Meeting Title: Magic Spoon Biweekly sync Date: 2026-04-03 Meeting participants: Demilade Agboola, Mary Burke, Michael Thorson
WEBVTT
1 00:01:32.470 ⇒ 00:01:33.649 Demilade Agboola: You’re a sister.
2 00:01:45.170 ⇒ 00:01:46.610 Demilade Agboola: I was just curious.
3 00:01:50.420 ⇒ 00:01:52.180 Demilade Agboola: Good morning.
4 00:01:59.850 ⇒ 00:02:01.199 Mary Burke: Hey Demi, how are you?
5 00:02:01.730 ⇒ 00:02:03.769 Demilade Agboola: Hi, Mary, how are you? I’m doing very well.
6 00:02:04.540 ⇒ 00:02:05.940 Mary Burke: Good, glad to hear it!
7 00:02:06.520 ⇒ 00:02:07.860 Demilade Agboola: Long time no see.
8 00:02:08.060 ⇒ 00:02:17.380 Mary Burke: I know, we haven’t chatted in a while, so I’m glad we had this… had this time just to catch up, check in. I know you met with Michael earlier this week, too, and he said that was really productive, so thank you for that.
9 00:02:17.380 ⇒ 00:02:24.210 Demilade Agboola: Oh, yeah, no, no problem. Always available to help, talk through different ideas, and open up,
10 00:02:24.500 ⇒ 00:02:28.130 Demilade Agboola: The world of, like, data to… the team.
11 00:02:28.340 ⇒ 00:02:37.630 Demilade Agboola: Just try and help maximize, like, the resources we have, and just utilize a lot of things. It was nice to be able to think about ways we could use dbt, and…
12 00:02:38.240 ⇒ 00:02:44.379 Demilade Agboola: For the, like, add value to, like, the Magic Spoon space.
13 00:02:45.660 ⇒ 00:02:47.729 Mary Burke: Yeah, no, thank you, we really appreciate it.
14 00:02:47.730 ⇒ 00:02:51.400 Demilade Agboola: No, it’s… it’s no… it’s no problem.
15 00:02:52.630 ⇒ 00:02:54.649 Demilade Agboola: I guess it’s two of us for this call.
16 00:02:56.810 ⇒ 00:02:59.089 Michael Thorson: Yeah, I don’t think JT can make it.
17 00:02:59.090 ⇒ 00:03:01.010 Demilade Agboola: Yeah, I think I saw him decline the call.
18 00:03:01.300 ⇒ 00:03:03.620 Michael Thorson: Yeah, he’s busy. He’s a busy one.
19 00:03:04.030 ⇒ 00:03:04.730 Demilade Agboola: Okay.
20 00:03:05.970 ⇒ 00:03:19.980 Demilade Agboola: Alright, so I’ll just kind of walk through… I have, like, a quick presentation, but nothing too serious. But then, obviously, the idea here would also be to hear your thoughts, your feedback, what’s going well, what we could attack more of, and how we could, you know.
21 00:03:20.080 ⇒ 00:03:23.829 Demilade Agboola: Fully, like, maximize what we have going for us.
22 00:03:25.310 ⇒ 00:03:26.379 Mary Burke: Great, sounds good.
23 00:03:26.680 ⇒ 00:03:27.500 Michael Thorson: Do it.
24 00:03:34.940 ⇒ 00:03:46.519 Demilade Agboola: So, for us, like, in terms of, like, operations and, like, delivery, I think generally speaking, our Perfect DBT flow is in a healthy status, and has been for a bit.
25 00:03:47.140 ⇒ 00:03:52.380 Demilade Agboola: I mean, obviously, we’ve had the issues with, like, Business Central and, you know, the invoice line tables.
26 00:03:53.120 ⇒ 00:03:59.339 Demilade Agboola: I think with, in terms of resolution, We were, you know, pretty…
27 00:03:59.720 ⇒ 00:04:05.680 Demilade Agboola: quick with resolution. We’re able to push the fix, and then also push a backfill
28 00:04:06.000 ⇒ 00:04:12.989 Demilade Agboola: To ensure that… Whatever contacts was lost was quickly recovered as quickly as possible.
29 00:04:13.560 ⇒ 00:04:21.109 Demilade Agboola: In terms of spins, I think we’ve gotten the data, like, Ingested, we’ve created the pipeline.
30 00:04:21.870 ⇒ 00:04:30.299 Demilade Agboola: And right now, it’s just in function of, like, internally, we’re just trying to have, like, the documentation, like, done, ready, and we’ll send it over to the team.
31 00:04:31.080 ⇒ 00:04:36.620 Demilade Agboola: Dbt, and, like, just testing in general.
32 00:04:36.980 ⇒ 00:04:40.079 Demilade Agboola: So I’m in final stages of, like, the dbt tests.
33 00:04:40.430 ⇒ 00:04:44.910 Demilade Agboola: And the idea here is… I will push the PRR today.
34 00:04:45.330 ⇒ 00:04:53.020 Demilade Agboola: Or, you know, get the PR out today. And the goal is just, basically, we want to provide visibility on troublesome data points.
35 00:04:53.420 ⇒ 00:04:55.550 Demilade Agboola: Like, I was talking to Michael earlier this week.
36 00:04:56.520 ⇒ 00:05:03.980 Demilade Agboola: Even some of the failures that happen within… Prefect, we can detect through dbt on the run.
37 00:05:04.130 ⇒ 00:05:08.259 Demilade Agboola: So something like the freshness went out of it, we can tell that, hey.
38 00:05:08.400 ⇒ 00:05:16.729 Demilade Agboola: The latest data is 12 hours ago, or 24 hours ago, or 36 hours ago, therefore, you know, something must have broken upstream.
39 00:05:17.300 ⇒ 00:05:23.089 Demilade Agboola: And we know that we need to, like, look into Prefect, so we don’t always… need to…
40 00:05:23.920 ⇒ 00:05:30.509 Demilade Agboola: constantly, like, go… because part of the monitoring process right now is I have to go, like, through the different systems and just kind of
41 00:05:30.610 ⇒ 00:05:35.209 Demilade Agboola: Do the manual check and ensure that There aren’t any issues with…
42 00:05:35.980 ⇒ 00:05:39.380 Demilade Agboola: Like, anything, to be honest, like, any of the runs and any of the flows.
43 00:05:40.060 ⇒ 00:05:44.710 Demilade Agboola: And this would kind of automate a part of it, especially, like, the…
44 00:05:45.050 ⇒ 00:05:52.769 Demilade Agboola: preferred part where the data might get stale, we can be able to see through dbt and say, hey, this data looks still.
45 00:05:52.920 ⇒ 00:05:58.210 Demilade Agboola: And so that will allow us to be a bit more proactive as well, and just catch it even earlier than we currently do.
46 00:05:59.080 ⇒ 00:06:02.750 Demilade Agboola: And then further downstream, in terms of, like, the models we’re building in dbt.
47 00:06:02.970 ⇒ 00:06:06.209 Demilade Agboola: part of what Michael and I talked about was being able to
48 00:06:07.210 ⇒ 00:06:13.170 Demilade Agboola: Think of, like, business-specific logic tests that we can ensure that we do, so that if.
49 00:06:13.170 ⇒ 00:06:13.570 Mary Burke: Yeah.
50 00:06:13.570 ⇒ 00:06:28.790 Demilade Agboola: is wrong, we can start to, like, know before, like, it hits the business stakeholders downstream. So it could be things around, like, what the expected revenue should be, so we can say, hey, revenue in a competitor day should not be less than $200,000, or whatever.
51 00:06:29.150 ⇒ 00:06:34.430 Demilade Agboola: And we can just kind of, like, ensure that we enforce these rules, and so if anything goes
52 00:06:34.580 ⇒ 00:06:40.019 Demilade Agboola: wrong, or if anything looks weird, we’re the first to know. Maybe it’s incomplete data.
53 00:06:40.210 ⇒ 00:06:42.499 Demilade Agboola: you know, an API, maybe it’s…
54 00:06:43.210 ⇒ 00:06:49.859 Demilade Agboola: data just stop being fresh. Whatever it is, we’ll just kind of be the first ones to have an idea of what’s going on and be able to jump on that.
55 00:06:50.580 ⇒ 00:07:00.739 Mary Burke: Yeah, I think, thinking through the, like, business-specific tests, are you imagining doing those, or building those out within Omni, or, the BI tool, or doing that within…
56 00:07:01.140 ⇒ 00:07:05.590 Mary Burke: Like, the data warehouse itself. Like, where are you imagining that?
57 00:07:06.380 ⇒ 00:07:08.299 Mary Burke: Those sense checks to take place.
58 00:07:08.590 ⇒ 00:07:13.440 Demilade Agboola: Yeah, so the goal… And that’s how I say it will be within DBT.
59 00:07:13.570 ⇒ 00:07:26.170 Demilade Agboola: And dbt will allow us to automatically come up with these tests, and then once the models run, every single time the models run, dbt will automatically apply those tests against the computer tables. So we have, like.
60 00:07:27.340 ⇒ 00:07:31.520 Demilade Agboola: A run, a completed table, and a test, so we ensure that
61 00:07:31.630 ⇒ 00:07:35.079 Demilade Agboola: Those tests, those tables that have just been created.
62 00:07:35.410 ⇒ 00:07:47.919 Demilade Agboola: meet our standard of, like, data quality. And where it doesn’t meet it, we’re alerted, well, let’s… we know that stuff is off, and we can look into it, and again, because of that, be the first ones to know that, hey, something is off here.
63 00:07:48.490 ⇒ 00:07:50.499 Demilade Agboola: Dbt also has, like, metadata.
64 00:07:51.270 ⇒ 00:07:56.710 Demilade Agboola: And so we can start to use those metadata to create, monitoring dashboards.
65 00:07:57.300 ⇒ 00:08:10.949 Demilade Agboola: So, because of the metadata of the test, we can start to get things around, like, what tests failed, what tests are passing, when last the test failed, like, things like that. And so, we can therefore use that metadata to create, like, a monitoring dashboard.
66 00:08:11.130 ⇒ 00:08:12.390 Demilade Agboola: So, you know.
67 00:08:12.390 ⇒ 00:08:12.730 Michael Thorson: Yeah.
68 00:08:12.730 ⇒ 00:08:21.490 Demilade Agboola: Anybody on the team can kind of look into that and kind of have an idea of, what issues we’re potentially having across our data infrastructure.
69 00:08:22.080 ⇒ 00:08:22.430 Demilade Agboola: Damn.
70 00:08:22.430 ⇒ 00:08:38.009 Mary Burke: Yeah, I think one thing that would be helpful, obviously, like, being able to monitor if things fail and what’s failed and for how long is really useful, but also, I think those, proactive alerts, like having it set up to Slack the team that something’s off, which
71 00:08:38.010 ⇒ 00:08:49.750 Mary Burke: We can also give context of, oh, this isn’t that important, and we can wait to fix it until DPT runs again, or this is actually pretty business critical, and we should jump on this and figure out what’s wrong. I think,
72 00:08:49.840 ⇒ 00:08:56.500 Mary Burke: Just… if giving that visibility without some having to proactively go into a dashboard would also be useful.
73 00:08:57.120 ⇒ 00:09:01.370 Demilade Agboola: Yeah, definitely. I think we shouldn’t, like, I think both of them will work hand in hand.
74 00:09:01.510 ⇒ 00:09:07.210 Demilade Agboola: I think for the data team, being able to proactively get those alerts will be very important.
75 00:09:07.530 ⇒ 00:09:11.899 Demilade Agboola: Because we would need to see it directly in Slack.
76 00:09:12.170 ⇒ 00:09:14.360 Demilade Agboola: And take action based off of that.
77 00:09:14.630 ⇒ 00:09:16.470 Demilade Agboola: Versus,
78 00:09:19.060 ⇒ 00:09:26.040 Demilade Agboola: a dashboard where you, again, have to go in manually to kind of see what’s going on, which, I mean, obviously has its importance.
79 00:09:26.520 ⇒ 00:09:28.599 Demilade Agboola: Of what’s going on over, like, a large…
80 00:09:28.950 ⇒ 00:09:36.770 Demilade Agboola: time frame, so we can say over the last two weeks, we’ve only had 2 failures, and it was because of this, or because of that. We can kind of have an idea of what was going on.
81 00:09:36.910 ⇒ 00:09:51.180 Demilade Agboola: But in terms of, like, yeah, being able to proactively jump on things and just, like, as quickly as possible, reduce the time from failure to up, like, from downtime to uptime as quickly as possible, yeah, proactive alerting will be the way to go.
82 00:09:51.970 ⇒ 00:09:53.060 Mary Burke: Okay, awesome.
83 00:09:53.890 ⇒ 00:09:57.269 Michael Thorson: Yeah, and just a… just a thought about this, like…
84 00:09:57.830 ⇒ 00:10:06.369 Michael Thorson: Well, probably, like, as we move forward and, like, can review the dbt, like, test logs over time and start to, like, track trends.
85 00:10:06.530 ⇒ 00:10:06.880 Demilade Agboola: Yeah.
86 00:10:06.880 ⇒ 00:10:25.279 Michael Thorson: like, I’ll probably want to build out just, like, a… like, almost like a manual input to start to tag failure groupings, so we start to, like, build upon, like, what types of failures and what solutions end up working, for our organization. So we, like, I feel like we don’t have, like, a decent…
87 00:10:25.520 ⇒ 00:10:42.770 Michael Thorson: like, organizational knowledge of, like, actually categorizing these errors, and, like, actually, like, knowing which ones we could potentially, like, self-service with a rerun, for example, versus, like, we actually need to go in and, like, push changes. I… I just noticed there’s, like, a lot of intermittent failures due to, like, timeout that I think…
88 00:10:43.330 ⇒ 00:10:52.799 Michael Thorson: could be solved for, but also, like, the shortest path can often just be, like, rerunning the prefig workflow when all others are, like, paused, so…
89 00:10:52.840 ⇒ 00:11:06.599 Michael Thorson: super excited to, like… I see this, like, living in Omni, potentially, but then, like, having kind of a UI where we can start to log, or, like, whoever’s monitoring, whether it be me or y’all, can start to, like, communicate and manage this thing together.
90 00:11:07.420 ⇒ 00:11:11.059 Demilade Agboola: Yeah, definitely, I agree. I agree, especially, like, Yvonne with Perfect.
91 00:11:11.290 ⇒ 00:11:14.469 Demilade Agboola: I… one of the things…
92 00:11:16.140 ⇒ 00:11:21.119 Demilade Agboola: looking at, and I’ll probably get Ashwini to look at very soon, is how…
93 00:11:21.250 ⇒ 00:11:29.930 Demilade Agboola: We can restructure the perfect environment to such a point that We know what… potentially…
94 00:11:30.090 ⇒ 00:11:42.030 Demilade Agboola: like, we know the infrastructure for different flows, basically. What should have large, extra large, and, like, because some of these failures are not necessarily always a function of the APIs themselves. I mean, some of them are, but some of them are also just, like.
95 00:11:42.980 ⇒ 00:11:51.019 Demilade Agboola: Because the perfect infrastructure isn’t set up in such a way that the right, infrastructure
96 00:11:51.610 ⇒ 00:11:56.870 Demilade Agboola: Like, it’s set to the default infrastructure, and it runs out of memory before,
97 00:11:56.980 ⇒ 00:12:01.239 Demilade Agboola: it’s done loading all the data, so it crashes. That’s what makes the flow crash.
98 00:12:01.560 ⇒ 00:12:12.859 Demilade Agboola: So being, like, potentially being able to put the right infrastructure to the right, flows, so we know that, hey, for this API, it might, like, the default might work 80% of the time.
99 00:12:13.040 ⇒ 00:12:32.529 Demilade Agboola: But just so that… that 1% of the time, or that, you know, the remaining 20% of time where it’s, like, beyond the normal default size, how about, like, make it the large? Or we can also look at potentially maybe autoscaling and how that works within Prefect. But the idea is, like, yes, I would like us to be more proactive with some of these solutions.
100 00:12:33.160 ⇒ 00:12:47.600 Demilade Agboola: And once we have a good idea of how we want to tackle it, we will then propose a solution to the team as to, like, how we want to ensure that we have less perfect failures, and obviously that would translate into better data quality downstream.
101 00:12:48.970 ⇒ 00:12:54.029 Michael Thorson: Nice. Yeah, that’s… I love the way you’re thinking about things, and, like, it being a little bit more…
102 00:12:54.540 ⇒ 00:13:03.049 Michael Thorson: Dynamic with, like, how we solve these problems, because, yeah, increased memory for every job is probably not the right solution.
103 00:13:03.050 ⇒ 00:13:07.539 Demilade Agboola: No, definitely not, definitely not. Because even, like, the failure.
104 00:13:07.710 ⇒ 00:13:16.290 Demilade Agboola: like, when we had the invoice line problem, I started looking through the logs of everything, and for everyone, it was just,
105 00:13:17.080 ⇒ 00:13:20.179 Demilade Agboola: Memory, run out of memory, run out of memory, run out of memory.
106 00:13:20.970 ⇒ 00:13:27.229 Demilade Agboola: And then, if you look through… even when you would manually, like, re-run it, it would run out of memory as well.
107 00:13:27.460 ⇒ 00:13:45.619 Demilade Agboola: And so when you put the infrastructure to large, it would handle it. So now, I think the goal should be, how do we auto-scale? Like, how do we ensure that, on the runs, because again, the default was fine for most runs before, right? So it’s not like a thing of, the default scale doesn’t ever work.
108 00:13:46.000 ⇒ 00:13:52.050 Demilade Agboola: Well, in those instances where we needed to, like, just be a little, like, have, like, a larger…
109 00:13:52.670 ⇒ 00:13:57.379 Demilade Agboola: memory assigned to that job. We just want to be able to be on top of it.
110 00:13:57.830 ⇒ 00:13:58.500 Michael Thorson: as possible.
111 00:14:02.700 ⇒ 00:14:05.750 Michael Thorson: Exciting times for data infrastructure.
112 00:14:06.280 ⇒ 00:14:09.110 Demilade Agboola: This is… this is the fun stuff.
113 00:14:09.110 ⇒ 00:14:12.140 Michael Thorson: I’m so… I’m, like, embarrassingly how stoked I am.
114 00:14:13.060 ⇒ 00:14:15.179 Demilade Agboola: This is the fun stuff.
115 00:14:15.560 ⇒ 00:14:16.220 Michael Thorson: Yeah.
116 00:14:16.810 ⇒ 00:14:24.690 Demilade Agboola: Yeah, so I think, in terms of Spins data, like I mentioned earlier, we’ve been able to load the data in, perfect pipelines have been set up.
117 00:14:24.960 ⇒ 00:14:28.820 Demilade Agboola: The base models have been done, so we’ve been able to, you know, chunk it up.
118 00:14:29.010 ⇒ 00:14:31.430 Demilade Agboola: In different ways, named accordingly.
119 00:14:31.930 ⇒ 00:14:36.270 Demilade Agboola: And so now, at this point, it’s just, how do we want to integrate it?
120 00:14:36.640 ⇒ 00:14:42.320 Demilade Agboola: into, like, our Omni setup, and how do we want to present it to our business stakeholders.
121 00:14:42.480 ⇒ 00:14:47.969 Demilade Agboola: So that we know, like, what downstream enrichment looks like, and how can you provide the most value.
122 00:14:48.230 ⇒ 00:14:51.699 Demilade Agboola: For any stakeholders, but that’s also been done as well.
123 00:14:53.090 ⇒ 00:14:59.189 Michael Thorson: Yeah, follow up here, I met with JT yesterday to do a quick, like, spins roadmap thing, and
124 00:14:59.320 ⇒ 00:15:01.530 Michael Thorson: Where we’re at with this project is…
125 00:15:01.650 ⇒ 00:15:14.809 Michael Thorson: kind of in, like, a pause before the next data release, everything’s looked good so far in terms of data quality, like, matching platform. The next steps are, like, using what we have to just stand up, like, a quick
126 00:15:15.030 ⇒ 00:15:32.859 Michael Thorson: kind of, like, a… I chose, like, the top 5 visualizations from our business stakeholders, and we want to rebuild them in Omni, just to, like, prove value, and I think the end customer, just so you’re aware, at first will likely be, like, leadership. So users that are, like, diving into
127 00:15:32.900 ⇒ 00:15:42.310 Michael Thorson: some core measures, but, like, not the full dataset. It’s gonna be, like, heavily filtered down in Omni, but I want it to be, like, very interactive and clickable, so…
128 00:15:42.410 ⇒ 00:16:01.090 Michael Thorson: I think I’ll take the responsibility, like, standing up those, since I think I have the most, like, business context. So I’ll set up the Omni models plus the visualizations, but I’ll obviously share them with this team so everyone can kind of get on board with, like, hey, this is the direction. We’ll probably have a kind of a series of work once we start
129 00:16:01.260 ⇒ 00:16:15.179 Michael Thorson: like, showing those, like, road testing those with our teams, and I’m sure there’s gonna be a ton of requests for, like, hey, can we get it sliced this way, that way, this way, that way, where we might top, Brainforge for some additional bandwidth there.
130 00:16:15.880 ⇒ 00:16:22.390 Demilade Agboola: Yeah, definitely. I think one of the first things sometimes, especially when, like, new data sources come in, It’s…
131 00:16:22.630 ⇒ 00:16:33.970 Demilade Agboola: some people struggle to, like, visualize and conceptualize how we’re gonna use that data, especially if it hasn’t necessarily been a pressing need of, like, I need this data, like, ASAP.
132 00:16:34.250 ⇒ 00:16:39.549 Demilade Agboola: But yeah, being able to show them, like, hey, this is the utility that we can get from these numbers.
133 00:16:39.680 ⇒ 00:16:54.579 Demilade Agboola: And this is how we can better enable you, like, enable you to make better decisions. I think that will be the first step. Once they start to see it and have an idea of, you know, hey, can we dice it this way, or slice it this way, or how can we utilize it?
134 00:16:54.670 ⇒ 00:17:00.689 Demilade Agboola: more in this dashboard or that dashboard, I think that will be the next step in helping us know how we want to…
135 00:17:00.750 ⇒ 00:17:15.880 Demilade Agboola: model the data within dbt and say, hey, I think we need to join it to more Shopify data, or we need to join it to more, like, this is how we want to utilize it, and I think that would come in more once we have a clearer idea of what’s going to be used downstream.
136 00:17:17.089 ⇒ 00:17:31.529 Michael Thorson: Totally. And I think, like, the… in terms of, like, the analytics engineering that we might lend… move towards is, like, they’re coming to a cool point where Spins data is in warehouse, Magic Spin’s sales data, like, which customer we’re selling to is in warehouse.
137 00:17:31.530 ⇒ 00:17:38.250 Michael Thorson: And we should have access to a little bit of, like, point-of-sales data directly from a few key retailers.
138 00:17:38.330 ⇒ 00:17:41.259 Michael Thorson: So, like, the way I see this maybe evolving is, like.
139 00:17:41.600 ⇒ 00:17:54.380 Michael Thorson: I really like starting to stack those three, because they’re reporting the same thing, it’s revenue, but it’s, like, there’s some time delays based on what report we’re looking at. So I see some potential value in tracking, kind of, like, Magic Spoon sales.
140 00:17:54.380 ⇒ 00:18:02.400 Michael Thorson: to our customer, point of sale, like, register sales, and then also spins, which is more of, like, a survey sale,
141 00:18:02.400 ⇒ 00:18:12.539 Michael Thorson: Dataset, and just starting to, like, use that as kind of, like, our one-stop shop for our leadership team to understand, like, what’s going on for our product at every, like, kind of stage of the business.
142 00:18:12.580 ⇒ 00:18:13.630 Demilade Agboola: I think…
143 00:18:14.280 ⇒ 00:18:31.999 Michael Thorson: Cool. Just, like, and Mary, to context, too, is, like, JT and I were talking, like, we’re kind of extending pacing to, like, also include, like, point-of-sales data, so it’s kind of like, how are we pacing to our customers, and how are we pacing in market? And if we can, like, start to layer those on, I think that’s kind of, like, a really valuable thing that…
144 00:18:32.000 ⇒ 00:18:34.689 Michael Thorson: Isn’t, like, unified yet, so…
145 00:18:34.690 ⇒ 00:18:36.380 Mary Burke: Yeah, great.
146 00:18:36.750 ⇒ 00:18:37.190 Michael Thorson: Cool.
147 00:18:40.770 ⇒ 00:18:47.080 Demilade Agboola: Okay, so, in terms of hours this week,
148 00:18:48.060 ⇒ 00:18:51.859 Demilade Agboola: So far, monitoring has been, like, 2 hours. I’ve spent, like.
149 00:18:52.260 ⇒ 00:18:57.190 Demilade Agboola: 7 hours this week, so I spent 4… on Wednesday.
150 00:18:57.460 ⇒ 00:19:00.720 Demilade Agboola: And I spend, like, one across the different days of the week.
151 00:19:01.050 ⇒ 00:19:05.839 Demilade Agboola: And so now… Ashwini spent one hour with the Spins documentation.
152 00:19:07.440 ⇒ 00:19:12.400 Demilade Agboola: So we’re still, like, within, like, the estimated range of ours.
153 00:19:14.980 ⇒ 00:19:20.109 Demilade Agboola: And I feel like, ultimately, we’ve been able to make some progress in terms of what we want to see out of…
154 00:19:20.330 ⇒ 00:19:21.909 Demilade Agboola: Our data so far.
155 00:19:22.390 ⇒ 00:19:24.309 Demilade Agboola: I mean…
156 00:19:25.690 ⇒ 00:19:32.690 Demilade Agboola: which kind of leads us to, like, the final part, like, so, like, I’m done presenting about, like, what we’ve been doing, so my thing, my question will be, like.
157 00:19:33.280 ⇒ 00:19:41.049 Demilade Agboola: Do we have any questions, or feedback, or things that we would like us to do more of, or, like, scope of attack? What would we like to…
158 00:19:41.460 ⇒ 00:19:45.050 Demilade Agboola: Prioritize about the next coming weeks?
159 00:19:47.090 ⇒ 00:19:56.830 Mary Burke: Yeah, I think, the number one priority is just getting the… when we have the new spins data release, which should be, like, maybe next Monday, Michael? I don’t have the calendar up, but…
160 00:19:56.830 ⇒ 00:19:58.210 Michael Thorson: I think it depends, yeah.
161 00:19:58.210 ⇒ 00:20:03.070 Mary Burke: Yeah, sometime soon. I think using, like Michael said, just getting a…
162 00:20:03.070 ⇒ 00:20:19.120 Mary Burke: being able to have that, like, filtered dashboard in Omni ready to at least, like, beta test with stakeholders, would be great. And then, yeah, I think continued work on the monitoring and alerting, just to make sure that the data warehouse is healthy and happy.
163 00:20:20.380 ⇒ 00:20:22.279 Demilade Agboola: We love a healthy, happy warehouse.
164 00:20:22.280 ⇒ 00:20:23.260 Mary Burke: Yeah.
165 00:20:26.340 ⇒ 00:20:27.720 Mary Burke: Does that sound right, Michael?
166 00:20:28.350 ⇒ 00:20:32.769 Michael Thorson: Yeah, I was gonna say, just let’s keep focusing on the, yeah, dbt test, and then…
167 00:20:32.980 ⇒ 00:20:47.489 Michael Thorson: all, like, I… open action to stand up just, like, what we want to observe with the next Spins release, so that, like, when we hit the ground running, I think we’re gonna have to collaborate, just, like, make sure it’s healthy. The Spins pipeline and the marts are healthy, so…
168 00:20:47.740 ⇒ 00:20:58.599 Demilade Agboola: Okay, if we have an idea of what the dashboard is and what we want it to look like, in terms of, like, what we just wanna… again, it could literally just be a very, like, meaning, like.
169 00:20:59.540 ⇒ 00:21:03.240 Demilade Agboola: very, like, an MVP concept dashboard, doesn’t have to be anything crazy.
170 00:21:03.990 ⇒ 00:21:11.420 Demilade Agboola: Pointing to some spins tables, taking the sum of, like, things, like, it’s, again, nothing crazy. Could be metric cards, could be quick…
171 00:21:11.570 ⇒ 00:21:12.790 Demilade Agboola: Bodcast, whatever.
172 00:21:12.980 ⇒ 00:21:18.099 Demilade Agboola: But the idea is if we have an idea of, like, those charts, and you can provide visibility to us.
173 00:21:18.570 ⇒ 00:21:21.779 Demilade Agboola: We can use some sort of mic monitoring to see that.
174 00:21:21.950 ⇒ 00:21:25.230 Demilade Agboola: To see what we want to see, and just to ensure that
175 00:21:25.450 ⇒ 00:21:36.000 Demilade Agboola: data is loading, like, the latest data is available, all those kind of stuff, and we can easily tell if things are breaking within the Spain’s pipeline.
176 00:21:37.480 ⇒ 00:21:40.769 Mary Burke: Just making sure that the mods are set up to support what those views are.
177 00:21:41.710 ⇒ 00:21:52.499 Michael Thorson: Yeah, the easiest one, honestly, is, like, and this is, like, the top line, is just… I can, like, also start a thread on this, fill out, like, a roadmap for BI for spins, but…
178 00:21:52.500 ⇒ 00:22:06.120 Michael Thorson: it’s really just, like, dollars over time is… and then dollars over time by products, because the marks you build should be, like… there’s kind of a regional view, and then there’s also, like, a top-line view, so it’s, like, starting to actually pull in…
179 00:22:06.320 ⇒ 00:22:20.659 Michael Thorson: like, just, like, dollars over time for each of the marts, and then, like, having slices by product, having slices by brand, having slices by region. It’s, like, those core dimensions, as, like, the filters, top line, if that makes sense.
180 00:22:20.960 ⇒ 00:22:21.730 Demilade Agboola: Monica?
181 00:22:21.730 ⇒ 00:22:32.640 Michael Thorson: Cool. I can start, like, a roadmap, though, and, like, start to get those thoughts done on paper, and then we can kind of, like, tag team and build a dashboard. I’m super curious to see what y’all’s take on it would be, so…
182 00:22:33.180 ⇒ 00:22:39.019 Demilade Agboola: Okay, sounds good. So you can just always send that to me, or you send that today, tomorrow, so, like, I can…
183 00:22:39.250 ⇒ 00:22:42.779 Demilade Agboola: Paste that into, like, our planned scope of attack for next week.
184 00:22:42.960 ⇒ 00:22:50.899 Demilade Agboola: But if… even if you can’t get across to me today, like, again, by Monday as well, just let me know, so I can also, like, Monday morning as well.
185 00:22:51.160 ⇒ 00:22:55.500 Demilade Agboola: I can also plan that into, like, how he plans to look at the week from a Brainford side.
186 00:22:57.190 ⇒ 00:23:12.350 Michael Thorson: That sounds great. And I’m… I’m just gonna create a new tab in our spins, kind of master tracker, since that is such a monolithic project. So I’ll have, like, we have a… we have an ETL roadmap, we have a MART roadmap, and then we’ll have a BI roadmap as well. It’s, like, obviously stacked on top.
187 00:23:12.640 ⇒ 00:23:15.420 Demilade Agboola: Okay, sounds good. Look forward to seeing that.
188 00:23:15.970 ⇒ 00:23:19.760 Demilade Agboola: Cool. You can just tag me, you can tag me, and let me know when it’s done.
189 00:23:20.550 ⇒ 00:23:22.380 Michael Thorson: Great. I’ll… I’ll hand it up.
190 00:23:22.660 ⇒ 00:23:23.630 Demilade Agboola: Thank you.
191 00:23:25.280 ⇒ 00:23:26.510 Mary Burke: Awesome, thanks, Debbie.
192 00:23:26.510 ⇒ 00:23:26.840 Demilade Agboola: Yay!
193 00:23:26.840 ⇒ 00:23:28.149 Mary Burke: Have a good weekend, guys!
194 00:23:28.650 ⇒ 00:23:29.660 Demilade Agboola: Have a good weekend.
195 00:23:30.910 ⇒ 00:23:31.750 Demilade Agboola: Bye.
196 00:23:31.750 ⇒ 00:23:32.310 Mary Burke: See ya!