam Kumaran: How’s the, weather where you’re at, Mary?
253 00:25:06.600 ⇒ 00:25:11.630 Mary Burke: It’s fine right now, but we’re supposed to get a foot of snow that everyone’s, like, freaking out about, but…
254 00:25:11.630 ⇒ 00:25:20.359 Uttam Kumaran: Yeah, in Texas, too, like, there’s this… every, like, 5 years, there’s, like, an ice storm, and everybody’s, like… everybody here is, like, freaking out.
255 00:25:20.360 ⇒ 00:25:21.390 Mary Burke: Yeah, it’s like… Yeah.
256 00:25:21.390 ⇒ 00:25:26.790 Uttam Kumaran: The most casual thing here, but because everybody knows others freak out and crash.
257 00:25:27.250 ⇒ 00:25:34.430 Uttam Kumaran: this, like, vicious cycle. I was telling Demi about it this morning, because Demi goes to Minnesota a lot, like, back and forth, and he’s like.
258 00:25:34.430 ⇒ 00:25:35.040 Mary Burke: Hmm.
259 00:25:35.300 ⇒ 00:25:40.269 Uttam Kumaran: He’s like, alright, just like a cute little ice storm. They know how to do snow there, yeah.
260 00:25:40.270 ⇒ 00:25:51.749 Mary Burke: No, one of my… one of my sisters lives in Tennessee, and she… but she grew up in the Northeast with all of us, and she’s freaking out about, like, the 6 inches of snow they’re gonna get, and went out and bought a generator. I was like, I think you’re gonna be fine.
261 00:25:51.750 ⇒ 00:25:55.780 Uttam Kumaran: Well… Well, the problem is, because those places don’t…
262 00:25:56.320 ⇒ 00:25:59.799 Uttam Kumaran: Have, like, this peak demand, typically at this time of the year.
263 00:25:59.900 ⇒ 00:26:01.190 Uttam Kumaran: It’s actually, like.
264 00:26:01.420 ⇒ 00:26:06.949 Uttam Kumaran: like, in Austin, the power grid just went out, and, like, some people, like, died. It went out for two days.
265 00:26:06.950 ⇒ 00:26:07.950 Mary Burke: 28?
266 00:26:07.950 ⇒ 00:26:12.460 Uttam Kumaran: 2020, or 2020, yeah, I think, yeah. So people here really…
267 00:26:12.460 ⇒ 00:26:13.140 Mary Burke: Yeah, that.
268 00:26:13.140 ⇒ 00:26:14.150 Uttam Kumaran: Freak out.
269 00:26:15.050 ⇒ 00:26:25.789 Uttam Kumaran: I’m like, well, I’ll just hop in the car, like, we can drive somewhere, or, like, I don’t know. But I’m also used to, like, East Coast, where I’m like, life goes on, like, what are you gonna do, you know?
270 00:26:27.140 ⇒ 00:26:29.210 Mary Burke: Oh, well, here’s hoping it’s tame.
271 00:26:29.210 ⇒ 00:26:41.480 Uttam Kumaran: Yeah, yeah. Cool, so I know we’re kind of coming up on contract end of this month. I just wanted to kind of, like, see where your head was at, and, like, I know we talked about a few other sources, it seems like we’re still probably…
272 00:26:41.530 ⇒ 00:26:56.339 Uttam Kumaran: like, two… two to three weeks away, maybe, from nailing, like, this initial piece. Audit, you know, basically should be out the door today. But, like, yeah, what do you think about, like, longer-term stuff, or… yeah, like, we’d love to hear that.
273 00:26:56.450 ⇒ 00:27:01.529 Mary Burke: Yeah, I think we’ve been, pretty impressed with the partnership so far. I know that a lot of,
274 00:27:01.530 ⇒ 00:27:19.300 Mary Burke: the team, like, for Michael and JT, there’s a lot of emphasis on the MMM and MMT work, and just getting that all set up, so I think we just want to make sure we can see that through all the way before, like, signing on to the next contract, just to make sure that that goes well, but assuming that all goes well, I think we’re,
275 00:27:19.440 ⇒ 00:27:26.590 Mary Burke: we are happy to… to continue working with you guys, and I think our focus will be implementing a lot of the…
276 00:27:26.810 ⇒ 00:27:42.279 Mary Burke: the kind of quick fixes that can come from the audit, too, and then just… I think with our, with our existing partners, some of the things that are… I think we mentioned this in our initial call, too, just the low-hanging fruit of, like, we don’t know when a pipeline breaks or times out.
277 00:27:42.280 ⇒ 00:27:42.800 Uttam Kumaran: Yes.
278 00:27:42.800 ⇒ 00:27:45.190 Mary Burke: Find out 3 months later that we’re missing, like, a few.
279 00:27:45.190 ⇒ 00:27:45.610 Uttam Kumaran: Yes.
280 00:27:45.610 ⇒ 00:27:48.820 Mary Burke: data, and it’s throwing off our financials.
281 00:27:48.820 ⇒ 00:27:49.410 Uttam Kumaran: Yes.
282 00:27:49.410 ⇒ 00:27:56.049 Mary Burke: getting those things, like, that initial QA, which isn’t like the fun stuff, but getting some of that done and set up.
283 00:27:56.050 ⇒ 00:27:58.170 Uttam Kumaran: No, it’s a lot of, like, insurance policy.
284 00:27:58.170 ⇒ 00:27:58.630 Mary Burke: Yeah.
285 00:27:58.630 ⇒ 00:28:00.940 Uttam Kumaran: it’s this… I mean, both Demi and I have worked
286 00:28:01.110 ⇒ 00:28:15.390 Uttam Kumaran: like, where you’re as a seat for a long time, and you just, like, really need to know that that’s happening, or dupes are entering, because someone else will find out, and then it’s like, oh god, like, how long has this been happening? What’s the impact? And so that’s exactly, like, for us.
287 00:28:15.390 ⇒ 00:28:29.189 Uttam Kumaran: that’s… that’s what we do. So we’ll go in and put tests everywhere, and also bring in, like, Slack learning, so we’ll start to get flagged when things are going wrong. Yeah. And then, I feel like you mentioned, like, look, people aren’t doing, like, tons of
288 00:28:29.220 ⇒ 00:28:42.979 Uttam Kumaran: need for, like, hourly data, maybe during some peak seasons, but for the most part, just some of these jobs, like, these are just things you have to, like, sort of curtail every, like, quarter. Like, jobs will just start taking longer, and it’s something just to take a look at.
289 00:28:43.020 ⇒ 00:28:51.739 Uttam Kumaran: But for the most part, I feel like things are good. I think, yeah, we’ll layer on tests. And then it’s really, like, I think the prefect ownership. I think I was surprised to see that, like.
290 00:28:51.770 ⇒ 00:29:04.149 Uttam Kumaran: okay, this is kind of, like, running on its own, and then there are some hiccups in there, and so I just want to make sure we… we, like, have ownership… we kind of, like, give you guys ownership over that, and, like, you guys really understand, like, how those pipelines are working, and…
291 00:29:04.200 ⇒ 00:29:08.249 Uttam Kumaran: any optimizations, and it’s easy to add more, right? So it’s not this, like, black box.
292 00:29:08.460 ⇒ 00:29:24.469 Mary Burke: Yeah, especially, too, I think, there’s definitely appetite from the team to be more, be more, like, in the weeds themselves, and be able to self-serve a little bit, too. So we’d love to… we’d love to take some information from you guys, too, on how we can do this, and how we can do things correctly.
293 00:29:24.860 ⇒ 00:29:25.820 Uttam Kumaran: Okay, okay.
294 00:29:25.950 ⇒ 00:29:36.580 Uttam Kumaran: Okay, cool, so then how about what I’ll do is I’ll probably just draft something that I can run by you next week, Mary, and then we’re gonna still… I think, Demi, let’s try to drive as much as we can out next week.
295 00:29:36.940 ⇒ 00:29:46.869 Uttam Kumaran: And then we can kind of collaborate on that doc, and then see if we can get something over. You know, I think that’s a good plan. And then, yeah, we’ll consider the prefect stuff. Yeah, yeah.
296 00:29:47.290 ⇒ 00:29:59.690 Mary Burke: If our team’s, like, a bottleneck at all, or if it’s… like, we’re always happy to jump on and talk through things live, too, if there are open-ended questions. I know that the spins data can be finicky at all those different reporting levels, and the.
297 00:29:59.690 ⇒ 00:30:15.119 Uttam Kumaran: Yeah, and I think they’ve… I think we’ve been kind of working… it’s been, I think, a lot more tighter since, like, the holidays, basically, and, like, we’ve kind of… all through the access issues, so yeah, I feel like we’re… we’re good there, so I’ll just try to get something over to you early next week, and then…
298 00:30:15.230 ⇒ 00:30:20.609 Uttam Kumaran: you know, I think we can think about it, yeah, if there’s any other, like… we have the dbt stuff, we have the…
299 00:30:20.760 ⇒ 00:30:38.000 Uttam Kumaran: prefect stuff, you know, we have continuing on, like, anything for MMM, and then, yeah, I guess I would love to hear if there’s any other… I have some from our earlier conversations about other sources and other models, so, like, I’ll just, like, weave that in, and we can maybe work on, like, what’s priority when, you know.
300 00:30:38.220 ⇒ 00:30:45.269 Mary Burke: Yeah, I think we have… yeah, there are a ton of improvements we can make, too, and I… there won’t be a shortage of things to do there.
301 00:30:45.450 ⇒ 00:30:52.199 Uttam Kumaran: Yeah, so that’s why if there’s, like… now that Ashwini’s, like, knee-deep in prefect stuff, it’s easy for him to kind of, like, get pipelines out and, like.
302 00:30:52.230 ⇒ 00:31:03.429 Uttam Kumaran: two-week span. I mean, he’s gonna… he’s not gonna like for me to say that, but I’m like, yeah, now that he’s in there, and, like, we’re able to ship things, like, I feel like roughly one to two weeks per connector seems, like, fair.
303 00:31:03.430 ⇒ 00:31:13.829 Uttam Kumaran: For stuff that’s not, like, super crazy. So, like, I want to start to knock those out, because then immediately we can hand it to team for modeling. You know, so that, like, without getting it landed, like, really…
304 00:31:14.030 ⇒ 00:31:15.120 Mary Burke: jammed up.
305 00:31:15.720 ⇒ 00:31:16.220 Mary Burke: Yeah, I think.
306 00:31:16.220 ⇒ 00:31:16.780 Uttam Kumaran: Yeah.
307 00:31:17.080 ⇒ 00:31:23.139 Mary Burke: I also think a lot of the improvements, too, can come from that dbt layer, especially with, like, our.
308 00:31:23.140 ⇒ 00:31:23.660 Uttam Kumaran: Yeah.
309 00:31:23.660 ⇒ 00:31:28.610 Mary Burke: I don’t know if you saw, our Shopify marts are kind of a mess, and making sure that that’s…
310 00:31:28.820 ⇒ 00:31:35.039 Mary Burke: we’re doing that correctly and making it clear, and there’s not… like, right now, our refunds aren’t coming in at all, and I don’t know why.
311 00:31:35.040 ⇒ 00:31:35.430 Uttam Kumaran: Okay.
312 00:31:35.430 ⇒ 00:31:44.919 Mary Burke: That’s been, like, an open ticket for a few weeks now, so it’s just, like, some of these, like, little pieces of just where we can improve. But I know I’m gonna get too detailed there, but…
313 00:31:44.920 ⇒ 00:32:00.880 Uttam Kumaran: No, no, that’s actually really helpful, like, if there are actually existing issues in each of those marts that you’re like, this is a huge flag, we should tackle that, like, that’s what would be great to hear, because that’s actually, I would say, more of a priority than, like, anything that’s, like, maybe net new.
314 00:32:00.880 ⇒ 00:32:01.990 Mary Burke: Yeah, yeah, I think just…
315 00:32:01.990 ⇒ 00:32:02.490 Uttam Kumaran: No.
316 00:32:02.490 ⇒ 00:32:06.270 Mary Burke: Making sure what we have is… is working correctly.
317 00:32:06.270 ⇒ 00:32:06.960 Uttam Kumaran: Yeah.
318 00:32:06.960 ⇒ 00:32:24.000 Mary Burke: And then we… I, like, our website used to be a lot more complicated with our, like, we have a custom bundling on the website, and we can put different product categories in one product and sell it for different prices, and we have, like, a lot of customized logic to account for that on the BI side, so making sure.
319 00:32:24.000 ⇒ 00:32:24.640 Uttam Kumaran: Yeah.
320 00:32:24.640 ⇒ 00:32:32.680 Mary Burke: Like, you guys can look at it and know what’s going on, and that we know what’s going on, too, and making sure that we’re aligned to how we’ve been defining things.
321 00:32:32.880 ⇒ 00:32:39.319 Uttam Kumaran: Okay, yeah, Demi, what we’ll do is we’ll just walk through end-to-end, and then we’ll start to consolidate logic, build some macros, and things like that.
322 00:32:39.730 ⇒ 00:32:40.400 Demilade Agboola: Okay.
323 00:32:41.250 ⇒ 00:32:44.379 Demilade Agboola: Alright, sounds good. Yeah, definitely. We’ll look into all of that.
324 00:32:44.760 ⇒ 00:32:52.220 Mary Burke: Awesome. Thank you guys. We really appreciate it, and everything’s been going great. So we’re excited to see the audit review, and then get the MMM stuff rolling.
325 00:32:52.850 ⇒ 00:32:55.839 Uttam Kumaran: Perfect. Alright, thank you, man. Thank you, guys.
326 00:32:56.130 ⇒ 00:32:57.530 Uttam Kumaran: Talk to you soon. Bye. Bye.