Meeting Title: Data Model Design Sync Date: 2025-07-14 Meeting participants: Demilade Agboola, Caio Velasco
WEBVTT
1 00:01:24.470 ⇒ 00:01:25.560 Caio Velasco: Hi Dimath.
2 00:01:25.910 ⇒ 00:01:26.919 Demilade Agboola: I’m how are you?
3 00:01:27.630 ⇒ 00:01:28.590 Caio Velasco: Good! How are you?
4 00:01:29.050 ⇒ 00:01:36.459 Demilade Agboola: I’m doing very well. So I think we could start off with the technical design document.
5 00:01:39.870 ⇒ 00:01:44.580 Demilade Agboola: So I think for us in terms of the document.
6 00:01:45.370 ⇒ 00:01:51.560 Demilade Agboola: We need to be able to firstly build the models on themselves, ensure data quality
7 00:01:52.730 ⇒ 00:01:58.989 Demilade Agboola: and then and well, this quality it will. You can decide this quality by comparing
8 00:01:59.540 ⇒ 00:02:03.529 Demilade Agboola: what currently exists and also do validation against the raw data.
9 00:02:04.160 ⇒ 00:02:14.690 Demilade Agboola: So we can have validation sessions with the stakeholders, because usually stakeholders work with the data closely, and
10 00:02:14.950 ⇒ 00:02:17.459 Demilade Agboola: they’re able to sense when things feel off
11 00:02:17.590 ⇒ 00:02:21.449 Demilade Agboola: as well as the fact. They also have access to the actual like
12 00:02:21.780 ⇒ 00:02:29.970 Demilade Agboola: distance. Do things like shopify things like dash so they can netsuite. They can compare. And just be sure that numbers work.
13 00:02:32.410 ⇒ 00:02:36.809 Demilade Agboola: So yes, we can have that we would have the data policy test.
14 00:02:37.080 ⇒ 00:02:38.950 Demilade Agboola: And then, in case of large
15 00:02:39.390 ⇒ 00:02:45.629 Demilade Agboola: data that doesn’t change, especially after a while, we can then add incremental processes.
16 00:02:46.090 ⇒ 00:02:52.120 Demilade Agboola: And then also, we need to also add things like data documentation
17 00:02:52.230 ⇒ 00:02:59.460 Demilade Agboola: as well as data, observability slash test depending on what is needed for be processed.
18 00:02:59.830 ⇒ 00:03:02.310 Demilade Agboola: and then the final step will be rollouts.
19 00:03:02.720 ⇒ 00:03:08.479 Demilade Agboola: The rollout will be how do we get the data to them? Because it’s in? Look up.
20 00:03:08.770 ⇒ 00:03:13.069 Demilade Agboola: And so usually, what happens is, once we have the marks, Emily.
21 00:03:13.840 ⇒ 00:03:17.320 Demilade Agboola: it goes into looker and to look, and all creates
22 00:03:18.120 ⇒ 00:03:26.120 Demilade Agboola: like just basically creates the the look ml version of it, but people can access it and use it in. Look up.
23 00:03:27.244 ⇒ 00:03:31.360 Demilade Agboola: I think that will be the high level idea of what the process is.
24 00:03:31.660 ⇒ 00:03:35.509 Demilade Agboola: do you have anything that you feel? I I might be overlooking.
25 00:03:37.820 ⇒ 00:03:41.447 Caio Velasco: Well, I think so. Basically, what we are proposing is
26 00:03:43.450 ⇒ 00:03:46.800 Caio Velasco: it’s almost like a full data governance kind of
27 00:03:49.920 ⇒ 00:03:53.119 Caio Velasco: when. And this is connected to our
28 00:03:54.020 ⇒ 00:04:08.759 Caio Velasco: work of rebuilding, refactoring, I mean, or or making it better, their models for each market that we are building, that that’s basically it right it could be for revenue, it could be for inventory, for anything else.
29 00:04:09.140 ⇒ 00:04:10.900 Caio Velasco: like something from end to end.
30 00:04:11.970 ⇒ 00:04:16.990 Demilade Agboola: Yes, so there there is, so at least from the the tickets assigned to us in
31 00:04:17.320 ⇒ 00:04:25.799 Demilade Agboola: in looker and nonlinear. It’s we we need to be able to create a design document. So first, st we can also say, these are the steps
32 00:04:26.180 ⇒ 00:04:29.110 Demilade Agboola: that we go through before we say things are done.
33 00:04:29.710 ⇒ 00:04:30.850 Demilade Agboola: And
34 00:04:31.070 ⇒ 00:04:43.000 Demilade Agboola: this is how far along we are. This is what we’ve done. So we’ve built out the models. We’ve done the top quality test. But we’ve not yet done documentation and observability yet. So that’s what’s still left.
35 00:04:43.723 ⇒ 00:04:47.970 Demilade Agboola: Or we’ve done this this this, but we may roll it out to looker.
36 00:04:48.500 ⇒ 00:04:53.669 Demilade Agboola: It allows us to be able to see, like what is happening, where things that might be going wrong.
37 00:04:54.180 ⇒ 00:04:59.370 Demilade Agboola: and obviously by the time we hand it over to them. It allows them to be able to maintain
38 00:04:59.660 ⇒ 00:05:04.649 Demilade Agboola: infrastructure because they understand where the documentation is. They also understand what they need to do
39 00:05:05.190 ⇒ 00:05:07.369 Demilade Agboola: to build out their own models, as well.
40 00:05:08.370 ⇒ 00:05:09.759 Caio Velasco: Okay, okay, okay.
41 00:05:10.376 ⇒ 00:05:23.969 Caio Velasco: and do, we need to get in is like, just, it’s literally like this overview that you just gave me. Or do we have to go into details of like, what would be the data quality test, for example, or or any other
42 00:05:24.090 ⇒ 00:05:25.070 Caio Velasco: practice.
43 00:05:26.447 ⇒ 00:05:33.372 Demilade Agboola: I think we can do both. I think we can start off with the high level I don’t know where how soon you know,
44 00:05:33.890 ⇒ 00:05:36.990 Demilade Agboola: Amber wants to meet with Zach and Utam and Alex
45 00:05:38.920 ⇒ 00:05:42.799 Demilade Agboola: but effectively, what we can do is we can do a high level version.
46 00:05:43.477 ⇒ 00:05:50.379 Demilade Agboola: We can turn that around today, if and then we can see if you want a more in depth version. That might be tomorrow.
47 00:05:50.770 ⇒ 00:05:59.510 Demilade Agboola: basically just a bit more time for select flesh it out, or you could use chat, gpt, to flesh it out. That’s also possible. If you do the high level version and can use chatgpt to make it
48 00:06:01.300 ⇒ 00:06:05.800 Demilade Agboola: the fleshed out version. So yeah, so you can send that also today, if that’s possible.
49 00:06:06.730 ⇒ 00:06:08.080 Caio Velasco: Okay, okay, cool.
50 00:06:08.631 ⇒ 00:06:13.129 Caio Velasco: And is this a deliverable that I should do you should do? Is it for amber?
51 00:06:15.150 ⇒ 00:06:19.370 Demilade Agboola: I think it’d be helpful if I don’t know.
52 00:06:20.060 ⇒ 00:06:34.829 Demilade Agboola: No, I could do it. It’s it’s fine. I will just need to sit down and just put the the thing exactly. I mean, I could use the meeting notes to be fair, like, I just said, it’s basically what the meeting. Yeah, just basically this, and just put into chat. Gpt.
53 00:06:35.160 ⇒ 00:06:37.399 Demilade Agboola: you can flesh it out a bit. Yeah, that’s fine.
54 00:06:38.210 ⇒ 00:06:39.440 Caio Velasco: No perfect. Then.
55 00:06:40.176 ⇒ 00:06:43.140 Caio Velasco: Okay. So I think for that part, it’s very clear.
56 00:06:43.970 ⇒ 00:06:44.530 Demilade Agboola: Okay.
57 00:06:44.900 ⇒ 00:06:56.590 Caio Velasco: 1. 1 other thing that I could ask you is like I could show briefly the the revenue work that I’m doing, so that you can at least understand what I’m doing, how I’m doing
58 00:06:56.930 ⇒ 00:07:08.399 Caio Velasco: and see if, like, at least, I’m going the right direction. Sometimes I think it takes a lot of time. But I I couldn’t do other way. I mean, I’m learning which, which is a good feedback for myself.
59 00:07:10.270 ⇒ 00:07:22.640 Caio Velasco: and yeah, but I’m not sure like, how could I jump to something more like in the middle of this process without losing a lot of learning points along the way, and then kind of getting stuck, most likely.
60 00:07:22.980 ⇒ 00:07:27.140 Caio Velasco: So. Well, I will share my screen like real quick, you know.
61 00:07:27.140 ⇒ 00:07:27.770 Demilade Agboola: Okay.
62 00:07:27.770 ⇒ 00:07:28.579 Caio Velasco: I’ll show you.
63 00:07:31.110 ⇒ 00:07:34.460 Caio Velasco: I think this one here.
64 00:07:38.850 ⇒ 00:07:40.730 Caio Velasco: You probably see my browser right.
65 00:07:41.800 ⇒ 00:07:44.480 Demilade Agboola: Yes, I can see your redshift square at the top.
66 00:07:45.030 ⇒ 00:07:51.325 Caio Velasco: Okay, perfect. So well, in the beginning, I was just basically using
67 00:07:52.330 ⇒ 00:07:55.570 Caio Velasco: more like this one, not even this one, actually. But
68 00:07:57.250 ⇒ 00:07:59.040 Caio Velasco: Let me see what where I was
69 00:07:59.870 ⇒ 00:08:04.280 Caio Velasco: in the beginning I was just trying to like select from tableau
70 00:08:04.710 ⇒ 00:08:09.709 Caio Velasco: and then going down into the the lineage. It’s lineage.
71 00:08:10.270 ⇒ 00:08:21.029 Caio Velasco: But then it’s like, Okay, there’s like each lineage has, like so many other things. And then at least, I was able to like, get our order. See? Along the process. Okay, nice.
72 00:08:21.170 ⇒ 00:08:27.500 Caio Velasco: But then, after I had to like restart from the beginning and say, like what I’m trying to do here.
73 00:08:27.730 ⇒ 00:08:34.759 Caio Velasco: So then I well, I also learning along the process that you know like this, for example, this 1st source
74 00:08:35.049 ⇒ 00:08:42.330 Caio Velasco: which builds the stage suborders it’s a very important one, and
75 00:08:42.840 ⇒ 00:08:50.579 Caio Velasco: it exists because I didn’t ever know. Like, okay, I remember the work from evil ingestion. So I mean, I get, I’m getting acquainted with with the with everything.
76 00:08:51.209 ⇒ 00:09:01.190 Caio Velasco: And while I go here in Hevo sometimes. And I check like, okay, this is the well, whatever like comments, orders, Ms customers of orders
77 00:09:01.360 ⇒ 00:09:07.099 Caio Velasco: also, just to help me go through columns. But then well, I go back to
78 00:09:07.420 ⇒ 00:09:11.050 Caio Velasco: the red redshift. And then what I’m doing
79 00:09:11.180 ⇒ 00:09:25.580 Caio Velasco: is, first, st I just query, basically to understand like what is in there. And when I do many columns like sometimes, they’re like 500 columns. But 90% are just new values, which I also don’t know why. But okay.
80 00:09:26.990 ⇒ 00:09:35.180 Caio Velasco: Then, after this, I noticed that this, for example, builds the stage suborders. So if I go to.
81 00:09:35.560 ⇒ 00:09:38.890 Caio Velasco: I actually had to bring the other tab here.
82 00:09:39.440 ⇒ 00:09:54.250 Caio Velasco: Oh, so if I go to DVD. Cloud State suborders, it’s getting directly from where’s Zoom?
83 00:09:55.670 ⇒ 00:10:18.750 Caio Velasco: Here from the source. And then I compile. I see what that’s that’s exactly that one. And then what I was trying to do was basically to rebuild this and understand why they are using these values. Why, what are these things? Why, they’re using this function. And then the basically the deduplication idea behind it. Based on something that is canceled, something delivered, whatever.
84 00:10:19.000 ⇒ 00:10:21.699 Caio Velasco: So this is like me learning, and then
85 00:10:21.820 ⇒ 00:10:42.940 Caio Velasco: I reproduce it here after with the cte. So I get my 1st select from above. I improve it kind of do the same as in Dbt. Cloud, and then I filter as if it was almost like the incremental and the duplication idea. And then, okay. And then this gives me a good idea of stay suborders.
86 00:10:43.230 ⇒ 00:10:56.129 Caio Velasco: And I was like, Okay, there is the same for stage orders. So I did the whole thing for stage orders from the source. Same idea understanding that, for example, now you have some subscription, did sales that
87 00:10:56.260 ⇒ 00:11:16.190 Caio Velasco: you had columns here in the 1st one and stage suborders, but they were new, and for some reason, in a stage orders they are not new. You have subscription details. So then, I generate some questions. For example, why subscriptions are just here. Is that why this is important, or is this something else? And then I put these questions
88 00:11:16.320 ⇒ 00:11:17.999 Caio Velasco: just so that I don’t forget
89 00:11:18.260 ⇒ 00:11:27.659 Caio Velasco: in this sheet like a Q&A, so basically, I was doing looking at this source, trying to build stage of orders, tracing an order
90 00:11:28.840 ⇒ 00:11:31.080 Caio Velasco: one of the selects from there.
91 00:11:31.664 ⇒ 00:11:42.929 Caio Velasco: And then some questions, and then I was well, this would be me going with Emily, maybe in the working session, or or just well, she can take time and come here and answer if she prefers.
92 00:11:43.671 ⇒ 00:11:46.408 Caio Velasco: So that’s the process that I’m doing.
93 00:11:47.270 ⇒ 00:11:52.229 Caio Velasco: the next one, just so that, you know, would be stage transactions.
94 00:11:52.350 ⇒ 00:12:05.979 Caio Velasco: which I understand now is basically well payments. I thought that transactions mean orders. But transactions like payment transactions. So okay, so it’s okay, quite easy. One same ideas before same query, same city.
95 00:12:06.810 ⇒ 00:12:15.220 Caio Velasco: And this is bringing credit card information good. And then I was the next one, which I also saw. A bit would be line items.
96 00:12:15.460 ⇒ 00:12:18.699 Caio Velasco: And then I also saw that Emily pointed out that.
97 00:12:18.930 ⇒ 00:12:24.020 Caio Velasco: for example, where is it dpt audit?
98 00:12:24.540 ⇒ 00:12:27.440 Caio Velasco: She said, that this
99 00:12:28.130 ⇒ 00:12:43.340 Caio Velasco: this one here stage split items has. I don’t know a lot of errors, or whatever. She was explaining a lot of things here. But when but then, let me bring you to the pain point I had before one time, 2 months ago she was going through things with me and explaining this thing.
100 00:12:43.680 ⇒ 00:12:59.209 Caio Velasco: I cannot understand this thing before going through all the work that I’m doing right now. So it’s like kind of like going step by step. I know. Now I will get to this one. I will build the stage line items. I will build the other one, too. Apparently there’s another one.
101 00:12:59.420 ⇒ 00:13:01.929 Caio Velasco: And then I was like, Okay, now I see what has happened.
102 00:13:02.150 ⇒ 00:13:14.029 Caio Velasco: I don’t see a way to do it differently, and at some point hit subscriptions, hit refunds, hit well the other parts, the other things
103 00:13:14.706 ⇒ 00:13:19.400 Caio Velasco: which are related to the linear tickets, and
104 00:13:20.230 ⇒ 00:13:37.119 Caio Velasco: maybe in those linear tickets, if someone had knowledge already, they could say, Okay, if you get to the point of rebuilding stage line items, you know, carry about these problems. So then, I know that I have to go to evil, get the main source and do the work that I’m doing.
105 00:13:37.280 ⇒ 00:13:58.400 Caio Velasco: But since I didn’t know this previously. Now, you know, we could go and re update all the linear tickets, which makes no sense. But just for for bringing you to the pain of not having a clear view of what the work was done. And now that I’m going through the work, I’m kind of understanding what the tickets should be, in the 1st place, right?
106 00:13:58.950 ⇒ 00:14:01.670 Caio Velasco: Which is like learning on the job for me, at least.
107 00:14:01.830 ⇒ 00:14:08.569 Caio Velasco: So that’s what I’m doing. Well, feel free to tell me anything, any feedback if it’s in the right direction.
108 00:14:09.251 ⇒ 00:14:27.299 Caio Velasco: I just hit subscriptions or credit card details. Now, for example, so I’m assuming that I am auditing transaction, and maybe I’m auditing subscription already because they appeared here. Is that making sense? This process.
109 00:14:28.950 ⇒ 00:14:35.289 Demilade Agboola: Yes, it makes sense that you need to go through and understand what’s going. I also go through.
110 00:14:36.790 ⇒ 00:14:37.980 Demilade Agboola: I
111 00:14:38.230 ⇒ 00:14:44.149 Demilade Agboola: I will say, though, that when I go through and everyone has their own process, I don’t necessarily
112 00:14:45.280 ⇒ 00:14:55.959 Demilade Agboola: break it so finely like I just kind of go through. Run it limited to 800. Get some particular ones. Look through, try and figure out what’s going on.
113 00:14:56.674 ⇒ 00:15:04.590 Demilade Agboola: So yeah, it’s it’s normal to go through the tables, but something I also want you to think of is they have a lot of models.
114 00:15:04.770 ⇒ 00:15:08.859 Demilade Agboola: I think it doesn’t make sense how their infrastructure is built.
115 00:15:10.280 ⇒ 00:15:15.000 Demilade Agboola: I want you to think of. How can we build 80?
116 00:15:15.280 ⇒ 00:15:22.580 Demilade Agboola: So which is part of wiring by auditing so without even actually building? And that’s kind of why, I said, the building can be separate from
117 00:15:22.780 ⇒ 00:15:26.269 Demilade Agboola: the auditing is, how can I build
118 00:15:27.020 ⇒ 00:15:32.020 Demilade Agboola: a a simple fact? Orders, or a simple facts of orders.
119 00:15:32.120 ⇒ 00:15:58.359 Demilade Agboola: or a simple fact transaction or a simple game orders. Or you know, basically, how can we simplify this so that we don’t have all this staging Oms, orders staging oms or borders. Then it goes into transactions, except, and then we have tableauxf, and it’s just so confusing. The name. Model names are not very clear, and sometimes the only things that happen are very small, very, very small changes.
120 00:15:58.660 ⇒ 00:16:10.260 Demilade Agboola: And it’s just the same thing to 4 different models. So how can we have a line where it’s just staging models, box models or staging models?
121 00:16:12.620 ⇒ 00:16:24.599 Demilade Agboola: fax models and aggregation models? So we’ve aggregated orders by dates, daily orders. Blah blah like, or daily orders by product or daily orders by, you know, states
122 00:16:24.740 ⇒ 00:16:28.289 Demilade Agboola: like, How do we just make it? Very simple
123 00:16:28.400 ⇒ 00:16:30.550 Demilade Agboola: that we don’t have like these
124 00:16:31.790 ⇒ 00:16:38.509 Demilade Agboola: many. Exactly. So I want you to have that in mind and think of like, how can we compress the lineage into place like
125 00:16:38.640 ⇒ 00:16:42.330 Demilade Agboola: into a much smaller, tidier infrastructure.
126 00:16:42.810 ⇒ 00:16:48.409 Demilade Agboola: As you’re doing this. So it’s not just like only understanding, but also kind of having a vision of what
127 00:16:48.540 ⇒ 00:17:01.980 Demilade Agboola: the future looks like. So that way we can talk to Emily, or whoever we can also see as we’ve seen it, we’ve seen what is currently existing. We understand how it works, and we understand the best way to be able to put this in a way that is scalable over time.
128 00:17:02.320 ⇒ 00:17:05.819 Demilade Agboola: because what they have is not scalable, and they ask part of why they are suffering.
129 00:17:06.619 ⇒ 00:17:07.389 Caio Velasco: Okay.
130 00:17:07.499 ⇒ 00:17:18.589 Caio Velasco: okay, no perfect. That’s a that’s a very good point. I’ll definitely keep that in mind. And for example, at least this part. Let’s call that I’m doing a bit of discovery, or maybe even auditing.
131 00:17:18.709 ⇒ 00:17:37.539 Caio Velasco: For example, now I see that there are tables that brings order Ids. There are tables that bring suborders. There are tables that bring line items, and maybe now a different problem with that line items from what she said. So when I would be thinking on
132 00:17:38.869 ⇒ 00:18:04.409 Caio Velasco: affect orders. Now, I would know that. Okay, this fact orders needs to have orders of order line. Item, blah blah blah transaction id or something else. Or maybe this would be like dim payments or something different, but without going through, the discovery would be even hard for me to think on what effect orders should have. You know, anyway, you know, like the basic things, sure. But still.
133 00:18:04.949 ⇒ 00:18:23.849 Caio Velasco: whatever they did along the whole lineage. You know, there might be many different things in there. But yeah, I get your point of trying to focus there. So maybe what I see is like me doing this as a discovery which kind of take days, I mean, you cannot lie, and I don’t work full time.
134 00:18:24.029 ⇒ 00:18:25.859 Caio Velasco: so I do the best I can.
135 00:18:26.349 ⇒ 00:18:43.789 Caio Velasco: And then the next phase would be okay. Now I have an idea of transaction. I have an idea of orders, of order, line items refund or something else. Then I would probably, from what you’re saying, stop a bit, and even try to build a fact order like a very basic one.
136 00:18:44.029 ⇒ 00:18:52.379 Caio Velasco: and see if the order that goes through the whole lineage they have. Now, it’s also appearing fact orders. Apparently we’re doing the right job.
137 00:18:52.589 ⇒ 00:18:54.169 Caio Velasco: So that does it make sense.
138 00:18:55.956 ⇒ 00:19:02.940 Demilade Agboola: Yes, I I will say, though, that it’s possible we’re trying.
139 00:19:03.350 ⇒ 00:19:07.540 Demilade Agboola: Do 2 things at the same time in the sense of
140 00:19:08.090 ⇒ 00:19:15.260 Demilade Agboola: right now. If you go to like Dvc. Cloud, I have. If you go to the new model structure I have created like a
141 00:19:15.810 ⇒ 00:19:17.930 Demilade Agboola: if it’s a new model structure.
142 00:19:22.980 ⇒ 00:19:29.230 Demilade Agboola: I go to staging unclear, and then you go to orders.
143 00:19:31.330 ⇒ 00:19:34.899 Demilade Agboola: I also like putting things in a way that makes sense.
144 00:19:35.030 ⇒ 00:19:39.870 Demilade Agboola: You can understand staging Oms activity, delivery, established legacy.
145 00:19:40.040 ⇒ 00:19:47.490 Demilade Agboola: redeliveries, orders. And it’s just based off what they did. It’s literally like the quite similar.
146 00:19:47.830 ⇒ 00:19:54.249 Demilade Agboola: Do what? It’s not the same. I just also listed. I subselect all I can list it out so you can also see what
147 00:19:54.390 ⇒ 00:19:58.840 Demilade Agboola: the column names are. But that is.
148 00:19:58.940 ⇒ 00:20:01.200 Demilade Agboola: if we’re going to now build effect orders.
149 00:20:01.450 ⇒ 00:20:14.290 Demilade Agboola: We can also build a fact orders, but also think of like, what are the things they’re doing in all the models and start adding them to the fact orders. Also I see that in Oms supporters there’s an orders column where they’re applying this business logic
150 00:20:14.480 ⇒ 00:20:22.439 Demilade Agboola: added to fact orders. There’s this, this, this where they are doing all of this added to fact. Orders like the idea is.
151 00:20:22.900 ⇒ 00:20:25.100 Demilade Agboola: I don’t want you to think that
152 00:20:25.530 ⇒ 00:20:28.729 Demilade Agboola: things only have to happen in parallel
153 00:20:30.399 ⇒ 00:20:35.130 Demilade Agboola: or serially sorry, like things only have to happen one after the other when it can happen in parallel.
154 00:20:36.300 ⇒ 00:20:42.809 Demilade Agboola: I think that allows you to be faster and allows you to be able to utilize things. It’s a bit better
155 00:20:43.150 ⇒ 00:20:51.709 Demilade Agboola: when you think of like, Hey, while you’re writing out the bits of logic that agents calculates revenue discounts this, that they’re also like
156 00:20:51.900 ⇒ 00:21:12.679 Demilade Agboola: in the different models that it would be in. You put that logic there for the framework of what factor does look like, or fact, transaction or fact refunds, or whatever or fact order refunds will look like, so that you don’t have to. Now, when you’re building again, or when you want to start building building from scratch, you already have a framework based off things you’ve put together
157 00:21:13.040 ⇒ 00:21:20.989 Demilade Agboola: of how the the model will look, because ultimately, at the end of day we need to be. Also, we. We need to deliver because people want to see.
158 00:21:21.150 ⇒ 00:21:24.119 Demilade Agboola: you know, results. People want to be able to use their data.
159 00:21:24.340 ⇒ 00:21:35.239 Demilade Agboola: And we we don’t want it to be like, we’re spending so much time auditing or spending so much time doing things that no one has any data available which is kind of why, even this inventory stuff.
160 00:21:35.350 ⇒ 00:21:46.340 Demilade Agboola: I’ll try and see if we can get it out this week, and they have that data, let them be able to use their data. And if there are any questions we can always come back and like, figure out some of the issues.
161 00:21:47.134 ⇒ 00:21:52.239 Demilade Agboola: But I I feel like perfection should not be the enemy of of progress.
162 00:21:52.740 ⇒ 00:21:56.359 Demilade Agboola: like, because we’re trying to be perfect, we should not make any progress.
163 00:21:57.600 ⇒ 00:22:01.750 Caio Velasco: Okay, okay? And and given all this that you just said.
164 00:22:03.130 ⇒ 00:22:05.430 Caio Velasco: So the deliverable would be
165 00:22:06.290 ⇒ 00:22:12.459 Caio Velasco: like an A 1st attempt. The 1st Mvp. Of a fact. Order table, for example.
166 00:22:17.500 ⇒ 00:22:19.150 Demilade Agboola: A bits.
167 00:22:19.991 ⇒ 00:22:22.819 Demilade Agboola: It also can also be in.
168 00:22:23.840 ⇒ 00:22:31.460 Demilade Agboola: It can also be for the audits. It can be quick documentation like it doesn’t have to be.
169 00:22:31.910 ⇒ 00:22:47.640 Demilade Agboola: This is currently how this exists there. Currently, a lot of duplications have been across these 4 different models. This can be compression to one model called fact Orders. Where we have a lot of this logic like this, this, this, this, this done here?
170 00:22:48.380 ⇒ 00:22:51.259 Demilade Agboola: again, just something. It can be a 1 page
171 00:22:51.400 ⇒ 00:22:53.900 Demilade Agboola: doesn’t again, doesn’t have bits with longest theme.
172 00:22:53.990 ⇒ 00:23:02.669 Demilade Agboola: but something that just shows that. You know we’ve gone through this document. This is these are things we’ve noticed that being done, these are the inefficiencies that are being there.
173 00:23:02.740 ⇒ 00:23:26.600 Demilade Agboola: Incremental model can be made better or more efficient. The fact that there are so many models means that every day your other run takes so long, because if it’s doing 4 different models that takes you a long time for each of the models, because others a lot of data. Generally speaking, it’s like, Okay, you just have a quick audit on what’s wrong and what is
174 00:23:27.640 ⇒ 00:23:33.740 Demilade Agboola: what they did wrong to get to that stage, and how things can be better. And the idea is
175 00:23:34.340 ⇒ 00:23:39.720 Demilade Agboola: they’re implementing what we’re implementing shows how things can be better.
176 00:23:40.180 ⇒ 00:23:44.019 Demilade Agboola: So it’s that balance of showing them what can be better
177 00:23:44.960 ⇒ 00:23:55.459 Demilade Agboola: first, st by like a quick documentation, so that we can present to them and the team Emily Zack, Alex. Whoever saying like, Hey, these are things we’ve seen are wrong with your infrastructure.
178 00:23:55.620 ⇒ 00:23:57.640 Demilade Agboola: and these are things we look to fix.
179 00:23:58.510 ⇒ 00:24:06.120 Demilade Agboola: And then, naturally fixing. It is the final deliverable, because that’s what they need to get data out to the team and business stakeholders.
180 00:24:08.210 ⇒ 00:24:12.089 Caio Velasco: Okay, okay, so it would. Okay. So maybe even
181 00:24:12.540 ⇒ 00:24:23.669 Caio Velasco: I see this as like I, I would have to have a discovery phase in the beginning to get acquainted, and then the auditing would be doing, as you said, like trying to move quickly with some small ideas.
182 00:24:23.830 ⇒ 00:24:26.980 Caio Velasco: And why? The discovery is important, because
183 00:24:27.590 ⇒ 00:24:32.739 Caio Velasco: I don’t know what what can, what, what fact, orders, or any other I can.
184 00:24:33.290 ⇒ 00:24:38.750 Caio Velasco: Potential table should have, you know, like it depends on the on, on how
185 00:24:39.510 ⇒ 00:24:46.999 Caio Velasco: I mean how Webber understands work. So if there is refunds, if there are subscriptions, if there’s cancellation, there’s a lot of things happening.
186 00:24:48.200 ⇒ 00:25:16.679 Caio Velasco: this has to be represented, in fact, orders. So the discovery process would kind of, as I see, take me towards at least understanding the domains that exist and how they affect each other. For example, you have an order, but they all can be canceled for some reason, can have shipment problem that would cancel payment problem that would cancel refunds for credit card issues or for shipping issues. So the thing is like a monster in itself.
187 00:25:17.150 ⇒ 00:25:20.590 Caio Velasco: And even though I get the idea of
188 00:25:20.720 ⇒ 00:25:30.319 Caio Velasco: proposing effect orders, I still have difficulties, understanding. How can I even know what should be in there without going through
189 00:25:31.120 ⇒ 00:25:33.170 Caio Velasco: a methodical discovery.
190 00:25:36.010 ⇒ 00:25:37.710 Demilade Agboola: That’s what you’re doing right now, aren’t you?
191 00:25:38.350 ⇒ 00:25:45.360 Caio Velasco: Yeah. Yeah. But then, what? As I see, maybe this would take long or too long in terms of what people expect.
192 00:25:45.490 ⇒ 00:25:47.209 Caio Velasco: or is it still on track.
193 00:25:47.957 ⇒ 00:25:51.049 Demilade Agboola: How long do you think it will take the their staff, invest.
194 00:25:51.550 ⇒ 00:26:02.740 Caio Velasco: So, for example, looking at what you did here, if I were to do the same as I’m doing, I still have to do this this the redeliveries, apparently, that I didn’t even know existed.
195 00:26:02.900 ⇒ 00:26:05.110 Caio Velasco: which is part of the discovery.
196 00:26:05.536 ⇒ 00:26:14.700 Caio Velasco: This one here that I don’t know what it’s about, and this would just be orders. If refunds or subscription or anything. It’s within here. I still don’t know.
197 00:26:14.850 ⇒ 00:26:15.810 Caio Velasco: I might
198 00:26:16.370 ⇒ 00:26:23.430 Caio Velasco: see when I’m doing this discovery, but I still don’t know, and if I don’t, and if it’s not here, then it means that discovery should
199 00:26:23.550 ⇒ 00:26:25.340 Caio Velasco: get, you know. Continue
200 00:26:26.660 ⇒ 00:26:31.109 Caio Velasco: So it’s like working in in the dark, you know, but I don’t know. Maybe
201 00:26:31.865 ⇒ 00:26:34.139 Caio Velasco: I have today, and maybe
202 00:26:35.940 ⇒ 00:26:38.455 Caio Velasco: let’s say that I take 2 more days
203 00:26:39.110 ⇒ 00:26:46.179 Caio Velasco: working part time as I do, I would be able to cover these and understand what what’s in there.
204 00:26:46.610 ⇒ 00:26:49.720 Caio Velasco: and assume that this is what should be in the fact orders.
205 00:26:49.830 ⇒ 00:26:55.130 Caio Velasco: And now, if it’s gonna be touching everything that is here,
206 00:26:59.730 ⇒ 00:27:00.549 Caio Velasco: one sec.
207 00:27:01.775 ⇒ 00:27:07.409 Caio Velasco: if it’s gonna be touching everything that is in here, I have no idea, because only after I see that I will know.
208 00:27:09.950 ⇒ 00:27:11.979 Demilade Agboola: Yes, but I want to.
209 00:27:11.980 ⇒ 00:27:19.400 Caio Velasco: Part of like line items, or, you know, credits. I already see it’s if this credits mean payment.
210 00:27:19.590 ⇒ 00:27:21.880 Caio Velasco: They just appeared in the
211 00:27:24.040 ⇒ 00:27:24.970 Caio Velasco: It’s 1 of the.
212 00:27:25.810 ⇒ 00:27:28.900 Demilade Agboola: Here’s what I want you to do. I want you to be able to
213 00:27:29.410 ⇒ 00:27:34.759 Demilade Agboola: have a good idea of what is happening right? But good idea is not a perfect idea.
214 00:27:35.200 ⇒ 00:27:42.280 Demilade Agboola: If you have a good idea, you can have a call with Emily. Run her through the things you’ve noticed and said, Hey.
215 00:27:43.032 ⇒ 00:27:50.350 Demilade Agboola: how do I identify refunds? And any might say, or refunds are found in transactions. This table she sent you that table.
216 00:27:51.082 ⇒ 00:28:01.949 Demilade Agboola: How do I do this, it tells you where you find it. I have noticed that this, this, this, this is how we calculate loyalty points. Okay.
217 00:28:02.280 ⇒ 00:28:20.489 Demilade Agboola: this is how we calculate discounts like you can give her like what you’ve seen, what you’ve observed, she can correct certain things or say actually, no, there’s more information to be found elsewhere. But the idea is, I don’t want you to think you have to have every single thing before you make progress. That’s what that’s all I’m trying to say
218 00:28:20.700 ⇒ 00:28:27.700 Demilade Agboola: you can have a good idea. You can have a decent idea. And then, next thing you’re saying, Hey, Emily, can I have a working session with you tomorrow.
219 00:28:27.820 ⇒ 00:28:50.310 Demilade Agboola: and then in that working session you can run her through again. High level. I have found out. This is how the orders are structured. This is where they live, it seems all orders are in one place, and the way to identify refund is by setting columns this column, that column. This column for discounts this column for loyalty. Is it the same before migration and after migration? She’ll say yes, no.
220 00:28:50.330 ⇒ 00:29:03.339 Demilade Agboola: This is how it is right before migration after migration. Are there are there any edge cases that need to know of, like those sort of questions that she would give you the information about, and you can be able to have it in one place.
221 00:29:03.420 ⇒ 00:29:06.000 Demilade Agboola: and then use that to also make you go faster.
222 00:29:06.570 ⇒ 00:29:10.729 Caio Velasco: Okay, okay, cool. Cool. No. Now, now, give me an idea. Okay.
223 00:29:11.230 ⇒ 00:29:16.250 Caio Velasco: okay, no. Perfect. Perfect. Then. I get it. I’ll
224 00:29:16.400 ⇒ 00:29:23.979 Caio Velasco: but I’ll continue what I’m doing just for the sake of the day. Then maybe tomorrow, I can already have the working session with her
225 00:29:24.744 ⇒ 00:29:35.719 Caio Velasco: and then go exactly through all this, and hopefully, I already know some of it. So when they can ask her like, Where is this discount coming from? Gift cards. Loyalty points. Blah blah. Okay, then.
226 00:29:35.720 ⇒ 00:29:38.350 Caio Velasco: And I guess Tuesday, with the idea of
227 00:29:38.470 ⇒ 00:29:44.129 Caio Velasco: what you just meant like focusing on the fact orders or day mortars or any other future.
228 00:29:45.120 ⇒ 00:29:47.000 Demilade Agboola: Okay. Alright. Sounds good.
229 00:29:47.160 ⇒ 00:29:49.040 Caio Velasco: Alright! Thank you, Emilia. That was helpful.
230 00:29:49.610 ⇒ 00:29:51.430 Demilade Agboola: Yeah, you’re welcome. No problem.
231 00:29:51.810 ⇒ 00:29:52.600 Caio Velasco: Okay.
232 00:29:52.860 ⇒ 00:29:53.430 Demilade Agboola: Bye.
233 00:29:53.430 ⇒ 00:29:54.540 Caio Velasco: See you later. Then bye-bye.