Meeting Title: Brainforge Interview w- Demilade Date: 2026-04-13 Meeting participants: Demilade Agboola, Christina Knudson
WEBVTT
1 00:00:36.930 ⇒ 00:00:37.800 Christina Knudson: Hello.
2 00:00:41.260 ⇒ 00:00:42.209 Christina Knudson: Are you able to hear me?
3 00:00:42.210 ⇒ 00:00:43.389 Demilade Agboola: Yeah. Probably you.
4 00:00:44.170 ⇒ 00:00:45.349 Christina Knudson: Good! How are you doing?
5 00:00:45.520 ⇒ 00:00:51.470 Demilade Agboola: I’m doing very well. My name is Dimulade, and I work with the Brainforge team.
6 00:00:52.710 ⇒ 00:01:00.399 Demilade Agboola: Yay… I work with the Brainford team, and I work as a senior analytics engineer on the team.
7 00:01:00.760 ⇒ 00:01:05.820 Demilade Agboola: And so yeah, this will just be an interview to get an idea of how
8 00:01:06.610 ⇒ 00:01:14.390 Demilade Agboola: You think, and how you solve, like, analytics engineering problems, so it’s not going to be live coding, it’s just going to be more…
9 00:01:14.950 ⇒ 00:01:19.489 Demilade Agboola: We’re gonna walk those scenarios and kind of get an idea of how you handle the different scenarios.
10 00:01:20.030 ⇒ 00:01:20.850 Christina Knudson: Sounds good.
11 00:01:21.010 ⇒ 00:01:23.039 Demilade Agboola: Alright, sounds good.
12 00:01:23.400 ⇒ 00:01:30.370 Demilade Agboola: So just, can you just, like, give me an idea of, like, your background and, the things, like, your tech stack and what you’ve worked with?
13 00:01:30.980 ⇒ 00:01:44.259 Christina Knudson: Yeah, so my background is in statistics, I have a, like, PhD in statistics, and then, with that, then, like, my dissertation was relating to, like.
14 00:01:44.260 ⇒ 00:02:03.099 Christina Knudson: making, like, statistical methodology, so a lot of, like, programming, making, like, an R package, and, like, calling C from R, and then when I started working as, like, a data scientist in the last, like, couple years, then I started getting more and more into dbt, and then in 2025, I was, like.
15 00:02:03.600 ⇒ 00:02:20.969 Christina Knudson: basically most of what I was doing with my time was, like, analytics engineering, because we, like, had to rebuild our whole tech stack, and then rebuild our data stack once we had, like, new data sources coming in, so, was working a lot on that. And so, yeah, mostly using, like, dbt, Snowflake,
16 00:02:21.340 ⇒ 00:02:33.580 Christina Knudson: some of our data we had in Redshift, and so used to that some, yeah, Python… R.C.
17 00:02:34.150 ⇒ 00:02:34.970 Christina Knudson: Okay. Yeah.
18 00:02:35.790 ⇒ 00:02:43.049 Demilade Agboola: Sounds good. Can you, like, just walk me through what you would consider the…
19 00:02:43.240 ⇒ 00:02:48.610 Demilade Agboola: More complex data pipeline or, like, data flow that you had to build out.
20 00:02:48.820 ⇒ 00:02:50.890 Demilade Agboola: And, like, wall media a bit complex.
21 00:02:51.910 ⇒ 00:02:59.370 Christina Knudson: Okay, yeah. So, like, one that I worked on that was maybe more… Complex was…
22 00:03:00.240 ⇒ 00:03:18.009 Christina Knudson: this last year, when we were rebuilding the, like, data stack, and then I was in charge of looking at, like, the volume of product sold, and that was, like, the piece that I was owning of it, and I guess what made it complex was, like, a couple things. One, like.
23 00:03:18.050 ⇒ 00:03:26.459 Christina Knudson: The raw data was coming in… Like, pretty… Like, in a…
24 00:03:26.630 ⇒ 00:03:33.190 Christina Knudson: A lot of different, like, tables, and so then there was, like, a lot of joins involved, and, like, lots of,
25 00:03:33.680 ⇒ 00:03:41.929 Christina Knudson: yeah, like, lots of transformations there and, like, join keys, and I guess part of what made it complex is that there wasn’t, a lot…
26 00:03:42.760 ⇒ 00:03:50.489 Christina Knudson: Or really much at all, like, documentation from the vendor providing the data, and so had to do a lot of, kind of, like.
27 00:03:51.250 ⇒ 00:03:53.100 Christina Knudson: You know, if I do this…
28 00:03:53.190 ⇒ 00:04:17.280 Christina Knudson: then what happens? Does this look reasonable? And then also, like, trying to figure out, like, okay, how much can I figure out on my own, and how much does it make sense just to ask? But also, when you ask, it might take, like, a week for them to get back to you, so obviously you have to, like, keep things moving and, like, doing the best that you can in that, in that meantime. So I guess that was, like, one piece that made it complicated, and then, like, another piece that made it complicated is…
29 00:04:17.800 ⇒ 00:04:22.890 Christina Knudson: It just, like… Needed to be, you know, like, the full…
30 00:04:23.800 ⇒ 00:04:38.930 Christina Knudson: pipeline, the full, like, getting all the way from raw all the way to, like, the mart, and then within the mart, there were, like, a few different flavors that they wanted, but, like, pretty similar, for some of the steps, like, analyses, so, like.
31 00:04:39.060 ⇒ 00:04:41.210 Christina Knudson: For…
32 00:04:41.220 ⇒ 00:04:59.460 Christina Knudson: just to kind of make that more concrete, like, we were selling the product, but then also, when the person becomes a customer, then we got some historical information about their, like, consumption of that product. So then, for some of the resulting MART datasets, we wanted to have
33 00:04:59.460 ⇒ 00:05:02.960 Christina Knudson: Like, both of those, historical and, like…
34 00:05:03.630 ⇒ 00:05:11.119 Christina Knudson: product that we sold in there, and then for other pieces, we just only wanted to know, like, what are we selling them? So…
35 00:05:11.450 ⇒ 00:05:13.749 Christina Knudson: Like, it didn’t…
36 00:05:13.920 ⇒ 00:05:20.189 Christina Knudson: Make it, like, that complex, but just learned how to, like, use macros that it would be, like,
37 00:05:20.410 ⇒ 00:05:24.500 Christina Knudson: Cleaner and, like, not repetitive, more reliable.
38 00:05:26.120 ⇒ 00:05:28.239 Demilade Agboola: Okay, that’s fair.
39 00:05:29.400 ⇒ 00:05:37.860 Demilade Agboola: Those are, like, very, like, fundamental parts of, like, building out any system, just the idea of, like, making it repeatable, and just also…
40 00:05:38.420 ⇒ 00:05:44.680 Demilade Agboola: Documentation and having to navigate the lack of clarity with Regard’s building out different things.
41 00:05:46.020 ⇒ 00:05:51.379 Demilade Agboola: literally had a meeting, like, 30 minutes ago, where we’re trying to figure out… so, a stakeholder…
42 00:05:51.540 ⇒ 00:05:55.520 Demilade Agboola: who’s handling their sales force, and, like, there’s no clarity on how things are…
43 00:05:56.660 ⇒ 00:06:05.949 Demilade Agboola: properly documented, and we’re having arguments about, like… well, not arguments, but we’re trying to go over, like, ARR, and, like, how the revenue is properly defined.
44 00:06:06.310 ⇒ 00:06:08.840 Demilade Agboola: So, it’s… it’s… it’s…
45 00:06:09.200 ⇒ 00:06:09.690 Christina Knudson: Yeah.
46 00:06:10.640 ⇒ 00:06:14.030 Demilade Agboola: In terms of, like, speaking of revenue, like.
47 00:06:14.350 ⇒ 00:06:16.790 Demilade Agboola: So let’s just try, like, a scenario.
48 00:06:17.320 ⇒ 00:06:23.149 Demilade Agboola: And you could take a minute or two to, like, before you answer, so that’s not… it’s not, like, an immediate response thing.
49 00:06:25.170 ⇒ 00:06:31.220 Demilade Agboola: So let’s say we have a client that wants to have, like, a daily revenue reporting model, or MART.
50 00:06:31.860 ⇒ 00:06:37.090 Demilade Agboola: And they have, like, 3 data sources, so say Stripe, Salesforce.
51 00:06:38.320 ⇒ 00:06:40.370 Demilade Agboola: Like, maybe Google Ads.
52 00:06:41.390 ⇒ 00:06:46.859 Demilade Agboola: And they want to have that built out in, like, you know, whatever cloud infrastructure you want.
53 00:06:47.050 ⇒ 00:06:55.110 Demilade Agboola: How would you design your solution, and what are the considerations you would, like, fact… that would factor into what decisions you make?
54 00:06:56.240 ⇒ 00:06:58.470 Christina Knudson: Okay. So I’m gonna, like…
55 00:06:58.940 ⇒ 00:07:08.920 Christina Knudson: type some notes to, like, organize this for myself. So you said that, you’d be getting revenue from 3 sources, Stripe, and then…
56 00:07:09.930 ⇒ 00:07:10.350 Demilade Agboola: goes for.
57 00:07:10.350 ⇒ 00:07:12.109 Christina Knudson: Oh yeah, Salesforce.
58 00:07:13.580 ⇒ 00:07:14.350 Christina Knudson: Okay.
59 00:07:14.690 ⇒ 00:07:22.470 Demilade Agboola: So… It’ll just be, like, a revenue reporting mat, where, you know, typical stuff about…
60 00:07:22.660 ⇒ 00:07:32.649 Demilade Agboola: the… what’s it called? Like, they want to see their revenue, they want to see, like, conversion rates, they want to see what ads were most effective, whatever. My…
61 00:07:33.240 ⇒ 00:07:39.849 Demilade Agboola: I want to be able to, like, like, how would you just design the solution? What technical considerations would you give?
62 00:07:40.340 ⇒ 00:07:47.549 Demilade Agboola: And, like… What would you try to, you know… Accommodating the entire infrastructure.
63 00:07:48.080 ⇒ 00:07:53.509 Demilade Agboola: So, like, there are a number of ways you could calculate, I just want to see how your mind, like, problem-solves these things.
64 00:07:54.490 ⇒ 00:08:08.720 Christina Knudson: Yeah, so I guess one thing I’m thinking about is, like, how much would these different sources be providing, like, duplicate records? Like, if it’s in Stripe, would it also be in Salesforce?
65 00:08:10.790 ⇒ 00:08:13.069 Demilade Agboola: Sorry, I didn’t have that question, I had a cough. Sorry.
66 00:08:13.070 ⇒ 00:08:22.219 Christina Knudson: Oh, okay. I was saying that, I was asking if the, like, Stripe and Salesforce would be providing, like, duplicate record…
67 00:08:22.480 ⇒ 00:08:27.060 Christina Knudson: records, or if it’s, like, if it’s in Stripe, then it’s not in Salesforce.
68 00:08:27.780 ⇒ 00:08:36.280 Demilade Agboola: Good question. So let’s say in this scenario, It does… Provide the same value.
69 00:08:36.590 ⇒ 00:08:41.309 Demilade Agboola: But Stripe allows us no payment date. So, it’s the same…
70 00:08:42.210 ⇒ 00:08:49.650 Demilade Agboola: like, Salesforce lets us know, like, you know, they’ve agreed to this, not the contract, but Stripe lets us know when they are actually paying, so…
71 00:08:50.210 ⇒ 00:08:51.449 Demilade Agboola: Same data, but…
72 00:08:52.420 ⇒ 00:09:02.490 Christina Knudson: Okay. Okay, so Stripe give us payment dates, Salesforce gives us contract info, Okay. And then…
73 00:09:04.470 ⇒ 00:09:07.300 Christina Knudson: Okay, so what I’m thinking about is, like.
74 00:09:07.430 ⇒ 00:09:13.110 Christina Knudson: the grain of the data, because when it’s coming in from, like, Stripe and Salesforce, then it’s probably, like.
75 00:09:13.240 ⇒ 00:09:20.370 Christina Knudson: well, like, Stripe would be, like, one record per payment, I would guess, and then, like, Salesforce is also, like, probably at the, like.
76 00:09:20.900 ⇒ 00:09:31.670 Christina Knudson: contract, or maybe, like, account level, so we’d need to think about, like, levels of granularity there, and if they, like, already match up, or if they need to, like.
77 00:09:34.340 ⇒ 00:09:42.490 Christina Knudson: Yeah, like, if you need to go into, like, one account and then, like, split out, like, different, like, payments,
78 00:09:44.490 ⇒ 00:09:49.950 Christina Knudson: And then, like, further down the line at some point. It sounds like we want to have it, like.
79 00:09:50.470 ⇒ 00:09:53.370 Christina Knudson: be daily revenue, and so then, like.
80 00:09:53.570 ⇒ 00:09:59.560 Christina Knudson: Further down, then you’d want to, like, aggregate by day to get, like…
81 00:09:59.780 ⇒ 00:10:04.990 Christina Knudson: The daily revenue, and then, like, for conversion, We’d probably want to, like.
82 00:10:05.730 ⇒ 00:10:10.430 Christina Knudson: well, not probably. We would need to, like, have some kind of… like…
83 00:10:11.010 ⇒ 00:10:16.890 Christina Knudson: grouping or window that we would, like, need to aggregate over, because obviously, like, we only have, like.
84 00:10:17.160 ⇒ 00:10:22.549 Christina Knudson: Did they pay, or not, or, like, whatever? So we’d want to know,
85 00:10:22.670 ⇒ 00:10:26.890 Christina Knudson: Like, do we want it, like, conversion by day, or, like, conversion by, like.
86 00:10:27.180 ⇒ 00:10:31.219 Christina Knudson: channel or conversion by whatever, and maybe it’s, like.
87 00:10:31.680 ⇒ 00:10:37.830 Christina Knudson: you know, a few of those, but the, like, grain is definitely something to think about there.
88 00:10:38.190 ⇒ 00:10:41.450 Christina Knudson: And… with the like.
89 00:10:43.700 ⇒ 00:10:47.959 Christina Knudson: kind of similar but different info in the different systems, then we’d want to, like.
90 00:10:48.200 ⇒ 00:10:58.239 Christina Knudson: think about with that granularity, like, how are we joining? How are we, like, deduping? Is there a risk of, like.
91 00:10:59.030 ⇒ 00:11:08.490 Christina Knudson: I don’t know, like, a one-to-many join, where really it should just be, like, a one-to-one join, and so, like, what kind of, like, keys would we need to…
92 00:11:08.810 ⇒ 00:11:20.339 Christina Knudson: join on in order to, like, make sure that we are getting, like, that one-to-one join, and, like, what kind of filtering do we need to do to make sure that we’re not, like, you know, accidentally including, like.
93 00:11:20.590 ⇒ 00:11:27.250 Christina Knudson: Contracts that not… that were never paid on, or, like, you know, something like that, so that we’re not, like, joining in things, and then, like.
94 00:11:27.380 ⇒ 00:11:31.409 Christina Knudson: Later having to dedupe when it would just make sense to just…
95 00:11:31.800 ⇒ 00:11:37.519 Christina Knudson: not pull those in in the first place, like the Salesforce contracts, for example.
96 00:11:37.740 ⇒ 00:11:44.510 Christina Knudson: Yeah, so then… What,
97 00:11:46.350 ⇒ 00:11:54.420 Christina Knudson: Yeah, how would I design the solution? So then, I would think about… since…
98 00:11:55.000 ⇒ 00:11:59.379 Christina Knudson: Stripe is processing all the payments, is that right, or is it just some of the payments?
99 00:11:59.800 ⇒ 00:12:08.619 Demilade Agboola: oh, let’s just assume, so that we don’t get too, like, convoluted that, oh, Stripe is doing some, you have some be hyper-line, yeah, I think that’ll be a bit too much.
100 00:12:09.030 ⇒ 00:12:09.380 Christina Knudson: Okay.
101 00:12:09.380 ⇒ 00:12:13.040 Demilade Agboola: Again, we’re trying to have just a simple scenario, and I just would like to…
102 00:12:13.660 ⇒ 00:12:19.730 Demilade Agboola: just to make it as simple as possible, but, like, good technical fundamentals, basically.
103 00:12:20.070 ⇒ 00:12:27.890 Christina Knudson: Okay, yeah. So then in that case, it would make sense to kind of make the Stripe data the, like, center of it, and, like…
104 00:12:28.160 ⇒ 00:12:29.769 Christina Knudson: Kinda have that…
105 00:12:29.900 ⇒ 00:12:41.610 Christina Knudson: I don’t know if, like, spine is the right word, but kind of, like, I know the left join would be on the left, like, the Stripe data would be on the left, and then, like, using Salesforce and Google Ads to, like.
106 00:12:41.680 ⇒ 00:12:57.529 Christina Knudson: add supplementary info, basically, so that we could see, like, okay, so here’s all the Stripe payments, and here’s the contract that is associated with it, and here is the, like, Google ad that, you know, led them to convert, things like that.
107 00:12:57.810 ⇒ 00:13:17.660 Christina Knudson: So yeah, you’d want to, like, definitely make sure that you understand, like, what are all the join keys to figure out how to identify, like, the correct Salesforce contract for that Stripe payment, and the, like, the… if there is one, like, the correct, like, Google ad for that Stripe payment, and maybe…
108 00:13:18.690 ⇒ 00:13:20.429 Christina Knudson: I’m not sure if it’s, like.
109 00:13:21.330 ⇒ 00:13:28.720 Christina Knudson: the Google ad leads to the contract, which then leads to the payment, or if you can, like.
110 00:13:28.920 ⇒ 00:13:34.070 Christina Knudson: directly say, like, this Google ad is connected to this Tripe payment, like.
111 00:13:34.300 ⇒ 00:13:36.840 Christina Knudson: My gut kind of feels like it would be, like.
112 00:13:37.210 ⇒ 00:13:44.180 Christina Knudson: the way that we connect Google Ads and Stripe would be, like, through Salesforce, but… okay, yeah.
113 00:13:44.180 ⇒ 00:13:49.789 Demilade Agboola: I’ll… yeah, so the good… idea would be… so if you have someone on a contract.
114 00:13:50.010 ⇒ 00:13:57.340 Demilade Agboola: They’re paying multiple times, and… The idea would be trying to… Figure out the lifetime value.
115 00:13:57.730 ⇒ 00:14:04.059 Demilade Agboola: So you can use that to figure out lifetime value, because if this ad lets them get in the contract, which has led to them
116 00:14:04.670 ⇒ 00:14:08.159 Demilade Agboola: paying for… Like, 10 months in a row.
117 00:14:08.790 ⇒ 00:14:14.249 Demilade Agboola: Or 20 months in a row, we can kind of figure out the lifetime value of that ad.
118 00:14:14.700 ⇒ 00:14:21.680 Demilade Agboola: That has led that customer to that contractor. That would be how you would want to think in terms of tying everything together.
119 00:14:22.320 ⇒ 00:14:23.950 Christina Knudson: Yeah, that sounds good.
120 00:14:24.110 ⇒ 00:14:30.599 Christina Knudson: And then you also, like, mentioned conversion rates, and so then one thing I would think about is, like.
121 00:14:30.840 ⇒ 00:14:41.479 Christina Knudson: what exactly is the conversion event? So, like, is the conversion event, like, having the contract in Salesforce and, like, the promise to pay, or is it, like…
122 00:14:41.940 ⇒ 00:14:50.940 Christina Knudson: it’s only truly considered a conversion once they, like, submit their first payment, or, like, I don’t know, something like that. I know sometimes, like, different
123 00:14:52.750 ⇒ 00:14:56.750 Christina Knudson: different, like… People can be different in, like, different…
124 00:14:56.850 ⇒ 00:15:09.090 Christina Knudson: can be interested in different, like, stages of the conversion, so, I would probably want to dig in more about, like, what are we thinking for conversion rates, like, as the definition of conversion.
125 00:15:10.160 ⇒ 00:15:10.550 Demilade Agboola: Okay.
126 00:15:11.000 ⇒ 00:15:11.880 Christina Knudson: Yeah.
127 00:15:13.860 ⇒ 00:15:18.949 Demilade Agboola: So, let’s, like, build on what we’ve done so far.
128 00:15:19.960 ⇒ 00:15:22.500 Demilade Agboola: So, I think the first question will probably be…
129 00:15:24.010 ⇒ 00:15:26.930 Demilade Agboola: How would you go about modeling the mods?
130 00:15:27.530 ⇒ 00:15:30.470 Demilade Agboola: Would you do more of a star schema?
131 00:15:30.730 ⇒ 00:15:37.529 Demilade Agboola: Normalized schema… How would, like, and what, like, how would you choose what to use?
132 00:15:38.300 ⇒ 00:15:44.609 Demilade Agboola: And… yeah, basically, how would you choose what to use? What guides your decision process on whether you go more star or normalized?
133 00:15:45.380 ⇒ 00:15:59.250 Christina Knudson: Yeah, so definitely, I’d want to, like, talk to the main stakeholders to see, like, how are they going to use this, because that’s going to, like, determine a lot about, like, how I would want to present this.
134 00:15:59.400 ⇒ 00:16:05.610 Christina Knudson: Like, in the mart, but, like…
135 00:16:06.630 ⇒ 00:16:11.229 Christina Knudson: I don’t know, mostly I’ve used, like, more of, like, the star schema, so…
136 00:16:11.510 ⇒ 00:16:15.320 Christina Knudson: I think that’s probably what I would default to, but,
137 00:16:17.060 ⇒ 00:16:26.089 Christina Knudson: I mean, I’m always learning new things, so if there’s, you know, a better way to do something, then I’m happy to do it that way.
138 00:16:26.500 ⇒ 00:16:27.340 Christina Knudson: Yeah.
139 00:16:28.570 ⇒ 00:16:36.930 Demilade Agboola: Okay, that’s fair. I think another follow-up question would be, so we built all of this out.
140 00:16:37.530 ⇒ 00:16:39.899 Demilade Agboola: And then realized that over time…
141 00:16:40.330 ⇒ 00:16:44.470 Demilade Agboola: One of our dbt models is, like, getting really slow.
142 00:16:45.100 ⇒ 00:16:45.760 Christina Knudson: Hmm.
143 00:16:46.080 ⇒ 00:16:47.000 Demilade Agboola: how…
144 00:16:49.780 ⇒ 00:16:56.570 Demilade Agboola: like, thinking of the worst case scenario, right? Like, so take, like, now this wasn’t something you built, someone else built it.
145 00:16:56.840 ⇒ 00:16:58.710 Demilade Agboola: What would you be looking for?
146 00:16:58.850 ⇒ 00:17:03.839 Demilade Agboola: To debugging and, like, making it run faster, like, improving efficiency of the model.
147 00:17:04.510 ⇒ 00:17:11.849 Christina Knudson: Yeah, so the first thing that I thought of was, like, what is the model materialization? Because it probably makes sense to
148 00:17:12.310 ⇒ 00:17:15.700 Christina Knudson: like… Set this up.
149 00:17:16.010 ⇒ 00:17:20.420 Christina Knudson: as, like, an incremental model. Like, obviously when you’re building, then…
150 00:17:21.140 ⇒ 00:17:32.529 Christina Knudson: you’re changing things a lot, and it may not make sense to make it incremental right away, but, like, once you feel like it’s, like, doing mostly the right thing, then I would want to, like, make it an incremental model, because, yeah, I expect, like.
151 00:17:32.870 ⇒ 00:17:38.780 Christina Knudson: A lot of payments to be coming through, and so it doesn’t make sense to be, like, refreshing these payments from, like.
152 00:17:38.990 ⇒ 00:17:55.140 Christina Knudson: 5 years ago, when those are probably pretty, like, set in stone, so it probably makes sense to just, like, set it up as an incremental model if it’s not already, and that should already make it, like, a lot faster, and especially since you said, like, over time it’s become
153 00:17:55.170 ⇒ 00:17:59.059 Christina Knudson: Slow, then that’s what’s making me think, like, oh, well, like.
154 00:17:59.480 ⇒ 00:18:07.610 Christina Knudson: Back when it was in a new company and had, like, 5 sales, then it was, like, this table wasn’t slow, but now that it’s been, like, you know…
155 00:18:07.740 ⇒ 00:18:17.889 Christina Knudson: long enough that they have a good number of, like, payments, and maybe it’s, like, the incremental piece that’s making it slow. And then if it’s not the incremental piece, then…
156 00:18:18.060 ⇒ 00:18:25.229 Christina Knudson: like, I’d want to look upstream, like, if it’s already set up as an incremental model, I mean.
157 00:18:25.400 ⇒ 00:18:26.970 Christina Knudson: Than I’d want to look like.
158 00:18:27.880 ⇒ 00:18:30.560 Christina Knudson: Upstream of what is…
159 00:18:31.050 ⇒ 00:18:43.920 Christina Knudson: like, what models are upstream of this, like, revenue mart, and see, like, what are the runtimes of those? Like, is it just something up there that is slow, and then that… even though, like, this
160 00:18:44.170 ⇒ 00:18:55.759 Christina Knudson: one mart is actually, like, doing its, like, little job pretty fast. If everything else upstream of it is slow, then that would make this look slow. So yeah, looking, like, upstream.
161 00:18:58.300 ⇒ 00:19:00.730 Demilade Agboola: And if you are…
162 00:19:01.040 ⇒ 00:19:05.690 Demilade Agboola: So let’s just add a bit more detail. So let’s say the query now is, like, 400 million rows.
163 00:19:05.900 ⇒ 00:19:07.549 Demilade Agboola: Which, I mean, obviously.
164 00:19:07.960 ⇒ 00:19:13.219 Demilade Agboola: The incremental suggestion will help quite a bit, because you don’t want to be refreshing 400 meals every day.
165 00:19:17.480 ⇒ 00:19:25.129 Demilade Agboola: I’m curious as to if there are any other things you think could be helpful, in terms of model efficiency.
166 00:19:26.210 ⇒ 00:19:27.660 Christina Knudson: Yeah, so, like.
167 00:19:27.940 ⇒ 00:19:28.280 Demilade Agboola: Yeah.
168 00:19:28.280 ⇒ 00:19:31.710 Christina Knudson: With the model itself, then, like.
169 00:19:32.540 ⇒ 00:19:52.370 Christina Knudson: I’d wanna see… like, if it definitely looks like this is the model that’s running slowly, then wanna… I’d wanna, like, look at the CTEs and see, like, okay, where is, like, filtering happening, or, like, are we pulling in things that we don’t actually need to be pulling in? Can we add some, like.
170 00:19:52.600 ⇒ 00:19:55.110 Christina Knudson: Basic things? Or, like.
171 00:19:55.700 ⇒ 00:20:05.179 Christina Knudson: maybe not basic, but, like, can we add things to the join so that it’s not needing to, like, then dedupe a whole bunch of things later?
172 00:20:07.120 ⇒ 00:20:10.960 Christina Knudson: Yeah, basically try to, like, Streamline it.
173 00:20:11.440 ⇒ 00:20:16.549 Christina Knudson: As early as possible, and as much as possible, to see if we can, like, make it
174 00:20:17.010 ⇒ 00:20:18.649 Christina Knudson: Faster that way.
175 00:20:19.430 ⇒ 00:20:20.890 Demilade Agboola: Okay, that’s fair.
176 00:20:21.500 ⇒ 00:20:28.380 Demilade Agboola: I have one more question. So, let’s say…
177 00:20:28.640 ⇒ 00:20:32.640 Demilade Agboola: So this is not more, like, the same scenario, this is, like, an entirely new scenario.
178 00:20:32.760 ⇒ 00:20:35.949 Demilade Agboola: So, because, you know, like, we’re a consulting firm.
179 00:20:36.940 ⇒ 00:20:41.830 Demilade Agboola: A client comes to you and says the… Need a dashboard.
180 00:20:42.260 ⇒ 00:20:46.109 Demilade Agboola: But they’re not necessarily, like, clear on what metrics they need, or how they need it.
181 00:20:46.380 ⇒ 00:20:52.180 Demilade Agboola: How do you go… from… Just this vague request to…
182 00:20:52.460 ⇒ 00:20:57.180 Demilade Agboola: Getting very good context on what you need to do in terms of building your models.
183 00:20:58.090 ⇒ 00:21:01.260 Demilade Agboola: and… That’s question one.
184 00:21:01.390 ⇒ 00:21:03.440 Demilade Agboola: Question two, stuff of that is.
185 00:21:03.610 ⇒ 00:21:09.389 Demilade Agboola: So you do tell them what you need to do, and then they give you, like, a really close deadline.
186 00:21:09.510 ⇒ 00:21:13.600 Demilade Agboola: So let’s say they need it in… They need it tomorrow, whatever.
187 00:21:14.130 ⇒ 00:21:21.500 Demilade Agboola: You know you could probably get it done by tomorrow, but you will take technical shortcuts that are not the most effective.
188 00:21:22.010 ⇒ 00:21:30.949 Demilade Agboola: How do you manage that scenario of balancing a stakeholder’s request with technical excellence.
189 00:21:31.760 ⇒ 00:21:32.390 Christina Knudson: Yeah.
190 00:21:32.900 ⇒ 00:21:43.659 Christina Knudson: Okay, so the first part then, like, how do you figure out what to put in the dashboard? Like, I would ask them, like, what kind of questions are they hoping to…
191 00:21:44.070 ⇒ 00:21:50.900 Christina Knudson: have answered when they look at the dashboard, like, what… kind of… Like…
192 00:21:50.900 ⇒ 00:21:52.100 Demilade Agboola: Sorry, sorry to interrupt.
193 00:21:52.310 ⇒ 00:21:56.539 Demilade Agboola: I’m still here, I’m just 10 seconds, I need to plug in my computer, it’s…
194 00:21:56.540 ⇒ 00:22:11.999 Christina Knudson: Okay, sounds good. Okay. Yeah, so I would, like, ask them, like, okay, what are you hoping that the dashboard is going to tell you? So then, from that, like, once I have that list of questions, then I can start to figure out, like, what to put in it, and, like.
195 00:22:12.220 ⇒ 00:22:16.540 Christina Knudson: I know that there’s a lot of questions that you don’t actually… want to ask
196 00:22:17.410 ⇒ 00:22:27.490 Christina Knudson: the stakeholder, but you’re, like, asking yourself in your head, like, okay, what kind of… like, to try to figure out, like, what’s the level of, like, aggregation, and, like.
197 00:22:27.710 ⇒ 00:22:31.800 Christina Knudson: Does it help to have it more as, like,
198 00:22:32.330 ⇒ 00:22:35.980 Christina Knudson: Like, plots, or more as, like…
199 00:22:36.010 ⇒ 00:22:47.599 Christina Knudson: Like, tables that they can have those numbers versus, like, kind of the more, like, overview, like, big picture that, like, a literal, like, picture or plot would give you.
200 00:22:47.600 ⇒ 00:22:57.129 Christina Knudson: So yeah, I’d want to know, like, what kind of things do they put in the dashboard, or what kind of things do they want to, like, get out of the dashboard? Try to figure out, like.
201 00:22:57.750 ⇒ 00:23:02.910 Christina Knudson: What also they’re, like, the level of… like.
202 00:23:03.980 ⇒ 00:23:14.609 Christina Knudson: quantitative comfort is. Like, if it’s somebody who’s, like, not very comfortable with, like, quantitative things, like, probably want to keep it,
203 00:23:15.090 ⇒ 00:23:17.849 Christina Knudson: Like, a little bit simpler, and, like.
204 00:23:18.830 ⇒ 00:23:30.049 Christina Knudson: just, like, more manageable for them, because if it’s, like, overwhelming, then they’re not going to want to use it, so try to, like, gear it towards whoever the, like, consumer of that is going to be.
205 00:23:30.340 ⇒ 00:23:34.309 Christina Knudson: And yeah, so basically, like, that process would involve, like.
206 00:23:35.000 ⇒ 00:23:43.649 Christina Knudson: a lot of, like, back and forth asking questions, and, like, trying to understand more about, like, what they’re hoping for,
207 00:23:43.930 ⇒ 00:23:52.219 Christina Knudson: Even though… like… They don’t know yet, but, like, try to, like, figure it out together.
208 00:23:52.470 ⇒ 00:23:57.809 Christina Knudson: And then if the, like, client gives a close deadline, then I guess I would, like.
209 00:23:58.060 ⇒ 00:24:03.430 Christina Knudson: lay out the options, like, okay, so if you want it done tomorrow, then that means that, like.
210 00:24:04.310 ⇒ 00:24:15.040 Christina Knudson: this is the scope of what can happen for tomorrow. Like, maybe you can’t have this, this, this, this, but, like, sounds like your top burning question right now is.
211 00:24:15.140 ⇒ 00:24:20.780 Christina Knudson: Like, this one question, so, like, for tomorrow, we could, like, have that…
212 00:24:21.000 ⇒ 00:24:29.229 Christina Knudson: Available, and then, like, this would be, like, like, my recommendation for, like, a timeline for, like, the other pieces of it.
213 00:24:29.630 ⇒ 00:24:37.519 Christina Knudson: And… with the, like… Technical shortcuts,
214 00:24:40.890 ⇒ 00:24:44.730 Christina Knudson: that is, like, a little bit trickier, because I think that a lot of stakeholders…
215 00:24:45.450 ⇒ 00:24:50.209 Christina Knudson: Don’t really understand what those shortcuts mean, and so…
216 00:24:50.370 ⇒ 00:24:52.329 Christina Knudson: I think it’s important to, like.
217 00:24:52.600 ⇒ 00:25:00.560 Christina Knudson: not just say, like, oh, we’ll have to, like, take technical shortcuts, but, like, kind of lay out some of those risks. Like, what that means is that, like.
218 00:25:00.970 ⇒ 00:25:10.559 Christina Knudson: this… May… be, like… we may… I don’t know.
219 00:25:10.680 ⇒ 00:25:17.559 Christina Knudson: be off by, like, 10% or something like that, because we think that, like.
220 00:25:17.900 ⇒ 00:25:20.670 Christina Knudson: If we don’t have time to, like, properly…
221 00:25:21.040 ⇒ 00:25:33.100 Christina Knudson: I mean, I wouldn’t say, like, all of this to them, but, like, to you. Like, okay, if we don’t have time to, like, properly, like, dedupe it and, like, check everything, then that means that, like, we might accidentally be, like.
222 00:25:34.260 ⇒ 00:25:40.149 Christina Knudson: Doing, like, a many-to-one join on whatever thing for, like, mostly for…
223 00:25:40.440 ⇒ 00:25:44.599 Christina Knudson: like, this one channel or something is, like, the biggest risk.
224 00:25:44.850 ⇒ 00:25:49.020 Christina Knudson: So, try to, like, explain to them, like, how much…
225 00:25:50.120 ⇒ 00:25:52.959 Christina Knudson: It could be off, or, like, what is the actual, like.
226 00:25:53.170 ⇒ 00:25:59.590 Christina Knudson: risk to them that they would see, since they don’t really know what’s, like, going on under the hood, and they don’t want to know what’s going…
227 00:25:59.690 ⇒ 00:26:01.409 Christina Knudson: They don’t want to know what’s good.
228 00:26:02.470 ⇒ 00:26:09.900 Christina Knudson: You know what I’m trying to say, yeah. I’m like, why can’t I say this? Yeah, yeah, so just, like, try to, like, keep it…
229 00:26:10.260 ⇒ 00:26:17.409 Christina Knudson: To what actually, like, matters to them, so that they can understand, like, what are the risks to them.
230 00:26:17.920 ⇒ 00:26:29.359 Demilade Agboola: That’s fair, that’s fair. So that was my final question. I was curious if you had any other questions, like, for me, about, like, Brainforge, or working in Brainforge, or anything along those lines.
231 00:26:30.270 ⇒ 00:26:35.940 Christina Knudson: Yeah, so I was wondering, like, if you could tell me what you like about working at Brainforge.
232 00:26:37.710 ⇒ 00:26:40.470 Demilade Agboola: So me, personally?
233 00:26:40.690 ⇒ 00:26:41.040 Christina Knudson: Yeah.
234 00:26:41.040 ⇒ 00:26:44.439 Demilade Agboola: I like to solve problems, I’m that kind of person.
235 00:26:46.260 ⇒ 00:26:52.340 Demilade Agboola: Bringford gives me that, like, feeling of being in different industries and, like, having to…
236 00:26:53.110 ⇒ 00:26:58.400 Demilade Agboola: figure out and solve the problems, because for me, one of the best things is whenever, like, I get to the bottom of…
237 00:26:59.480 ⇒ 00:27:05.309 Demilade Agboola: whatever problem it is, I get there’s a… there’s, like, a release of, like, endorphins. I’m just like, yes!
238 00:27:05.430 ⇒ 00:27:08.679 Demilade Agboola: So for me, Brantford gives me the opportunity on the daily.
239 00:27:09.370 ⇒ 00:27:13.180 Demilade Agboola: But also beyond that, there’s also the,
240 00:27:13.970 ⇒ 00:27:17.750 Demilade Agboola: Clarity, lack of micromanagement a lot of times.
241 00:27:18.300 ⇒ 00:27:25.879 Demilade Agboola: you know, we’re across different time zones, I’m in Malta, sometimes in Texas, we have colleagues in Pakistan, you know, so, like.
242 00:27:26.470 ⇒ 00:27:41.139 Demilade Agboola: no one’s micromanaging. We believe in strong communication and strong clarity, like, hey, today I’m going to be closing early, but I will be up tomorrow to, like, get this done before the client meeting. Things like that, like, we’re just…
243 00:27:41.300 ⇒ 00:27:46.100 Demilade Agboola: Claring off in what we’re doing and how things are working the different…
244 00:27:46.430 ⇒ 00:27:52.240 Demilade Agboola: across the team, such that we don’t have clashes, so I really do like that part as well, of working in Brainforge.
245 00:27:53.040 ⇒ 00:27:55.920 Christina Knudson: Yeah, okay, cool. And then, like.
246 00:27:57.470 ⇒ 00:28:05.609 Christina Knudson: When things get hard on the team, can you describe, like, how the team handles that?
247 00:28:06.760 ⇒ 00:28:16.289 Demilade Agboola: I think it depends on what a hard is, but generally speaking, It’s usually a pain where…
248 00:28:16.510 ⇒ 00:28:20.050 Demilade Agboola: We will have, like, a huddle or a Zoom call where we can all hop on.
249 00:28:20.370 ⇒ 00:28:26.190 Demilade Agboola: We all try to gain clarity on what the scenario is. So, is it a thing of, like…
250 00:28:27.440 ⇒ 00:28:30.450 Demilade Agboola: We got bad data into a dashboard.
251 00:28:30.990 ⇒ 00:28:34.989 Demilade Agboola: And now the client is questioning so much.
252 00:28:36.410 ⇒ 00:28:41.870 Demilade Agboola: And so, again, we will… let’s use that as a scenario. We will ask, like, okay, so what happened?
253 00:28:42.220 ⇒ 00:28:51.399 Demilade Agboola: And someone might be like, hey, we got… maybe we didn’t get client instructions from the client, so that factored into the modeling. It could be…
254 00:28:51.620 ⇒ 00:28:59.689 Demilade Agboola: a thing where internally, we didn’t QA as best as we could, and so that got the bad data in there.
255 00:28:59.950 ⇒ 00:29:06.289 Demilade Agboola: But usually we try to figure out what the root cause was. If it’s an internal thing, we might think of, like, how do we…
256 00:29:06.490 ⇒ 00:29:11.269 Demilade Agboola: Number one, fix the problem, like, immediately, but number two, also, like, build processes to…
257 00:29:12.130 ⇒ 00:29:14.750 Demilade Agboola: Prevent such things from happening.
258 00:29:15.140 ⇒ 00:29:28.449 Demilade Agboola: So for instance, that’s part of why, unless it’s really, really critical, we don’t really do PRs on Fridays, we try to avoid that, because we have had scenarios where, bad data was pushed, and Saturday, Sunday.
259 00:29:28.890 ⇒ 00:29:30.580 Demilade Agboola: The teams that were using it.
260 00:29:30.910 ⇒ 00:29:38.130 Demilade Agboola: were, like, the clients’ teams that were using it had bad data and were so, like, frustrated, and it wasn’t fixed until Monday. So, like, just things like that.
261 00:29:38.270 ⇒ 00:29:44.529 Demilade Agboola: If it’s a thing of, like, hey, we didn’t get clear requirements, and that affected our data quality and modeling quality.
262 00:29:44.700 ⇒ 00:29:50.360 Demilade Agboola: Okay, so what do we need to do to ensure that, like, we always know what we’re doing before we make that step?
263 00:29:50.760 ⇒ 00:29:54.880 Demilade Agboola: Yeah, so generally, those are kind of the processes.
264 00:29:55.650 ⇒ 00:29:59.469 Demilade Agboola: So we’ve tried and get, like, resolution, as well as just, like, future-proofing.
265 00:30:00.790 ⇒ 00:30:02.159 Christina Knudson: Okay, cool.
266 00:30:03.200 ⇒ 00:30:04.770 Christina Knudson: Can I ask one more question?
267 00:30:05.040 ⇒ 00:30:05.870 Demilade Agboola: Sure, go for it.
268 00:30:05.870 ⇒ 00:30:10.510 Christina Knudson: Okay, like… with…
269 00:30:10.810 ⇒ 00:30:22.719 Christina Knudson: anything data, or probably most things in life. You’re never, like, 100% done, but at some point, you probably need to stop doing things for the client, like…
270 00:30:23.110 ⇒ 00:30:27.610 Christina Knudson: There’s some, like, scope, and then that… Needs to just be, like.
271 00:30:27.740 ⇒ 00:30:35.070 Christina Knudson: you know, completed at some point, so, like, How… Like, how does that…
272 00:30:35.230 ⇒ 00:30:43.960 Christina Knudson: work? Like, how do you determine, like, when you’ve, like, fulfilled the, like, needs of it, and… like…
273 00:30:44.610 ⇒ 00:30:54.079 Christina Knudson: That you just have to accept that, like.01% of the data are gonna be doing this weird thing still, or, like, something like that.
274 00:30:55.880 ⇒ 00:30:58.529 Demilade Agboola: Hmm, that’s a really good question.
275 00:30:58.810 ⇒ 00:31:06.239 Demilade Agboola: I think it depends on… what we communicate to the client, if we can figure out what the…
276 00:31:06.590 ⇒ 00:31:07.940 Demilade Agboola: Discrepancies are.
277 00:31:08.280 ⇒ 00:31:13.509 Demilade Agboola: So we try to do our best to be able to isolate whatever discrepancies there are.
278 00:31:14.700 ⇒ 00:31:19.330 Demilade Agboola: get a root cause. Some things are beyond our control. So, for instance, if we can say, hey.
279 00:31:19.550 ⇒ 00:31:27.220 Demilade Agboola: Our data is dependent on this API. This API has these issues that do not match up to what you’ll see on the front end.
280 00:31:27.430 ⇒ 00:31:33.810 Demilade Agboola: And so, if you look on the front end, you might see these numbers, but you look on our dashboard, you might see different numbers.
281 00:31:34.060 ⇒ 00:31:34.650 Christina Knudson: D.
282 00:31:34.650 ⇒ 00:31:45.560 Demilade Agboola: beyond our control. This is just an API thing, and usually takes about 24 hours for the API to be up to date, or whatever, or the dashboard to be up to date with the API. And we can break it down.
283 00:31:45.780 ⇒ 00:31:50.399 Demilade Agboola: ensure that, hey, we’ve done the best of our abilities of what’s going on.
284 00:31:50.860 ⇒ 00:31:58.400 Demilade Agboola: I think usually that’s what it comes down to. We try our best, and even, like, part of the call, like I said, I had, like, 30 minutes prior to this call.
285 00:31:58.590 ⇒ 00:32:05.849 Demilade Agboola: was just being able to… I’m going to, like, go back and just kind of break down the different scenarios in which our numbers don’t seem to match.
286 00:32:06.110 ⇒ 00:32:11.219 Demilade Agboola: And come up with just a plan of attack from their end, as well as our end, on how that will go.
287 00:32:11.620 ⇒ 00:32:13.810 Demilade Agboola: And how we should handle things.
288 00:32:14.490 ⇒ 00:32:18.110 Demilade Agboola: But I think, ultimately, that’s it. I think it comes down to communication and…
289 00:32:18.320 ⇒ 00:32:25.049 Demilade Agboola: the ability to show expertise, I think, ultimately, we’re the experts, that’s why we’re hired, that’s why we’re being paid.
290 00:32:25.360 ⇒ 00:32:33.160 Demilade Agboola: And it’s important for us to be able to say, hey, we understand this is what you want to see, we understand this is what you desire, we understand this is what was scoped.
291 00:32:34.100 ⇒ 00:32:37.279 Demilade Agboola: Unfortunately, there are these restrictions.
292 00:32:37.670 ⇒ 00:32:40.879 Demilade Agboola: And as long as we can clearly state and…
293 00:32:41.450 ⇒ 00:32:49.270 Demilade Agboola: Ensure that within the entire conversation, we maintain that scope of, like, professionalism as well as expertise.
294 00:32:49.830 ⇒ 00:32:55.039 Demilade Agboola: a lot of clients are able to understand whatever that is, and in some cases, they might also push
295 00:32:55.220 ⇒ 00:33:00.229 Demilade Agboola: For, like, they might put pressure on the other, like, block that makes it hard for it to happen.
296 00:33:00.350 ⇒ 00:33:06.100 Demilade Agboola: Because they’re like, hey, if the issue’s the API, I’m gonna reach out to the vendor and, like, ensure that… Yeah. Yeah, things like that.
297 00:33:06.360 ⇒ 00:33:10.689 Demilade Agboola: So, ultimately, I think it’s just about being clear.
298 00:33:10.920 ⇒ 00:33:20.040 Demilade Agboola: I’m being transparent and saying, hey, we’re 90% of the way there. This last 10% is being, a hassle because of
299 00:33:20.210 ⇒ 00:33:21.380 Demilade Agboola: This, this.
300 00:33:21.910 ⇒ 00:33:22.610 Demilade Agboola: Bean.
301 00:33:22.890 ⇒ 00:33:25.509 Demilade Agboola: So we just try to maintain clarity, as well as just.
302 00:33:25.880 ⇒ 00:33:34.449 Demilade Agboola: within the team, so if someone else comes on the project and is, like, trying to figure out what’s going on and why it’s not matching, they’re not starting from zero. There’s some clarity in terms of either
303 00:33:34.570 ⇒ 00:33:40.510 Demilade Agboola: Documentation in the files, collected code files, or just in the general repo.
304 00:33:40.620 ⇒ 00:33:42.420 Demilade Agboola: So we can figure out what’s going on with that.
305 00:33:43.280 ⇒ 00:33:43.890 Christina Knudson: Okay.
306 00:33:44.670 ⇒ 00:33:46.199 Christina Knudson: Okay, cool, thank you.
307 00:33:46.480 ⇒ 00:33:50.279 Demilade Agboola: That sounds good. I see you’re based in, Minnesota?
308 00:33:50.560 ⇒ 00:33:51.110 Christina Knudson: Yeah.
309 00:33:51.370 ⇒ 00:33:55.540 Demilade Agboola: Okay. So my girlfriend lives in West St. Paul, so I do visit quite a.
310 00:33:55.540 ⇒ 00:33:56.150 Christina Knudson: Oh!
311 00:33:56.150 ⇒ 00:33:57.100 Demilade Agboola: Good. Yeah.
312 00:33:57.200 ⇒ 00:34:00.290 Demilade Agboola: Wow. And she lives there, so that’s…
313 00:34:00.480 ⇒ 00:34:06.610 Demilade Agboola: I… I literally was there, in December. I should be there in, like, 2 weeks’ time, 3 weeks’ time, so, yeah.
314 00:34:07.290 ⇒ 00:34:07.970 Christina Knudson: Wow.
315 00:34:08.210 ⇒ 00:34:10.879 Demilade Agboola: It’s always cold, though, that’s my issue with Minnesota.
316 00:34:10.880 ⇒ 00:34:16.270 Christina Knudson: Yeah, I’m like, why did you go in December, but also… You know?
317 00:34:16.610 ⇒ 00:34:23.940 Demilade Agboola: I mean, to be fair, it got really buckleshit, because she was telling me how bad it got in, like, January, February, so I was kind of happy I went in December, November.
318 00:34:23.940 ⇒ 00:34:37.100 Christina Knudson: Yeah. That’s true. December is definitely better than January or February. Like, January or February, it’s like, yeah, the cold starts to, like, really just sink into, like, your bones, and it’s pretty hard to shake, so…
319 00:34:37.230 ⇒ 00:34:39.580 Christina Knudson: Yeah, amazing.
320 00:34:39.580 ⇒ 00:34:40.770 Demilade Agboola: So now, with getting.
321 00:34:40.770 ⇒ 00:34:47.900 Christina Knudson: I know, it was, like, 70 degrees Celsius here. I mean, Fahrenheit. I don’t know what that is in Celsius, I can never…
322 00:34:48.150 ⇒ 00:34:52.319 Demilade Agboola: That’s probably, like, 2022, maybe 19, actually.
323 00:34:52.639 ⇒ 00:34:58.979 Christina Knudson: Okay, yeah, it was pretty nice yesterday, so hopefully she got to, like, get out and enjoy the sun a little bit.
324 00:34:58.980 ⇒ 00:35:03.670 Demilade Agboola: Oh, no, she’s taking her walks, she’s enjoying it, she’s having a great time now.
325 00:35:03.670 ⇒ 00:35:06.260 Christina Knudson: Okay. Nice.
326 00:35:06.410 ⇒ 00:35:10.369 Demilade Agboola: Okay, yeah, so I just, I just remembered that when I was looking through her profile, I was like, oh.
327 00:35:10.630 ⇒ 00:35:12.059 Demilade Agboola: Minnesota auction, interesting.
328 00:35:12.060 ⇒ 00:35:12.660 Christina Knudson: Yeah.
329 00:35:12.850 ⇒ 00:35:14.259 Christina Knudson: Yeah, that is cool.
330 00:35:14.260 ⇒ 00:35:15.100 Demilade Agboola: Alright, then.
331 00:35:15.600 ⇒ 00:35:17.210 Christina Knudson: Okay, well, thank you so much.
332 00:35:17.370 ⇒ 00:35:26.490 Demilade Agboola: I’ll… thanks for the time, I’ll definitely be reaching out to Kayla, and just letting her, know my briefing’s on the call, and she’ll definitely be in touch with you.
333 00:35:26.900 ⇒ 00:35:28.600 Christina Knudson: Okay, sounds good. Thank you.
334 00:35:28.850 ⇒ 00:35:29.620 Demilade Agboola: Bye.
335 00:35:29.860 ⇒ 00:35:30.610 Christina Knudson: Bye.