Meeting Title: Urban Stems Project Sync Date: 2025-06-23 Meeting participants: Demilade Agboola, Caio Velasco
WEBVTT
1 00:03:29.210 ⇒ 00:03:29.950 Caio Velasco: Hello!
2 00:03:35.810 ⇒ 00:03:39.710 Demilade Agboola: Hi, Kyle, how are you doing.
3 00:03:40.120 ⇒ 00:03:41.439 Caio Velasco: Good! How are you?
4 00:03:43.270 ⇒ 00:03:49.089 Demilade Agboola: I’m okay. I’m hoping this week’s a better week than last week last week was was chaotic, was fucking chaotic.
5 00:03:49.410 ⇒ 00:03:54.914 Caio Velasco: I can’t imagine that I also can’t stand it, but try to do my best to organize it.
6 00:03:56.910 ⇒ 00:03:58.690 Demilade Agboola: Yeah, like.
7 00:03:59.370 ⇒ 00:04:13.169 Demilade Agboola: I know, like also, last week I wasn’t the most present on the urban settings project. It didn’t took all my attention. So, but this week this week should be fine. So I should be back like more present on the projects.
8 00:04:13.540 ⇒ 00:04:14.570 Caio Velasco: Mess.
9 00:04:14.570 ⇒ 00:04:15.565 Demilade Agboola: Yeah.
10 00:04:16.680 ⇒ 00:04:21.910 Demilade Agboola: But yeah, I I saw your like, your thoughts on the linear ticket about the cost estimate.
11 00:04:22.559 ⇒ 00:04:35.050 Demilade Agboola: I I think what we can do now is we can do some of we can do an estimate of what we’ve done so far and say, Oh, we’ve saved you. I don’t know how much we save them. But let’s say we save them $500 a month.
12 00:04:36.140 ⇒ 00:04:45.270 Demilade Agboola: We can say, Hey, we’ve done this so far. This is potentially how much we can save because we’re still looking at further deprecating or like turning off tables.
13 00:04:45.370 ⇒ 00:04:49.339 Demilade Agboola: So just having an estimate is is always just always good, because at least
14 00:04:49.530 ⇒ 00:04:55.159 Demilade Agboola: it at least allows us to number one. See how much we’ve done so far for ourselves.
15 00:04:55.300 ⇒ 00:05:24.249 Demilade Agboola: and number 2 also allows us to know how much to communicate to the urban stems team that, like, hey, like, we are producing some value, even though this is still the slow part of what we’re still trying to do. But there is some business value. And then amber can always look at sales, can look at it and decide on like, how do we push those numbers to them, or do we just wait until we’re done with everything before we push those numbers to them that allows us to at least keep track of certain things. I think that’s the important thing.
16 00:05:25.400 ⇒ 00:05:26.870 Caio Velasco: For that part.
17 00:05:28.160 ⇒ 00:05:34.940 Caio Velasco: I never really work like with the costs. Part. How would you see that? For example, we turned off
18 00:05:35.080 ⇒ 00:05:39.029 Caio Velasco: a pipeline XYZ. In stitch.
19 00:05:39.650 ⇒ 00:05:46.399 Caio Velasco: How would I know how much that pipeline was costing before? And now that it’s off, we would have the difference.
20 00:05:47.460 ⇒ 00:05:54.389 Demilade Agboola: I think it would depend on like the usage does it? Does the pipeline show like how much it’s been ingesting.
21 00:05:55.970 ⇒ 00:05:58.420 Caio Velasco: No, I cannot even open here. Let me see.
22 00:05:58.730 ⇒ 00:06:08.639 Demilade Agboola: Because if it shows how much it’s been ingesting, it’s like over the last couple of weeks or months. Or let’s just say month period. Maybe it’s ingested Xy zeros.
23 00:06:08.860 ⇒ 00:06:14.810 Demilade Agboola: And then, if you know how stitch does their calculations like. Usually it’s a
24 00:06:15.840 ⇒ 00:06:20.860 Demilade Agboola: a function of like maybe number of rows or number of columns, or whatever like their metric is.
25 00:06:22.690 ⇒ 00:06:25.679 Demilade Agboola: and we can then compare or do the math on it.
26 00:06:36.710 ⇒ 00:06:42.029 Demilade Agboola: Neither any heads, and I need block out my current. I need some heads down time with around stamps.
27 00:06:42.320 ⇒ 00:06:44.930 Demilade Agboola: I need like 3 h when nobody disturbs.
28 00:06:45.812 ⇒ 00:06:53.109 Caio Velasco: Sometimes I need like a week to nobody, so to get things done.
29 00:06:53.300 ⇒ 00:06:54.130 Demilade Agboola: Yeah.
30 00:06:54.230 ⇒ 00:06:55.439 Demilade Agboola: Now I’ll be.
31 00:06:55.820 ⇒ 00:07:01.209 Caio Velasco: Let me see. So, for example, these are the the pipelines. I know that this one we turned off
32 00:07:02.570 ⇒ 00:07:04.630 Caio Velasco: probably 12 days ago, makes sense.
33 00:07:05.290 ⇒ 00:07:07.190 Caio Velasco: If I go in that
34 00:07:09.160 ⇒ 00:07:14.970 Caio Velasco: Low details rules this billing period. Maybe something here.
35 00:07:15.300 ⇒ 00:07:19.489 Demilade Agboola: Yeah, but there are rows, though. Can you go on the like rows on the right.
36 00:07:20.480 ⇒ 00:07:22.639 Caio Velasco: Roles on the right.
37 00:07:22.640 ⇒ 00:07:24.150 Demilade Agboola: Like the yellow. Yeah.
38 00:07:24.590 ⇒ 00:07:25.370 Caio Velasco: Yeah.
39 00:07:25.630 ⇒ 00:07:29.859 Demilade Agboola: So there are rows loaded, I don’t know why, saying like 0 rows this billing period.
40 00:07:32.290 ⇒ 00:07:34.339 Demilade Agboola: so I’m guessing. Turn it off on the 12, th right.
41 00:07:35.220 ⇒ 00:07:43.809 Caio Velasco: Yeah, this will this one, I think we probably paused it 12 days ago makes sense. I don’t remember, but I think this name reminds me of something. We turned off
42 00:07:45.510 ⇒ 00:07:49.660 Caio Velasco: and it well, I can’t even well, it’s easier to just come here.
43 00:07:52.230 ⇒ 00:07:56.840 Caio Velasco: So thing.
44 00:08:00.160 ⇒ 00:08:05.559 Caio Velasco: Yeah, it was active. And now it’s inactive. So we turned off this one or shoot.
45 00:08:06.620 ⇒ 00:08:07.105 Demilade Agboola: Okay.
46 00:08:10.940 ⇒ 00:08:16.070 Caio Velasco: So this was the number of rows being jested always, basically every day.
47 00:08:16.430 ⇒ 00:08:19.450 Caio Velasco: And now we have 2 well days that
48 00:08:19.740 ⇒ 00:08:27.610 Caio Velasco: these those roles weren’t being jested. I just don’t know how much that would mean, but can try to see what I find
49 00:08:28.070 ⇒ 00:08:32.059 Caio Velasco: that makes sense right like the number of roles, and now they they’re not being jested.
50 00:08:32.929 ⇒ 00:08:42.839 Demilade Agboola: Yes, but what I want to be sure of is, if this is representation of like the rose was ingesting each day. Give me one second how and look at stitch, pricing.
51 00:08:43.340 ⇒ 00:08:44.000 Caio Velasco: Update.
52 00:08:44.680 ⇒ 00:08:47.250 Demilade Agboola: Stitch, price.
53 00:08:47.250 ⇒ 00:08:49.399 Caio Velasco: Oh, yeah, because it can be incremental. Right?
54 00:08:49.580 ⇒ 00:08:55.230 Demilade Agboola: Yeah, yeah, I like, I want to be sure that they charge based off the increase in number of rows.
55 00:08:57.740 ⇒ 00:09:00.700 Demilade Agboola: Oh, stitch data.
56 00:09:26.920 ⇒ 00:09:28.369 Caio Velasco: I’ll just be here.
57 00:09:28.740 ⇒ 00:09:33.550 Demilade Agboola: Interesting stitch seems up. Stitch seems quite cheap, like just.
58 00:09:35.690 ⇒ 00:09:41.210 Caio Velasco: Let me see a 1 that is still one. Then we at least have something, and this one
59 00:09:43.470 ⇒ 00:09:44.319 Caio Velasco: about it.
60 00:09:45.730 ⇒ 00:09:46.480 Caio Velasco: Oh, wow!
61 00:09:47.400 ⇒ 00:09:49.380 Caio Velasco: Each day was loading more and more and more.
62 00:09:50.210 ⇒ 00:09:56.980 Demilade Agboola: Yeah, so like, that’s what I was saying. Like, I I don’t know if it’s a function of like how they charge it, whether it’s based on the number of rows
63 00:09:57.140 ⇒ 00:10:05.090 Demilade Agboola: or like number of new rules. So like, I know, like Fivetron used to have burning charge based on active rules, so that when the world changes, that’s when they charge you.
64 00:10:05.310 ⇒ 00:10:10.979 Demilade Agboola: But I don’t know if they do the same, or if it’s just based on the rows that they have loaded.
65 00:10:11.810 ⇒ 00:10:15.709 Caio Velasco: Okay, I think it might just be based on the load.
66 00:10:16.400 ⇒ 00:10:21.789 Caio Velasco: Oh, well, at least now I can try to like chat, dpt stuff and see like.
67 00:10:21.920 ⇒ 00:10:28.059 Caio Velasco: what is the price stage pricing evil pricing. Now, it makes sense. Yeah, I didn’t think about it would be by role.
68 00:10:28.784 ⇒ 00:10:32.770 Caio Velasco: By the way, we don’t need to lose time with this. Now it’s okay.
69 00:10:35.200 ⇒ 00:10:40.709 Caio Velasco: Well, if you find something, let me know. But at least now I have like something on my mind that’s good.
70 00:10:43.150 ⇒ 00:10:55.700 Caio Velasco: or the next part would be the Whoa is this one here which leads to our
71 00:10:55.810 ⇒ 00:10:58.169 Caio Velasco: the others also down here?
72 00:11:00.660 ⇒ 00:11:20.139 Caio Velasco: yeah. So this is what we were talking about last time, like the legacy revenue models, and I understood that there was stuff being done. Pre migration post migration. We want to identify both so that we can calculate in both scenarios and well see differences, or find new things from the past, and use these new things in in the new ones.
73 00:11:21.690 ⇒ 00:11:28.580 Caio Velasco: it makes sense for me, basically what we are saying, that we want to calculate an equation before in an equation after.
74 00:11:29.070 ⇒ 00:11:31.070 Caio Velasco: For, let’s say, revit
75 00:11:32.580 ⇒ 00:11:35.889 Caio Velasco: The question that I have is, first, st what is.
76 00:11:36.120 ⇒ 00:11:42.140 Caio Velasco: what was, what is migration use like being used using Dbt, or is it something different?
77 00:11:45.110 ⇒ 00:11:59.860 Demilade Agboola: okay, before I answer that question, I’ll just send you a link. It’s just to chase documentation on billing frequently asked questions. So yeah, they charge you, based on volume of rows. So it doesn’t really matter if it’s a new row or old row you could check, but like, just to be sure.
78 00:11:59.960 ⇒ 00:12:00.660 Demilade Agboola: but.
79 00:12:00.660 ⇒ 00:12:01.480 Caio Velasco: Nice.
80 00:12:01.520 ⇒ 00:12:02.900 Demilade Agboola: Yeah.
81 00:12:04.590 ⇒ 00:12:07.040 Demilade Agboola: So you might just that might just be helpful.
82 00:12:07.520 ⇒ 00:12:09.620 Caio Velasco: Okay, cool. I’m gonna put this.
83 00:12:10.260 ⇒ 00:12:10.660 Caio Velasco: Yes.
84 00:12:10.660 ⇒ 00:12:20.189 Demilade Agboola: That just helps you so like. If that just allows you to know if you should add the number of rows, or if it’s just if the volume of like the new rows, the newly added rows
85 00:12:20.560 ⇒ 00:12:25.569 Demilade Agboola: so that will just help you cut estimate based off each table how much that ends up costing.
86 00:12:26.530 ⇒ 00:12:27.200 Caio Velasco: Correct.
87 00:12:28.870 ⇒ 00:12:31.889 Demilade Agboola: And then you could do you could do a zoom for Hevo, and just like
88 00:12:31.990 ⇒ 00:12:38.169 Demilade Agboola: do a summitation. It will be an estimate. Obviously, we don’t necessarily know their contracts. We could ask for their contracts.
89 00:12:38.290 ⇒ 00:12:45.380 Demilade Agboola: but it’s just an estimate based off the like, the frequent pricing and just saying like, hey, this is based off what we know
90 00:12:46.550 ⇒ 00:12:48.060 Demilade Agboola: as your pricing.
91 00:12:48.670 ⇒ 00:12:50.980 Demilade Agboola: This is how much it should be.
92 00:12:51.610 ⇒ 00:13:05.869 Demilade Agboola: So yeah, like, I said, I don’t think it’s going to be a lot. As I said, you might literally just be like $200 or something. It’s not. It’s not like we’re trying to save them the world. But we’re making their days. We’re cutting off efficiencies. And that’s part of what we’re doing.
93 00:13:06.470 ⇒ 00:13:06.960 Demilade Agboola: cool.
94 00:13:06.960 ⇒ 00:13:07.710 Caio Velasco: Helpful.
95 00:13:09.660 ⇒ 00:13:18.354 Demilade Agboola: So yeah, for the for the, for this, the inactive models and streams.
96 00:13:19.420 ⇒ 00:13:22.349 Demilade Agboola: yeah. So the migration is usually a function of
97 00:13:26.380 ⇒ 00:13:28.200 Demilade Agboola: tools in production.
98 00:13:30.198 ⇒ 00:13:37.841 Demilade Agboola: So it’s a function of like what they were using in production. And then, or maybe what they were using for ingestion. And then
99 00:13:38.310 ⇒ 00:13:45.030 Demilade Agboola: they maybe change that. And so those changes have allowed them to make modifications in the models
100 00:13:47.070 ⇒ 00:13:49.900 Demilade Agboola: like, for instance, I know recently just moved to shopify.
101 00:13:50.220 ⇒ 00:13:55.629 Demilade Agboola: I’m not sure what they’re using, for, to be honest, I should I could ask about that. But they just moved to shopify.
102 00:13:56.370 ⇒ 00:14:05.889 Demilade Agboola: And so, for instance, like the data doesn’t come in same way. Shopify models. Data differently like the aspect of shopify is slightly different from what they used before, but, like
103 00:14:06.690 ⇒ 00:14:17.529 Demilade Agboola: everything still calculates revenue. So what they will end up doing is they just kind of like, create this big model where they just kind of put different things and do case wins and like. So it ends up looking really
104 00:14:18.230 ⇒ 00:14:18.990 Demilade Agboola: weird.
105 00:14:19.750 ⇒ 00:14:20.300 Caio Velasco: Beautiful.
106 00:14:20.430 ⇒ 00:14:30.990 Demilade Agboola: It’s very hard to understand, very hard to debug in some cases, when like, there’s an error like going to the exact points to figure out what the error is and like, why things are not working.
107 00:14:31.300 ⇒ 00:14:38.810 Demilade Agboola: It’s just very frustrating. And obviously I don’t like Emily.
108 00:14:39.730 ⇒ 00:14:53.080 Demilade Agboola: Emily has had to suffer, quote, unquote, to be able to get to that point where she can at least be useful to the data. But the problem is, it’s it’s just bad like if you don’t. If you don’t. If you don’t understand how modeling works.
109 00:14:53.230 ⇒ 00:15:00.420 Demilade Agboola: you’re constantly like struggling. And that’s kind of part of the problems so ideally. What we want to be able to get to is get to a point where
110 00:15:01.160 ⇒ 00:15:05.100 Demilade Agboola: everything is done in such a way that, like
111 00:15:05.460 ⇒ 00:15:25.159 Demilade Agboola: the things that don’t change anymore exist like the models that we know. Oh, this is an old system that does not currently exist. They’re only 1 point. So that, like that part is on its own, and then the currently active part can be on its own as well. And so, if things are breaking, things are changing, you know they are only tracing
112 00:15:25.180 ⇒ 00:15:45.140 Demilade Agboola: the active part. You know. They’re not going to look through old code, and you know that part is still fine. So you’re going through the new code. You’re like, okay, so this is what shopify is doing. This is what’s happening. Oh, there’s a break. It like this is what’s breaking? Maybe the skew is not coming as you should like whatever, and that allows you to like pinpoint and be faster troubleshooting
113 00:15:45.636 ⇒ 00:15:48.009 Demilade Agboola: things like that like. So the idea is
114 00:15:48.300 ⇒ 00:15:51.560 Demilade Agboola: also allows for better numbers, because once we
115 00:15:52.590 ⇒ 00:16:09.020 Demilade Agboola: once we know how we’re separating things and putting things together, it’s much easier to figure out if we’re making a miscalculation like truly like, when I 1st joined the project, and I hadn’t gotten full scope of some of these things. I remember they were having a revenue inflation problem.
116 00:16:09.740 ⇒ 00:16:13.360 Demilade Agboola: Right like the revenue was we had, and it should be.
117 00:16:14.150 ⇒ 00:16:29.799 Demilade Agboola: And Emily is going through the. And it’s like, I’m not sure if it’s this line, or if it’s that part I I think it’s this case. It’s hard to like troubleshoot these things when there isn’t a proper where there isn’t proper infrastructure. So the idea is, we know how
118 00:16:29.870 ⇒ 00:16:46.850 Demilade Agboola: data should be modeled. We know, understand that much better than they do, and we have the time to be able to show them like this is what it should be done. So it’s important that we like go in there and properly understand the models and break them apart. And I think
119 00:16:47.030 ⇒ 00:17:01.720 Demilade Agboola: sometimes it’s just about like getting the dirty working like you will be confused. But the idea is that confusion should lead to like better questions. I feel like confusion is a good thing, because that allows you to ask questions that lead to clarity. So I feel like
120 00:17:03.500 ⇒ 00:17:05.329 Demilade Agboola: It might just be helpful to like
121 00:17:05.700 ⇒ 00:17:19.090 Demilade Agboola: dive in there. Figure out like what’s going on here. What you don’t figure out, write it down as questions, and so, if you have your working session with Emily, you have very good questions about, like, Oh, I’ve been going through this model.
122 00:17:19.290 ⇒ 00:17:30.259 Demilade Agboola: I don’t like what’s going on here like, I can tell you. Some of the dates, for instance, like some of the dates, are just migration dates like case when 2024, 11, 0, 6, or something is like
123 00:17:30.450 ⇒ 00:17:33.610 Demilade Agboola: they made their migration in November of last year.
124 00:17:33.900 ⇒ 00:17:40.389 Demilade Agboola: So the case one statements will reflect that things like that like you can like. Once you start to go through
125 00:17:40.970 ⇒ 00:17:43.630 Demilade Agboola: the revenue and kind of go up.
126 00:17:44.431 ⇒ 00:17:54.190 Demilade Agboola: Some things will make sense. Some things won’t make sense things that don’t make sense like. Write them down in notes, pad, or like in a sheet. And then you can ask any of those questions.
127 00:17:55.530 ⇒ 00:18:08.189 Caio Velasco: Okay. Okay, okay, cool. And from your knowledge, you said that you were working more with the inventory than the revenue. So you have probably not much knowledge about how their revenue is defined. Right.
128 00:18:08.350 ⇒ 00:18:16.100 Demilade Agboola: Yeah. The the things I have knowledge on about revenue are things that kind of relate to like subscriptions.
129 00:18:16.870 ⇒ 00:18:24.569 Demilade Agboola: I’m sorry. Well, not but they’re related to eventually at the end of the day. So things like subscriptions, and how some things are calculated with subscriptions.
130 00:18:25.155 ⇒ 00:18:31.559 Demilade Agboola: But it’s not. It’s not a lot. It’s it’s journeying like I don’t know some of the discounts.
131 00:18:31.890 ⇒ 00:18:38.349 Demilade Agboola: I don’t understand how they use them. I know at 1st the subscriptions. I think
132 00:18:39.050 ⇒ 00:18:43.340 Demilade Agboola: it’s you. Take the people pay for subscriptions, and then
133 00:18:43.470 ⇒ 00:19:03.710 Demilade Agboola: it’s deducted. It’s counted as revenue every time they make a purchase, but you might have to clarify that as well you can look through. If if it makes sense sure, doesn’t make sense, you can ask Emily, and be like, Hey, when it comes to revenue and subscriptions. Is it counted as revenue immediately, or is it counted as revenue every single time they make the order? Things like that like
134 00:19:06.060 ⇒ 00:19:11.079 Demilade Agboola: How do you do taxes and all that stuff? Because, again, these are
135 00:19:11.420 ⇒ 00:19:14.139 Demilade Agboola: the things in the tableau items. Xf.
136 00:19:14.290 ⇒ 00:19:19.309 Demilade Agboola: so like, potentially what you might end up just having to do is you might have to
137 00:19:19.410 ⇒ 00:19:26.840 Demilade Agboola: look at tableau items Xf, and then kind of go up right like
138 00:19:27.603 ⇒ 00:19:31.369 Demilade Agboola: so when I mean up, I mean, like upstream, so like the
139 00:19:31.920 ⇒ 00:19:46.019 Demilade Agboola: figure out like, where where does the revenue come from? When you see other totals? When you see discount totals like what, what models are they coming from? And like just going through those models and figuring out how exactly
140 00:19:46.886 ⇒ 00:19:48.519 Demilade Agboola: that data is coming in.
141 00:19:50.730 ⇒ 00:20:01.359 Caio Velasco: Okay? So okay, so that that bring us here. So you’re saying that double items except might be the downstream model.
142 00:20:02.000 ⇒ 00:20:07.930 Caio Velasco: Better link. Think that that we have, like better things related to revenue or or inventory.
143 00:20:08.720 ⇒ 00:20:14.010 Demilade Agboola: To be honest initially. Yeah, like they initially had it for everything.
144 00:20:14.614 ⇒ 00:20:23.749 Demilade Agboola: And that’s part of why we’re trying to like migrate stuff and make things better, because I mean ideally, it’s not great for you to have everything in a spot.
145 00:20:25.840 ⇒ 00:20:27.539 Demilade Agboola: Okay, so that’s okay.
146 00:20:28.450 ⇒ 00:20:32.110 Caio Velasco: Okay. Okay. I get it. No, no.
147 00:20:32.110 ⇒ 00:20:33.910 Demilade Agboola: I just want to show you something.
148 00:20:34.580 ⇒ 00:20:34.949 Caio Velasco: Yes.
149 00:20:39.250 ⇒ 00:20:40.110 Demilade Agboola: Hmm!
150 00:20:41.940 ⇒ 00:20:45.360 Demilade Agboola: Everyone wants something to grab the screen.
151 00:20:48.910 ⇒ 00:20:53.519 Demilade Agboola: We live in business.
152 00:20:54.360 ⇒ 00:21:09.420 Caio Velasco: And even when you say inventory, what is the definition of the inventory? You mentioned subscription? Okay. But is it inventory? Something that is at the end of the day related to the order process. So basically tracking an order from 0 to
153 00:21:09.850 ⇒ 00:21:10.950 Caio Velasco: delivered.
154 00:21:12.020 ⇒ 00:21:17.269 Demilade Agboola: Yes, but they have a very complicated inventory system.
155 00:21:17.610 ⇒ 00:21:25.490 Demilade Agboola: So they have. So they have the general inventory systems. Where, hey?
156 00:21:25.690 ⇒ 00:21:44.740 Demilade Agboola: This place receives an order. Right? So that’s that’s like a 1 to one. Think of that as a 1-to-one. But in certain cities, especially like big cities, they have something called a hub and spoke model. So there’s a hub where it gets the inventory, and there’s a spoke way where eventually the the inventories go to as final destinations.
157 00:21:46.182 ⇒ 00:21:52.190 Demilade Agboola: So that complicates that because you get things you don’t want to count
158 00:21:52.330 ⇒ 00:22:02.132 Demilade Agboola: until it hits this book. When it’s at the hub. It doesn’t count as this, like things like he ends up have adding a new layer of like complexity to certain things.
159 00:22:02.610 ⇒ 00:22:12.369 Demilade Agboola: but yeah, eventually is just basically about like trying to fulfill the orders. And they also have ways. They go about it because it’s flowers, and it’s perishable.
160 00:22:12.945 ⇒ 00:22:27.954 Demilade Agboola: They have things called buffers. So, for instance, if they’re going to, if they they have orders of 25 in. In a certain place, they might send 30, in case 5 of them perish things like that.
161 00:22:28.580 ⇒ 00:22:44.290 Demilade Agboola: so it’s like, so they have buffers. They have Qas there like there’s there’s a lot going on behind the scenes. So the inventory calculation is, it’s it’s a bit more complicated than just we sent out 2 things. This is where like it ends up, having a lot of like math
162 00:22:44.400 ⇒ 00:22:52.061 Demilade Agboola: in it. And then they have their definitions of things quantity received available for sale.
163 00:22:53.260 ⇒ 00:23:05.139 Demilade Agboola: so available for sale will be everything that is in Qa plus everything that is in like received minus this minus that minus like this. There’s just a lot going on there to be honest.
164 00:23:06.080 ⇒ 00:23:09.520 Caio Velasco: Okay, this is, okay, okay, okay.
165 00:23:10.580 ⇒ 00:23:16.639 Demilade Agboola: So this is tableau it and xf, so he has like models upstream.
166 00:23:17.944 ⇒ 00:23:24.329 Demilade Agboola: So the idea is, it will just be about figuring out like which ones particularly matter to this
167 00:23:24.540 ⇒ 00:23:25.770 Demilade Agboola: for revenue.
168 00:23:26.140 ⇒ 00:23:32.410 Caio Velasco: But a quick question, for before, for you to get where you are. What what do you do? You go to studio.
169 00:23:32.980 ⇒ 00:23:34.540 Demilade Agboola: Yes, studio.
170 00:23:35.380 ⇒ 00:23:37.180 Demilade Agboola: Give me a second.
171 00:23:38.680 ⇒ 00:23:40.439 Demilade Agboola: Alright. So studio.
172 00:23:40.840 ⇒ 00:23:44.560 Demilade Agboola: And then I went to. I searched by times access.
173 00:23:44.910 ⇒ 00:23:46.200 Caio Velasco: Yeah, perfect.
174 00:23:47.170 ⇒ 00:23:56.380 Demilade Agboola: So once you get here like, I can see that they have like a lot of columns and everything because we got it from redshift. You can see, like the things that he has
175 00:23:57.170 ⇒ 00:24:06.250 Demilade Agboola: right? So it has. Item, total order, total price, all of that stuff, right order, ready revenue
176 00:24:06.924 ⇒ 00:24:10.249 Demilade Agboola: credits used refund total order, subtotal.
177 00:24:10.480 ⇒ 00:24:11.759 Demilade Agboola: Well, that.
178 00:24:12.620 ⇒ 00:24:17.590 Demilade Agboola: And so you can kind of like, just try and see where it’s coming from, like what’s going on with this.
179 00:24:18.310 ⇒ 00:24:25.679 Demilade Agboola: My gut, feeling again without looking too deeply into this, is that a lot of the revenue numbers are coming from?
180 00:24:25.810 ⇒ 00:24:27.240 Demilade Agboola: If you go to graph.
181 00:24:28.120 ⇒ 00:24:30.890 Demilade Agboola: So this is the graph. I just expand it.
182 00:24:31.040 ⇒ 00:24:33.050 Demilade Agboola: and that’s where I go where I was before.
183 00:24:33.680 ⇒ 00:24:38.460 Demilade Agboola: My guess is, it’s coming from supporters. Calculation that flows into
184 00:24:39.715 ⇒ 00:24:45.950 Demilade Agboola: other other level build. So you can just kind of like, go through this flow
185 00:24:46.540 ⇒ 00:24:53.120 Demilade Agboola: and see what’s happening. If not I. It could be this, but I doubt I think this is products.
186 00:24:53.770 ⇒ 00:24:56.410 Demilade Agboola: This is more of product than
187 00:24:57.070 ⇒ 00:24:59.770 Demilade Agboola: than revenue. But you could also just check.
188 00:25:02.110 ⇒ 00:25:10.559 Caio Velasco: Okay. And and you’re saying that maybe they only use shopify now for to sell everything they they sell.
189 00:25:11.330 ⇒ 00:25:15.169 Demilade Agboola: Yeah. So shopify is the source of truth right now.
190 00:25:16.070 ⇒ 00:25:21.549 Caio Velasco: Okay? And if that’s well, that’s true, then
191 00:25:22.940 ⇒ 00:25:33.650 Caio Velasco: theoretically at least, shopify must have everything well defined in terms of. If an order, it’s, you know, requested.
192 00:25:33.850 ⇒ 00:25:39.720 Caio Velasco: That’s you know you have to. I mean, for example, if they have subscription. Is that done in spotify as well.
193 00:25:41.574 ⇒ 00:25:45.690 Demilade Agboola: Subscriptions. I believe we’ve also been shopify, but I’m not sure.
194 00:25:47.030 ⇒ 00:25:47.680 Caio Velasco: Okay.
195 00:25:47.680 ⇒ 00:25:48.680 Demilade Agboola: But in a show.
196 00:25:49.380 ⇒ 00:25:58.849 Caio Velasco: Okay, okay, yeah. But theoretically, I mean, of course, in a beautiful world you would just go into shopify documentation and ask, like, How do you define revenue?
197 00:25:59.230 ⇒ 00:25:59.580 Demilade Agboola: Yeah.
198 00:25:59.580 ⇒ 00:26:04.520 Caio Velasco: It’s not related to urban stem. It’s related to shopify right as a source.
199 00:26:06.200 ⇒ 00:26:07.010 Demilade Agboola: Definitely.
200 00:26:07.290 ⇒ 00:26:08.040 Caio Velasco: Okay.
201 00:26:08.690 ⇒ 00:26:13.969 Demilade Agboola: Then also with this, like trying to understand
202 00:26:14.740 ⇒ 00:26:26.999 Demilade Agboola: some of their logic as to why they’re doing things like this. So we have dollar dollar columns as discount auto revenue. This, this, this. It’s a lot of these a lot like the guy who was there before. Steve.
203 00:26:27.200 ⇒ 00:26:30.359 Demilade Agboola: he used a lot of like what’s it? Called ginger.
204 00:26:33.000 ⇒ 00:26:38.359 Demilade Agboola: So I mean, you might need to just compile before you before you read.
205 00:26:39.190 ⇒ 00:26:44.390 Demilade Agboola: So if you compile it, it would compile the ginger into like regular simple for you.
206 00:26:47.030 ⇒ 00:26:54.729 Demilade Agboola: And it didn’t. Okay, all right. Yeah. So you cannot read it as regular sequel. If change is a bit too much.
207 00:26:55.420 ⇒ 00:26:56.160 Caio Velasco: Yes.
208 00:26:56.160 ⇒ 00:27:04.350 Demilade Agboola: Yeah, so yeah, so it’s just kind of like understanding. Oh, for this that is here.
209 00:27:05.514 ⇒ 00:27:10.180 Demilade Agboola: Alright. So you can see this is, can you see what I mean by like the dates?
210 00:27:12.490 ⇒ 00:27:17.659 Demilade Agboola: Right? So, for when the subscription id is null, and then discount total.
211 00:27:17.770 ⇒ 00:27:19.669 Demilade Agboola: If this option id is null.
212 00:27:19.920 ⇒ 00:27:28.160 Demilade Agboola: then the discount total is this divided by case when blah blah blah blah blah, this
213 00:27:28.490 ⇒ 00:27:33.519 Demilade Agboola: when it says no, and this is equal to 0. By the way, this is bad sequel
214 00:27:34.220 ⇒ 00:27:36.170 Demilade Agboola: you start with the less. Just
215 00:27:36.650 ⇒ 00:27:41.480 Demilade Agboola: so, you know, I said, it’s possible you start with the less with the most precise
216 00:27:41.620 ⇒ 00:27:43.480 Demilade Agboola: up until the less precise.
217 00:27:44.970 ⇒ 00:27:49.349 Demilade Agboola: So, for instance, when subscription Id is null.
218 00:27:50.670 ⇒ 00:27:55.800 Demilade Agboola: and I’m saying my subscription id is no, and this is this is a subset of this.
219 00:27:56.280 ⇒ 00:28:04.960 Demilade Agboola: because every single time this is true, this would have been true first, st so it would already have ascribed
220 00:28:05.130 ⇒ 00:28:08.229 Demilade Agboola: this 1st part, even when what I’m trying to do is this.
221 00:28:08.640 ⇒ 00:28:09.949 Demilade Agboola: do you know what I’m trying to say.
222 00:28:11.510 ⇒ 00:28:19.020 Caio Velasco: I think so, because I mean, what? Why are they subset of each other? If it’s either subscription or not? A subscription.
223 00:28:19.260 ⇒ 00:28:22.650 Demilade Agboola: So this is a subset of this condition is what I’m trying to say.
224 00:28:23.780 ⇒ 00:28:27.270 Demilade Agboola: It’s like, it’s like me saying, Oh,
225 00:28:29.040 ⇒ 00:28:34.830 Demilade Agboola: When if there’s a group of people, and I’m saying, when the person is a boy.
226 00:28:35.420 ⇒ 00:28:41.850 Demilade Agboola: do this right, and then after that, and I say, when the person is a boy and 6 feet tall
227 00:28:41.970 ⇒ 00:29:06.030 Demilade Agboola: do this, but the truth of the matter is the way these conditions work. Is. It ascribes the 1st condition first, st and then it goes looking for the next condition to ascribe. So what has happened now is, if you’ve done something very general, as your 1st case, when the second case, when doesn’t hit because it’s no longer, it’s no longer looking for it. If that makes any sense, it’s already done. The 1st case when statement
228 00:29:06.420 ⇒ 00:29:10.379 Demilade Agboola: so you generally start from more precise. So like if I’m trying to do something
229 00:29:10.590 ⇒ 00:29:26.680 Demilade Agboola: there, I’ll say, oh, when the person is 6 foot tall and is a boy. Give them this, then, when they are boys, do this, because at that point it handles the 6 foot people differently. 6 foot boys differently from just when they are just boys, if that makes any sense.
230 00:29:26.950 ⇒ 00:29:28.420 Demilade Agboola: So you do, you.
231 00:29:28.762 ⇒ 00:29:29.790 Caio Velasco: From a SQL.
232 00:29:30.230 ⇒ 00:29:37.239 Caio Velasco: Database kind of perspective, right? Not from only a logic perspective, from a logic perspective could be either.
233 00:29:37.450 ⇒ 00:29:42.349 Caio Velasco: but from maybe you’re talking more like how the computer would read prose and everything.
234 00:29:42.690 ⇒ 00:29:47.239 Demilade Agboola: Yeah. So you start from more precise and then go down to to
235 00:29:47.770 ⇒ 00:29:56.279 Demilade Agboola: less precise, if not you. You could realize, like I could go in there and probably find out that some of these things don’t hold true, because describe the 1st thing 1st
236 00:29:57.170 ⇒ 00:30:03.349 Demilade Agboola: and not the next thing. But yeah, this kind, this kind of kind of what I mean by like you go through, and you’re trying to understand what’s going on.
237 00:30:04.255 ⇒ 00:30:08.229 Demilade Agboola: So the the 1st questions you probably would ask would be, okay. So
238 00:30:10.970 ⇒ 00:30:16.829 Demilade Agboola: what is condition? In which? 1st of all, why is subscription? Id? No, in certain cases
239 00:30:17.540 ⇒ 00:30:19.380 Demilade Agboola: right? Things like that.
240 00:30:20.040 ⇒ 00:30:20.560 Demilade Agboola: And.
241 00:30:20.560 ⇒ 00:30:21.869 Caio Velasco: Yeah, yeah, exactly. Yeah.
242 00:30:21.870 ⇒ 00:30:31.950 Demilade Agboola: Yeah. And then next question would be, what does add on equals? 0 mean? What does? Why does? Why are we separating separate? Why are we taking some skus apart?
243 00:30:33.000 ⇒ 00:30:42.839 Demilade Agboola: Why is it only reform, total, or whatever like the actual column is in these cases, maybe in those cases that’s fine
244 00:30:43.495 ⇒ 00:30:49.580 Demilade Agboola: and then, when when we add on equals to one, and also
245 00:30:50.150 ⇒ 00:30:53.199 Demilade Agboola: just confirm, I believe this is the date of the migration.
246 00:30:53.680 ⇒ 00:30:57.799 Demilade Agboola: Alright, so can we take this model out
247 00:30:59.050 ⇒ 00:31:07.279 Demilade Agboola: and just say, like, Hey case, when all of that is this, we put it in a different model. Do all the calculations for that in one place.
248 00:31:08.571 ⇒ 00:31:16.700 Demilade Agboola: And then so we’ll have our tax total order. Subtotal credits given refund total all before this time
249 00:31:17.120 ⇒ 00:31:23.499 Demilade Agboola: in one, in one model. This is our like legacy model, where, like things are not changed anymore.
250 00:31:23.730 ⇒ 00:31:27.819 Demilade Agboola: And then, if this is still some active logic we’re using right now.
251 00:31:28.130 ⇒ 00:31:29.170 Demilade Agboola: Right.
252 00:31:29.900 ⇒ 00:31:32.400 Demilade Agboola: Let’s have that in a column right now.
253 00:31:32.941 ⇒ 00:31:42.229 Demilade Agboola: Also, if this was a condition that was only done because, you know, back, then, the system didn’t recognize this, this, this, this. Okay, cool.
254 00:31:43.080 ⇒ 00:31:57.799 Demilade Agboola: We know how we we’re going to partition, that’s all. As I’m trying to say. So like, the idea is, we’re just trying to partition everything to the point where, like, we can calculate the promo, total order totals everything for the different like
255 00:31:58.050 ⇒ 00:32:05.790 Demilade Agboola: cases. We understand what the different cases are, we can put them into different boxes rather than just having like this model where everything comes in.
256 00:32:07.379 ⇒ 00:32:09.019 Demilade Agboola: Yeah. So.
257 00:32:09.020 ⇒ 00:32:13.279 Caio Velasco: And then, and as I see, those cases would be
258 00:32:14.980 ⇒ 00:32:37.699 Caio Velasco: when when I took look at linear would be each ticket that that we are outdating, because I remember that I saw, like an audit subscription, all the discount, all the refunds, because when we go into table, except then, of course, you have multiple cases. So we would like to understand each one of them first, st and then we’ll move forward. That’s the relation with those tickets.
259 00:32:38.020 ⇒ 00:32:39.280 Caio Velasco: There’s ones.
260 00:32:39.830 ⇒ 00:32:40.500 Caio Velasco: Yeah, that’s right.
261 00:32:40.500 ⇒ 00:32:59.930 Demilade Agboola: But I I would also want you to see one of the things that can help. Efficiency is using them as linked so like. Yes, subscriptions and refunds and discounts are like separate things. Yes, but there’s also some sort of relationship to them, like as your understanding. Why? Because look at it. If you look at the
262 00:33:00.900 ⇒ 00:33:02.410 Demilade Agboola: compiled code.
263 00:33:02.880 ⇒ 00:33:13.469 Demilade Agboola: it’s basically some of the similar calculations being done, which is why they use just to do across everything. It’s kind of similar calculation from subscriptions for credit, for revenues like
264 00:33:13.810 ⇒ 00:33:22.179 Demilade Agboola: it’s not in a way, the difference in terms of like the final value. But the setting processes are the same the way things are handled. Pre migration will still be
265 00:33:22.400 ⇒ 00:33:28.139 Demilade Agboola: different. Like the way when we’re caring for certain. Like product skews.
266 00:33:28.410 ⇒ 00:33:32.979 Demilade Agboola: it appears to be like, you know, so like understanding what those similarities are.
267 00:33:33.790 ⇒ 00:33:56.170 Demilade Agboola: So it would help you just be more efficient. So that if you learn, maybe one thing you might as well that one thing you’ve learned applies to the 3 different things, and you can push those 3 things at once rather than like, you know, only pushing subscriptions and then having another thinking of it like, Oh, I’m coming back to discounts, or I’m coming back to revenue sometimes you might. What you might learn can push all 3 at the same time forward.
268 00:33:56.170 ⇒ 00:33:56.770 Caio Velasco: Yeah.
269 00:33:57.310 ⇒ 00:34:08.679 Caio Velasco: yeah, yeah, no perfect perfect. Now, it’s interesting to see how like your view of it, you would start even from the model and then get the ideas. For some reason it’s interesting that the way I work it’s I would
270 00:34:09.190 ⇒ 00:34:12.640 Caio Velasco: kind of 1st spend time in shopify.
271 00:34:13.610 ⇒ 00:34:24.330 Caio Velasco: how how shopify works, and then, you know, go to your urban stems and then maybe check this. The models like, okay, is this model representing what shopify has to offer?
272 00:34:25.000 ⇒ 00:34:25.610 Demilade Agboola: Okay.
273 00:34:25.610 ⇒ 00:34:36.349 Caio Velasco: And then I would go into those details, as you mentioned, for example, like why, there is a created at Utc less than 2,014. Why are you restricting? Only for that date?
274 00:34:36.610 ⇒ 00:34:53.560 Caio Velasco: Yeah, we have no idea. And well, and then I would move from there. Because, yeah, it seems that there is a lot of overlaps that you don’t need, because at the end of the day that’s that’s what I was trying to say. Revenue has an equation, and you have. And this model has to respect that equation and calculate
275 00:34:53.719 ⇒ 00:34:55.469 Caio Velasco: the roles as they come.
276 00:34:56.139 ⇒ 00:34:56.539 Demilade Agboola: That’s right.
277 00:34:56.540 ⇒ 00:35:05.900 Caio Velasco: There’s, you know, I know that there’s also the thing of like you have order, and the order has to have can have multiple things inside, and some of the themes can
278 00:35:06.040 ⇒ 00:35:18.390 Caio Velasco: be refunded. Some other things, and probably won’t be so. It’s like crazy. But at the end of the day the equation is just one. It’s like, revenue is price times, quantity minus all the crazy stuff.
279 00:35:19.350 ⇒ 00:35:21.180 Caio Velasco: Definitely, definitely.
280 00:35:21.550 ⇒ 00:35:26.772 Caio Velasco: Okay, okay, okay, so at least, like, I have something where to start from.
281 00:35:27.420 ⇒ 00:35:34.869 Caio Velasco: always intrigued by how someone created this. Someone created this this code from from 0 right.
282 00:35:36.050 ⇒ 00:35:52.069 Demilade Agboola: Yeah, that’s their former Dbt person, Steve. What Emily just does is if things are going. If things are not like working properly, she just kind of goes around and kind of, because I think it’s here, or I think it’s, you know, that kind of thing to kind to fix it.
283 00:35:52.511 ⇒ 00:35:57.829 Demilade Agboola: But I want us to be able to not only just fix it, but like, have a good understanding of what’s going on.
284 00:35:57.960 ⇒ 00:36:02.860 Demilade Agboola: and let let that transition into like what we’re building for them. So that, like
285 00:36:03.350 ⇒ 00:36:09.270 Demilade Agboola: the structure is like, I said they, they have good structure. I need to delete this by the way
286 00:36:09.490 ⇒ 00:36:20.730 Demilade Agboola: they have good structure. And they know, okay, this is what’s going on in our models. This is where we have facts, team staging all that stuff. They know where to find different things.
287 00:36:21.238 ⇒ 00:36:27.640 Demilade Agboola: And they understand that like, Hey, this is for inventory, for products. This is for, like, you know.
288 00:36:27.830 ⇒ 00:36:36.629 Demilade Agboola: whatever. So they can go there and start to see what the union of products looks like. What the marks of whatever looks like.
289 00:36:37.480 ⇒ 00:36:38.940 Demilade Agboola: Yeah. So.
290 00:36:39.790 ⇒ 00:36:40.650 Caio Velasco: Okay. Okay.
291 00:36:40.650 ⇒ 00:36:44.560 Demilade Agboola: Helping them have better structure. That’s 1 of the most important things.
292 00:36:45.080 ⇒ 00:36:46.839 Caio Velasco: Yeah, yeah, no, I agree.
293 00:36:48.560 ⇒ 00:36:50.460 Caio Velasco: Let’s see if I have another question here.
294 00:36:57.500 ⇒ 00:37:01.040 Caio Velasco: If I at some point I have a draft
295 00:37:01.270 ⇒ 00:37:07.129 Caio Velasco: of like revenue. Whatever is it? How would I validate that
296 00:37:07.340 ⇒ 00:37:11.560 Caio Velasco: 1st 1st layer would be? Do they have, I mean.
297 00:37:12.020 ⇒ 00:37:26.329 Caio Velasco: just like going to shopify with their account, and seeing how much money they have now, or how much money they had from a day to another, from a window of days of like a period. How would someone validate that.
298 00:37:27.890 ⇒ 00:37:33.099 Demilade Agboola: I would say we could potentially use your dashboard and compare
299 00:37:34.641 ⇒ 00:38:00.129 Demilade Agboola: and if it doesn’t match, that’s fine, like if it now, if it’s very far apart. It’s probably problem. If it’s like little differences, it’s up to us to figure out like, why do we have like a 2,000 difference, it’s probably something big. If it’s a $200 differences potentially a thing of like, hey, based on understanding of how revenue works.
300 00:38:02.100 ⇒ 00:38:09.480 Demilade Agboola: Because one thing to remember is while we’re trying to understand what they’re doing now, there is no guarantee that what they’re doing now is perfect.
301 00:38:10.430 ⇒ 00:38:34.789 Demilade Agboola: Right? So I want you to take things on. That’s why, like I’m saying like things like dissect the model. But obviously, at the end of the day, you’re going to have to understand which might come from like reading, or just like asking important questions. It’s like, okay, this is how you’re calculating revenue now. But what about this scenario? Have you considered that scenario is everything like? Is this sound?
302 00:38:35.040 ⇒ 00:38:44.090 Demilade Agboola: Right? So it’s possible that when we’re done with all this analysis we come up with the model. That we believe is more accurate, and you can be like, yes, this is how I can show you. It’s more accurate.
303 00:38:44.240 ⇒ 00:38:55.370 Demilade Agboola: This is concerns you hadn’t considered before. This is what you hadn’t thought of before, and now we’re considering it. And so now, your revenue is down by 10% or 5%, or whatever.
304 00:38:55.620 ⇒ 00:39:10.640 Demilade Agboola: This is the difference, right? Or it’s up by 5%. This is why, we think it’s up by 5%. Now, obviously, I don’t expect that we would have like a hundred percent difference. It’s 100% difference. I think that that would suggest that we’re doing something wrong.
305 00:39:11.229 ⇒ 00:39:36.060 Demilade Agboola: But yes, it might not match exactly. But once we can figure out okay, these, this in this scenario in these rows. This is what you are calculating for revenue, and based off the logic and conversations with key stakeholders. And how things have you know, worked in the past. I actually don’t think this is revenue. This should not count as revenue. It’s a this redelivery. It’s a whatever situation.
306 00:39:36.110 ⇒ 00:39:41.760 Demilade Agboola: and that’s why, there’s a slight difference in our output versus what is in the dashboard.
307 00:39:42.360 ⇒ 00:39:43.609 Demilade Agboola: So there’s that
308 00:39:43.720 ⇒ 00:39:52.350 Demilade Agboola: also, if you want to do like line by line. So that’s for aggregated data. But if you actually still want to do like line by line, you can always take the
309 00:39:53.370 ⇒ 00:39:59.130 Demilade Agboola: other details like the other Id and go to shopify and always confirm, so that also works as well.
310 00:39:59.830 ⇒ 00:40:09.060 Caio Velasco: Okay, but even shopify, I mean, who is paying urban stamps? It’s shopify, right? Shopify at some point of the month would send money to
311 00:40:09.450 ⇒ 00:40:11.340 Caio Velasco: to urban stems, or is it?
312 00:40:11.880 ⇒ 00:40:27.080 Caio Velasco: I mean, it depends how this, the pay payment gateway system, or those things. But I believe that at some point someone is sending a hundred 1,000 at that day has to match with what we are doing at some point.
313 00:40:27.240 ⇒ 00:40:32.640 Caio Velasco: No like, that’s what I mean. That would be the best possible scenario of validation.
314 00:40:33.130 ⇒ 00:40:33.890 Caio Velasco: Right.
315 00:40:35.490 ⇒ 00:40:40.750 Demilade Agboola: But I think it’s probably shopify that gives them the money. Again, you could ask Emmy, and just confirm
316 00:40:41.523 ⇒ 00:40:45.789 Demilade Agboola: unfortunately, she has too much because it’s possible
317 00:40:46.250 ⇒ 00:40:52.150 Demilade Agboola: that shopify’s business logic is slightly different from how urban systems calculates it.
318 00:40:52.640 ⇒ 00:40:53.720 Demilade Agboola: And
319 00:40:54.530 ⇒ 00:41:19.629 Demilade Agboola: they might need to push back on shopify and go. Hey? Actually based on our internal calculations, we should get more revenue than you’re currently giving us. So like, it’s very important to like, because sometimes that’s what happens. They just realize that, hey? We should actually get a bit more revenue, because this is what we think, because shopify at the end of day is not infallible. It’s the same way a bunch of them is calculating. Revenue shopify is also trying to calculate their revenue.
320 00:41:20.170 ⇒ 00:41:32.749 Demilade Agboola: They might like some. They might be an oversight there might be misunderstanding. Shopify my reason, the logic, general logic that they would use for most of their customers, which doesn’t always work in open stems use case things like that. So.
321 00:41:32.750 ⇒ 00:41:33.360 Caio Velasco: Okay.
322 00:41:34.290 ⇒ 00:41:40.680 Demilade Agboola: I, I would say yes, go in with the mentality of like.
323 00:41:41.530 ⇒ 00:41:43.660 Demilade Agboola: This is what I consider to be the truth.
324 00:41:44.350 ⇒ 00:41:51.479 Demilade Agboola: and understand like, test it for the different use cases trying to understand what’s going on. And if there’s a disparity
325 00:41:51.600 ⇒ 00:41:54.200 Demilade Agboola: trying to understand what that disparity is.
326 00:41:54.520 ⇒ 00:42:01.329 Demilade Agboola: So sometimes, if you have that disparity. You can ask us a question to Emily, or like another stakeholder, maybe Perry or fat or
327 00:42:01.450 ⇒ 00:42:03.970 Demilade Agboola: trying to remember his name. I think it’s Fabio.
328 00:42:07.700 ⇒ 00:42:11.909 Demilade Agboola: I was trying to remember his name, Felipe. You can ask Felipe and be like
329 00:42:12.580 ⇒ 00:42:13.520 Demilade Agboola: That’s
330 00:42:13.730 ⇒ 00:42:24.729 Demilade Agboola: this is what we’ve discovered this this this, and then they can say, actually, no, you might need to change your logic to fit this. That’s fine like, that’s not something to consider. But let’s let us know.
331 00:42:25.050 ⇒ 00:42:42.419 Demilade Agboola: like, I want us to be open minded, but not so open minded like we should understand what like, what matters to them, but also ask questions that show that we’re actually just thinking about the like use cases and potential things they may have not looked at or considered. So it’s that balance of
332 00:42:43.083 ⇒ 00:42:50.839 Demilade Agboola: you know we don’t like what things look like before, but like we can also see some of the numbers. Just don’t add up.
333 00:42:51.230 ⇒ 00:42:51.950 Demilade Agboola: or
334 00:42:52.670 ⇒ 00:43:16.239 Demilade Agboola: and this is why it doesn’t add up. Or have you considered this use case? Where, like the flowers spoil this happens that happens. And how do you account for like wastage, for instance, maybe, does that count? Does that also count in revenue? Or does that? Is that a buffer thing? So things like that like you. Just you’re asking those questions. But you also like understanding what they’re currently doing as well.
335 00:43:17.090 ⇒ 00:43:17.890 Caio Velasco: Okay?
336 00:43:21.959 ⇒ 00:43:26.020 Caio Velasco: Last question just out of interest. Do they have any
337 00:43:26.200 ⇒ 00:43:29.859 Caio Velasco: prediction models? Do they work with machine learning at all?
338 00:43:30.860 ⇒ 00:43:31.870 Demilade Agboola: I wouldn’t say.
339 00:43:31.870 ⇒ 00:43:33.159 Caio Velasco: Something, for later.
340 00:43:33.640 ⇒ 00:43:38.289 Demilade Agboola: But I do know Perry is supposed to be responsible for forecasting.
341 00:43:38.850 ⇒ 00:43:44.469 Demilade Agboola: so I don’t. You might have to ask her how she does her forecasting how things go for the
342 00:43:44.570 ⇒ 00:43:45.170 Demilade Agboola: I think.
343 00:43:45.170 ⇒ 00:43:51.910 Caio Velasco: Because I, when you mentioned the buffer that the 1st thing came on my mind, how would you calculate, Buffer? How do you predict that.
344 00:43:53.770 ⇒ 00:44:00.290 Demilade Agboola: Yeah, those are important questions. Maybe but that’s more event related. But like, potentially, yeah, that’s 1 of those things where
345 00:44:01.314 ⇒ 00:44:17.799 Demilade Agboola: we could also ask like, How do you have buffer at the times where you have too much in the buffer? Other times you have too little in the buffer, and maybe we can help them figure out ways in which they can have, like a good amount of buffer for the different locations. I don’t know. We’ll we’ll see.
346 00:44:19.950 ⇒ 00:44:25.490 Caio Velasco: Okay, perfect. I mean, I think I have a lot of information. I’ll start to like, you know. Put my head on this, and
347 00:44:25.680 ⇒ 00:44:27.555 Caio Velasco: well, let’s see what happens.
348 00:44:28.070 ⇒ 00:44:29.390 Demilade Agboola: Sounds good.
349 00:44:29.620 ⇒ 00:44:31.370 Caio Velasco: Thank you very much. I appreciate it.
350 00:44:31.940 ⇒ 00:44:33.870 Demilade Agboola: No problem. Have a great day.
351 00:44:33.870 ⇒ 00:44:34.609 Caio Velasco: You too.
352 00:44:35.160 ⇒ 00:44:36.030 Demilade Agboola: Right.