Meeting Title: Brainforge Interview w- Awaish Date: 2026-04-14 Meeting participants: Adegbite Ayoade, Awaish Kumar


WEBVTT

1 00:09:56.530 00:09:57.750 Awaish Kumar: Hello.

2 00:09:58.700 00:09:59.669 Adegbite Ayoade: Hi, Heather.

3 00:10:00.620 00:10:01.750 Awaish Kumar: Hey, how you doing?

4 00:10:02.270 00:10:03.380 Adegbite Ayoade: I’m good, and you?

5 00:10:07.760 00:10:08.480 Awaish Kumar: Okay.

6 00:10:08.480 00:10:09.369 Adegbite Ayoade: I’m doing fine.

7 00:10:10.000 00:10:11.320 Awaish Kumar: Can I pronounce your name?

8 00:10:12.820 00:10:16.779 Adegbite Ayoade: Are you? So… or just call me the officials.

9 00:10:18.300 00:10:19.110 Awaish Kumar: Sorry.

10 00:10:20.170 00:10:21.010 Adegbite Ayoade: It been…

11 00:10:21.210 00:10:22.220 Awaish Kumar: Evan, okay.

12 00:10:22.430 00:10:23.150 Adegbite Ayoade: Yeah.

13 00:10:25.240 00:10:26.879 Awaish Kumar: Yeah, we can,

14 00:10:28.960 00:10:35.500 Awaish Kumar: Okay, so in this session, we are just going to talk a little bit more about you, your background.

15 00:10:35.740 00:10:39.929 Awaish Kumar: And a little bit about brain foods and what we do here.

16 00:10:41.120 00:10:44.690 Awaish Kumar: So, yeah, in the beginning, we’ll just start with your introduction.

17 00:10:45.190 00:10:47.589 Awaish Kumar: And then some follow-up questions, and after that.

18 00:10:48.130 00:10:51.420 Awaish Kumar: I will leave some time for you to ask any questions at the end.

19 00:10:52.490 00:10:54.280 Adegbite Ayoade: Okay. Yeah, that’s fine.

20 00:10:54.900 00:10:56.949 Awaish Kumar: Let’s start with your introduction.

21 00:10:58.550 00:11:05.869 Adegbite Ayoade: Okay, so, I actually started as a web developer, and

22 00:11:06.010 00:11:24.970 Adegbite Ayoade: I had to resign from that web development job when I gained admission to, you know, further my education, and at that point, I was looking for what I can do while, you know, learning. I don’t want to, like, leave the old tech space just because I want to go back and further my education. So, while researching.

23 00:11:25.040 00:11:35.010 Adegbite Ayoade: I came across, data, data analytics, actually. So, after doing, like, self-study for, like, 3 to 6 months.

24 00:11:35.010 00:11:47.280 Adegbite Ayoade: I got my first training gig to teach somebody data analytics, you know, SQL, Power BI, Tableau, Excel and Python, right? And right after that.

25 00:11:47.710 00:12:00.219 Adegbite Ayoade: that person sort of now becomes, like, my number one client. They started telling people about me. So when he traveled out, he started referring, you know, students from UK, started doing master’s projects, you know.

26 00:12:00.260 00:12:17.559 Adegbite Ayoade: From there, I… after graduating, I now participated in an academy, and that account gave me my first, you know, company experience. So, I came second, and I was, given a job offer with this, company that,

27 00:12:17.560 00:12:21.599 Adegbite Ayoade: The mission is to make data available for,

28 00:12:22.080 00:12:38.700 Adegbite Ayoade: anybody on the street, so you can understand what is happening when you see a particular chart. So that role, I started as a research analyst, and, we sort of built a product. So after building that product and see how that product is, you know,

29 00:12:38.700 00:12:45.190 Adegbite Ayoade: is performing, and, you know, people are happy for that product. That whole thing just pushed me more.

30 00:12:45.400 00:12:59.479 Adegbite Ayoade: into analytics, where I now started building reports, answering business questions, and, you know, supporting decision making, basically. So from there, I moved into a role where I now literally became

31 00:12:59.690 00:13:01.310 Adegbite Ayoade: The only data person.

32 00:13:01.450 00:13:15.149 Adegbite Ayoade: In that team, and I had to build the entire infrastructure from ingestion down to, you know, transformation, to analytics, and then to building, dashboards. And at this point, I was already, you know, moving, like.

33 00:13:15.150 00:13:21.940 Adegbite Ayoade: into the data engineering part, so I use mostly open source to build, you know, all of this,

34 00:13:22.080 00:13:28.090 Adegbite Ayoade: Pipelines and, you know, sources from different places, so that the organization can have more

35 00:13:29.080 00:13:36.799 Adegbite Ayoade: a better way, basically, to make informed, decisions. So, from there, I started realizing that it’s…

36 00:13:38.220 00:13:55.189 Adegbite Ayoade: data is beyond just moving data from one place to another. It’s more useful when you can actually solve business problems. So I started moving towards analytics engineering, where I get to relate more with stakeholders. And right now, I work as a senior analytics engineer in

37 00:13:55.190 00:14:11.479 Adegbite Ayoade: consulting company where I’ve had the opportunity to work across different companies, solving different data problems, from documentation to data auditing, you know, to building models that save products, amongst many other things, right? So, at this point, I think,

38 00:14:11.480 00:14:28.070 Adegbite Ayoade: My strengths, basically sit between, communicating with stakeholders and getting what they need, and transforming that into a technical requirement that, you know, analysts and, data science or data, machine learning engineer cannot consume the products that I’m building.

39 00:14:29.710 00:14:30.500 Awaish Kumar: Okay.

40 00:14:31.060 00:14:33.669 Awaish Kumar: So, can you give me one of…

41 00:14:34.380 00:14:37.949 Awaish Kumar: Example of one… of the… one of the projects.

42 00:14:38.260 00:14:43.350 Awaish Kumar: That you did, end-to-end, and… basically…

43 00:14:43.860 00:14:49.499 Awaish Kumar: Yeah, just brief me about the whole project, what tools and technologies were used.

44 00:14:49.720 00:14:54.740 Awaish Kumar: What was your contribution, and what were… The final outcome.

45 00:14:55.980 00:15:05.380 Adegbite Ayoade: Okay, so, I think I will keep from, my current role, basically at Bridge LLC. So,

46 00:15:06.140 00:15:21.560 Adegbite Ayoade: we have the product inventory, you know, that is coming from different sources. Basically, it’s, S3-based ingestion, and the business rely on, you know, pricing, availability, and I think, reporting, right?

47 00:15:21.560 00:15:31.450 Adegbite Ayoade: But the issue was that when error occurs, right, which is quite common in a retail beta, the only way to fix that is probably just,

48 00:15:31.570 00:15:36.879 Adegbite Ayoade: backfilling the data all over again. And this basically sort of, like, create a bottleneck for…

49 00:15:36.890 00:15:52.050 Adegbite Ayoade: the analytics team, right? So, when I came in, we designed an ingestion system that basically just extracts, from S3 and loads directly into Snowflake, using staging tables, basically. So, from there.

50 00:15:52.050 00:16:04.919 Adegbite Ayoade: I connect dbt to now build, you know, modular models that is clean, that the analysts can use, directly. So, their whole entire pipeline. So, we were using dbt Core.

51 00:16:04.920 00:16:20.690 Adegbite Ayoade: So I had to use Airflow to sort of, like, build the orchestration workflow for that pipeline. So my contribution, basically, in that project would be that I led the design and implementation of, you know, the OFUSTAR from the…

52 00:16:20.690 00:16:34.699 Adegbite Ayoade: model structure in dbt to be in the snowflake ingestion, pattern, right? And, as an added advantage, I developed, you know, a streamlined application that would help the business, side

53 00:16:34.700 00:16:45.970 Adegbite Ayoade: for interaction, maybe they need to, quickly update some set of values without having to reload the entire ingestion pipeline again, so they can just use that app. So that app serves as,

54 00:16:46.360 00:17:05.010 Adegbite Ayoade: a modification layer, basically, for the pipeline, and we sort of implemented, like, a data validation structure into the field, so that when something fails, it doesn’t get to the analytics team before we actually flag that, you know, something has failed.

55 00:17:06.920 00:17:13.380 Awaish Kumar: Okay, so… How… How many years of experience do you have in DBT?

56 00:17:14.000 00:17:17.059 Adegbite Ayoade: I have, close to 6 years of experience in e-writing.

57 00:17:18.170 00:17:26.940 Awaish Kumar: Okay, and so, yeah, just tell me, like, If you think…

58 00:17:27.980 00:17:33.970 Awaish Kumar: If we have a table, which is really big, and data volume.

59 00:17:35.490 00:17:37.290 Awaish Kumar: Lots of columns, also.

60 00:17:37.820 00:17:44.090 Awaish Kumar: So… And it takes a lot of time to carry that table.

61 00:17:44.810 00:17:51.079 Awaish Kumar: So how would you go about… optimizing… The query performance.

62 00:17:52.310 00:18:04.930 Adegbite Ayoade: Okay, so, first, I think you can use, incremental approach for first loading that table, basically, so that, it’s not…

63 00:18:05.410 00:18:21.779 Adegbite Ayoade: reloading the old thing all over again, right? Then, after doing incremental, the second will be, selecting the actual columns that are needed. Because I’ve worked, on a project where I had to convert a stop procedure

64 00:18:21.780 00:18:30.779 Adegbite Ayoade: into a DBT model, right? And the stop procedure basically has close to a thousand columns. So what I did was, I had to look for

65 00:18:30.840 00:18:34.210 Adegbite Ayoade: The coulombs that were actually used in production.

66 00:18:34.270 00:18:39.839 Adegbite Ayoade: And then trace that back to, you know, to the stop procedure, and first pull that out.

67 00:18:39.840 00:18:54.320 Adegbite Ayoade: So, once I pull that out, I know that those columns serve as the final layer that I’m sending out to the, to the consumer or to the user. So, column pruning, right, which is going to mean that I’m looking at the exact columns that are actually needed

68 00:18:54.380 00:19:01.500 Adegbite Ayoade: and pull out those columns from the, you know, our numerous columns that we have. Then, I will try to look at,

69 00:19:01.830 00:19:22.089 Adegbite Ayoade: at what area are we using filter? What kind of transformations are we performing, right? Are these transformations, things that we can push a bit back, so that the final layer that the dashboards are connected to is not, you know, consuming too much power to aggregate any sort of value? Can we have, like, the aggregated data

70 00:19:22.090 00:19:28.979 Adegbite Ayoade: sent to the good layer, so that when the dashboards are feeding on it, the kind of workload that they are doing is not much.

71 00:19:28.980 00:19:39.780 Adegbite Ayoade: Then I also look at the kind of joinments that are there. Do we have, you know, a lot of subqueries within the query, and we turn all of this into a city so that

72 00:19:39.930 00:19:46.749 Adegbite Ayoade: When we are looking at which part of our code is taking too much time to, you know, to bring our results.

73 00:19:46.750 00:19:59.240 Adegbite Ayoade: We can know exactly the path to go into, because CTA will give us that. We’ll be able to, like, test and optimize every part of the code. But when we have a lump sum, long-line query that, you know, has different join-ins.

74 00:19:59.240 00:20:15.349 Adegbite Ayoade: has occurred within it, then it becomes a problem to actually figure out where the problem is coming from. So, I will look at it in that approach, right? Break it down, you know, take some things to, city approach, and try to avoid many-to-many journeys, right?

75 00:20:15.350 00:20:26.559 Adegbite Ayoade: Then, aside that, I would also… using that same, if this is just one model, right, I can break that into different layers, right? So I can have maybe

76 00:20:26.870 00:20:28.539 Adegbite Ayoade: an intermediate layer.

77 00:20:28.540 00:20:52.540 Adegbite Ayoade: that basically focus on transformation, right? And then I can have, like, a domain-focused mat that basically just serves what that particular unit is actually, you know, looking to consume, basically. And like I mentioned earlier as well, we can use incremental approach so that the table is not, loading all over again. Now, also, depending on the warehouse that we are using.

78 00:20:52.540 00:21:06.179 Adegbite Ayoade: We can set some, implementations around clustering and partitioning. So if we are using Snowflake, easier for us to, you know, set, clustering, to be filtered by, you know, a particular column, or columns.

79 00:21:06.340 00:21:13.470 Adegbite Ayoade: So that informations that are going out there are more optimized for, you know, the dashboard that is basically consuming it.

80 00:21:14.890 00:21:19.090 Awaish Kumar: Okay, so how… How can you apply partitioning in Snowflake?

81 00:21:20.330 00:21:26.149 Adegbite Ayoade: Okay, so applying partitioning in Snowflake. So let’s say, the…

82 00:21:27.370 00:21:35.680 Adegbite Ayoade: So, in your configuration, right, we have, I’m trying to get a mental model now,

83 00:21:37.900 00:21:44.710 Adegbite Ayoade: How do you… okay, so… Let’s say we have a model, right, that has,

84 00:21:46.820 00:21:54.130 Adegbite Ayoade: I don’t want to… sorry, okay. So, I think for Snowflake, actually, right, Snowflake has,

85 00:21:54.240 00:22:13.770 Adegbite Ayoade: partition enabled, right? I think so. It has partition enabled, basically, right? Its own partition is sort of, like, automatic, so it takes you to, like, split your data. Once it’s growing too much, it starts splitting it into, like, a macro, partitions, basically. Maybe… it could be by coulomb range, it could be by, you know.

86 00:22:14.240 00:22:18.899 Adegbite Ayoade: values, amongst many other things. But if it is,

87 00:22:19.790 00:22:27.790 Adegbite Ayoade: setting up, like, clustering, right? You can use, you know, your filter basically around where,

88 00:22:28.240 00:22:31.440 Adegbite Ayoade: Like, the rear line to sort of, like, set

89 00:22:31.840 00:22:38.460 Adegbite Ayoade: your… your clustering. So, it could be, cluster by maybe, created updates, or, you know.

90 00:22:38.460 00:22:52.429 Adegbite Ayoade: combination of user ID and created updates, or maybe a combination of last updated and user ID, or something like that. So that way, it helps you to, like, automatically implement that, on the data warehouse side.

91 00:22:53.240 00:22:55.549 Awaish Kumar: So, for example, if I have a table.

92 00:22:55.910 00:22:59.090 Awaish Kumar: If I have a model, in Postgres.

93 00:22:59.780 00:23:02.260 Awaish Kumar: Okay. I run it through dbt.

94 00:23:02.570 00:23:05.389 Awaish Kumar: And what I want to do is, I want to…

95 00:23:06.010 00:23:09.219 Awaish Kumar: It… and it has… it’s a very big table, it has indexes.

96 00:23:10.670 00:23:15.280 Awaish Kumar: So, if I load it… load the data into the model.

97 00:23:16.730 00:23:19.410 Awaish Kumar: With the indexes, it takes a lot of time to load.

98 00:23:19.770 00:23:22.099 Awaish Kumar: What I’m trying to do is I’m trying to

99 00:23:23.260 00:23:27.780 Awaish Kumar: Delete the indexes, load the data, and recreate the indexes.

100 00:23:27.960 00:23:31.019 Awaish Kumar: So, how can we do it using dbt?

101 00:23:32.180 00:23:39.390 Adegbite Ayoade: Okay, so you want to… so you have a large table that has, let’s say, maybe a million rows.

102 00:23:39.660 00:23:46.390 Adegbite Ayoade: And it has indexes so that it’s faster when, being queried, right? But then.

103 00:23:46.880 00:23:56.700 Adegbite Ayoade: when you are loading, of course, indexes will slow down your loading process, so you want to delete, load your data, then reapply indexes all over again. Am I right?

104 00:23:57.120 00:24:14.839 Adegbite Ayoade: Okay, so, I think it’s a very simple process, so we can use the, hook operation in dbt, so that after the table has been loaded, then we now reactivate the indexes on them, basically.

105 00:24:15.860 00:24:17.770 Awaish Kumar: Yeah, so how can we use that?

106 00:24:18.910 00:24:22.100 Adegbite Ayoade: Okay, so, we can say that in the…

107 00:24:22.210 00:24:40.569 Adegbite Ayoade: configuration of the table, right? So we can have, like, maybe a pre-hook that first drop the index on the table, and based on the column that those indexes are applied to, then we have the second pre-hook that, you know, applies those, indexes back.

108 00:24:40.730 00:24:44.829 Adegbite Ayoade: Right? Once the data has now been fully created.

109 00:24:46.410 00:24:47.150 Awaish Kumar: Okay.

110 00:24:47.850 00:24:51.670 Awaish Kumar: Okay, so it’s a poster that we can use afterwards.

111 00:24:52.970 00:25:10.879 Adegbite Ayoade: Yeah, and I think another approach might be to just use, maybe use a macro hook, right? That just, you know, creating, especially if you don’t want to use, you know, pre-hook again, because that would also, like, you know, cost more, operational awareness.

112 00:25:10.880 00:25:12.739 Awaish Kumar: There are post hooks as well.

113 00:25:12.880 00:25:13.409 Awaish Kumar: So, yeah.

114 00:25:14.260 00:25:15.490 Awaish Kumar: Then we just post them.

115 00:25:15.680 00:25:16.900 Adegbite Ayoade: Who’s the… yeah.

116 00:25:18.740 00:25:19.450 Awaish Kumar: Okay.

117 00:25:20.400 00:25:23.939 Awaish Kumar: Okay, sounds cool. I just…

118 00:25:24.250 00:25:32.419 Awaish Kumar: I’ll ask a few more things regarding your communication. So, how do you communicate with your…

119 00:25:32.780 00:25:37.880 Awaish Kumar: stakeholders, right? If you have a… disagreement.

120 00:25:40.390 00:25:48.310 Awaish Kumar: if you… Did your analysis, and you came up with some numbers, and your stakeholder, or the manager.

121 00:25:48.410 00:25:50.979 Awaish Kumar: Does not agree with your numbers.

122 00:25:51.140 00:25:55.039 Awaish Kumar: So how, then, what are… what is the process you follow?

123 00:25:55.570 00:25:57.820 Awaish Kumar: To… to defend yourself.

124 00:25:59.590 00:26:05.340 Adegbite Ayoade: Okay, so, first, I want to identify

125 00:26:05.900 00:26:20.560 Adegbite Ayoade: that kind of stakeholders, basically, because when I’m dealing with stakeholders, I like to use the power-interest grid to sort of, like, classify the stakeholders into, you know, high power interest in the project.

126 00:26:20.560 00:26:28.290 Adegbite Ayoade: high power, low interest, low power, you know, low interest, or, you know, high interest, low power. So, depending on the spectrum that…

127 00:26:28.300 00:26:47.749 Adegbite Ayoade: that particular stakeholder is forced to, that would help me to determine how best to sort of, like, you know, approach them, right? If it is a stakeholder that is in high power interest, the first thing would be that I want to go back to the metrics definition that we agreed to while we’re starting the project, right?

128 00:26:47.750 00:26:54.050 Adegbite Ayoade: what were the filters that, you know, that were applied to maybe the chart or the visualization that I’ve built?

129 00:26:54.050 00:27:01.110 Adegbite Ayoade: What are the level of aggregation that you have carried out, and all of that, because sometimes,

130 00:27:01.520 00:27:05.379 Adegbite Ayoade: We have usually seen problems, you know, around is…

131 00:27:05.910 00:27:17.479 Adegbite Ayoade: Active users, basically. What does active users mean, right? So, for me, I could say, oh, active users is as long as you log in, you know, once a week, you’re an active user, right?

132 00:27:17.680 00:27:21.759 Adegbite Ayoade: For the stakeholder, they could say, oh, if you log in 3 times a week.

133 00:27:21.830 00:27:29.999 Adegbite Ayoade: we consider you an active visa, so that would be the first thing, right? Let’s… let me be sure that I have the accurate metric definitions that, you know.

134 00:27:30.000 00:27:45.750 Adegbite Ayoade: we both agreed to when we started out on this project, right? The second thing would be that I would try to look at my logics, right, and sort of, like, try to make it transparent to that particular stakeholder that, okay, this is how the data was.

135 00:27:45.840 00:27:56.039 Adegbite Ayoade: And after I applied, you know, my transformation logic and all of that, this is how it is, right? So that way, I’m not, pushing because… let’s now assume that maybe the

136 00:27:56.090 00:28:15.500 Adegbite Ayoade: this particular stakeholder does not have, like, you know, the technical background to understand my codebase and all of that. Me trying to explain what I’ve done in terms of transformation basically helps that person feels like, okay, I’m not putting them in a box, as to say I’m training the dragons around me, right? Then another thing that I would also look at is

137 00:28:15.970 00:28:27.089 Adegbite Ayoade: what… where are they looking at the analysis from? What is their perspective? Are they looking at it from, you know, their understanding of the business, or are they looking at it from what the data is actually saying?

138 00:28:27.090 00:28:36.619 Adegbite Ayoade: Right? So this also, you know, basically helped me to understand how best to come in. Maybe it will now be to involve, you know, another stakeholder that, you know, maybe have

139 00:28:36.620 00:28:44.859 Adegbite Ayoade: more technical background, you know, that could also, contribute to the matter, right? Then, I think the last thing would be,

140 00:28:45.520 00:29:01.630 Adegbite Ayoade: try to build a single source of truth, basically, right? Once we identify, oh, okay, the problem is maybe coming from metric definition, right? Which is also important that while you are meeting with these stakeholders at the beginning of a project.

141 00:29:01.630 00:29:23.009 Adegbite Ayoade: you are documenting everything that, you know, you are agreeing on, and after each meeting, you send them, like, a summary of what that meeting covers, what you agreed on regarding, you know, a particular metric definition, or, you know, things that they said they will get back to you. So that way, even if there’s going to be, like, any form of disagreement, that this is not what I agreed to at the beginning.

142 00:29:23.020 00:29:40.190 Adegbite Ayoade: then you can pull out emails or pull out documents that says, oh, this was, you know, what your, your recommendation basically looks like when we’re starting out this. And outside of all of this, staying calm and focused actually helps a lot, because

143 00:29:40.970 00:29:51.419 Adegbite Ayoade: Of course, whatever it is that I’m building, I’m building it for them, and they need it so that they can make, you know, that informed decision to be able to, you know, drive the business forward.

144 00:29:51.420 00:30:03.899 Adegbite Ayoade: So if, at the end of the whole thing, they feel like I’m actually not solving that problem for them, what I need to start working on is, how best can we, you know, collaborate more? Maybe,

145 00:30:04.120 00:30:20.449 Adegbite Ayoade: bring in someone else that might, you know, be shadowing every single meeting that we are doing, that could say, oh, yeah, I think I was there during this meeting, this was what you guys agreed on, you know, amongst many other things, like that. So that way, we can both get,

146 00:30:20.450 00:30:33.349 Adegbite Ayoade: another person’s perspective that maybe, like, somebody even neutral, to look at the project and, you know, tell us what they feel about, the metrics that we have got turned out and all of that. So, that… that’s how I basically, go around it.

147 00:30:35.460 00:30:37.770 Awaish Kumar: Okay, great. I think…

148 00:30:38.890 00:30:43.989 Awaish Kumar: I’m good with my questions, so I will just leave some time for you to ask any questions.

149 00:30:44.980 00:30:54.489 Adegbite Ayoade: Okay, okay. So, I know that, this is, a consulting arm, and,

150 00:30:55.300 00:31:13.930 Adegbite Ayoade: How does, Brief, I said Brainford. Yeah, Brainforge currently, designs SourceX for this, for this role, because while I… when I was, engaging with, Dame Lady, he said I would be working, you know, directly with him, so how… how is SourceX defined for this particular role?

151 00:31:15.480 00:31:17.720 Awaish Kumar: Sorry, how are the… what…

152 00:31:18.450 00:31:23.979 Adegbite Ayoade: how is sources defined? Like, what defines sources for digital, basically?

153 00:31:24.210 00:31:25.770 Awaish Kumar: Success defines.

154 00:31:25.880 00:31:26.750 Awaish Kumar: Yes.

155 00:31:27.450 00:31:30.779 Awaish Kumar: So, for this role, as Demon already mentioned,

156 00:31:31.940 00:31:34.460 Awaish Kumar: That you will be working directly with him.

157 00:31:34.950 00:31:46.220 Awaish Kumar: That’s true. And, the success is measured by, you can say, like, what’d you say?

158 00:31:47.110 00:31:55.090 Awaish Kumar: Adoptability is fast learning, so how fast you can adopt to the environment and… and keep up.

159 00:31:56.350 00:32:00.490 Awaish Kumar: So, in the first week, for example, if you are on board.

160 00:32:00.640 00:32:05.749 Awaish Kumar: If you are assigned to a new client, and you onboard in the first week, and at least understand the

161 00:32:05.950 00:32:11.190 Awaish Kumar: fully structure, that’s… it’s, like, the shortest, like, I would say.

162 00:32:11.800 00:32:19.360 Awaish Kumar: In the first week of your time, if you… If you are able to… Come on up.

163 00:32:20.730 00:32:26.060 Awaish Kumar: on… get onboarded on a client, and understand the tools tech stake for them, and…

164 00:32:27.530 00:32:29.939 Awaish Kumar: And the models and everything.

165 00:32:30.720 00:32:32.999 Awaish Kumar: That’s… that’s good enough.

166 00:32:33.320 00:32:38.450 Awaish Kumar: And maybe submit some… Like, small bug fixes or something like that.

167 00:32:38.710 00:32:41.690 Awaish Kumar: That’s… that’s, like, the first week.

168 00:32:42.310 00:32:43.040 Adegbite Ayoade: Okay.

169 00:32:43.700 00:32:50.990 Awaish Kumar: Then, on the longer term, what the success at Brainford looks like is it’s a fast-paced environment.

170 00:32:51.580 00:32:54.870 Awaish Kumar: Number one… number two, it is, like,

171 00:32:56.930 00:33:01.749 Awaish Kumar: There would be some kind of context switching that you might be assigned to more than…

172 00:33:03.630 00:33:09.259 Awaish Kumar: More than two clients, so at the same time, you might be working in two or three clients.

173 00:33:09.790 00:33:22.200 Awaish Kumar: That means on the same day, you might have to deliver some work for client A and some work for client B, at least for two of them. So, yeah, there will be some kind of context switching.

174 00:33:22.340 00:33:26.330 Awaish Kumar: There will… as a startup, we obviously have some…

175 00:33:26.920 00:33:35.090 Awaish Kumar: Fast-pacing things that we get from clients, and we try to deliver them as soon as possible if something is urgent.

176 00:33:35.250 00:33:38.000 Awaish Kumar: On, like, you know, resolving it on the same day.

177 00:33:38.560 00:33:46.489 Awaish Kumar: So, on the longer term, if you are able to context switch between multiple clients and deliver data models.

178 00:33:46.900 00:33:55.000 Awaish Kumar: Maybe, yeah, so with, with shortest turnaround time, yeah, that’s, that’s what it looks like. And then…

179 00:33:55.170 00:34:01.490 Awaish Kumar: There are some continuous things, like improving workflow, giving suggestions about How you can improve.

180 00:34:02.970 00:34:11.989 Awaish Kumar: like, option A or B, or how we improve workflows, how we improve our… Pr process, like, whatever

181 00:34:12.199 00:34:13.839 Awaish Kumar: Workflow you’re following.

182 00:34:13.980 00:34:23.660 Awaish Kumar: So, we are giving suggestions, writing playbooks regarding that, like the standard operating procedures for

183 00:34:23.860 00:34:27.369 Awaish Kumar: for creating a new model, if you have something to add.

184 00:34:27.580 00:34:33.900 Awaish Kumar: then you can basically share that knowledge with the team. So these are the, basically, things that

185 00:34:34.139 00:34:39.549 Awaish Kumar: That, yeah, making… That will make you success… succeed in this role.

186 00:34:40.790 00:34:57.899 Adegbite Ayoade: Okay, okay, alright, thank you for that. I think I have just one more question, and, where do you see the biggest gap, today in, Brainforge, process? Basically, maybe it’s, onboarding to a new client, or, you know,

187 00:34:57.960 00:35:01.279 Adegbite Ayoade: Delivering your clients’ projects, things like that.

188 00:35:04.060 00:35:06.780 Awaish Kumar: I think it’s just, like,

189 00:35:07.320 00:35:17.469 Awaish Kumar: We are growing rapidly, so we have a lot of clients coming in, a lot of work, a lot of, obviously, people are working

190 00:35:18.620 00:35:22.420 Awaish Kumar: On multiple times, so, like, now that we have Anyways…

191 00:35:23.240 00:35:26.160 Awaish Kumar: A lot more work for analytics engineer.

192 00:35:27.210 00:35:31.450 Awaish Kumar: Yeah, we started hiring, so it’s more like delivery side of it.

193 00:35:31.690 00:35:34.109 Awaish Kumar: Where we need more help.

194 00:35:35.480 00:35:43.030 Awaish Kumar: So we have people supporting our processes, we have Demonade leading analytics engineering, we have,

195 00:35:45.170 00:35:49.760 Awaish Kumar: So, yeah, but once it ends up on the…

196 00:35:50.660 00:35:57.430 Awaish Kumar: plates of the engineers, it’s where we would like more help, so the… No.

197 00:35:57.690 00:35:59.679 Awaish Kumar: Client comes in, you get the…

198 00:36:00.110 00:36:10.120 Awaish Kumar: plan by Tamilare, and then, obviously, we need to work on that, execute that plan, make sure that it is delivered on time with,

199 00:36:11.470 00:36:18.679 Awaish Kumar: With, like, the… with the… Like, the… To see…

200 00:36:18.960 00:36:28.929 Awaish Kumar: Quality work delivered on time, that’s what is… is we’re looking for, so… Yeah, like,

201 00:36:29.060 00:36:35.569 Awaish Kumar: Handling, data quality checks, process for Data quality checks,

202 00:36:35.990 00:36:39.460 Awaish Kumar: Ensuring that the models that we’re building are

203 00:36:39.730 00:36:47.019 Awaish Kumar: Doesn’t meet those data quality standards that are expected by the client.

204 00:36:47.880 00:36:51.700 Awaish Kumar: So, this is what we are, like, trying to do here.

205 00:36:52.770 00:36:53.480 Awaish Kumar: Yeah.

206 00:36:54.960 00:36:58.510 Adegbite Ayoade: Okay, okay. So I think,

207 00:36:58.850 00:37:04.829 Adegbite Ayoade: This last question is, for you, actually. What’s, like…

208 00:37:05.010 00:37:08.670 Adegbite Ayoade: How does it feel, like, working at Greenforge?

209 00:37:10.840 00:37:16.460 Adegbite Ayoade: like, maybe the most exciting part of, you know, working at Brain Forge, and things like that.

210 00:37:17.670 00:37:26.470 Awaish Kumar: Yeah, I already mentioned a few things in my… both of my answers. There is a… I like being a startup, I like being… working in a fast-paced environment.

211 00:37:26.620 00:37:30.509 Awaish Kumar: I like being somewhere where you get a lot of things to learn.

212 00:37:30.620 00:37:36.490 Awaish Kumar: So it’s an environment where I… I learn, maybe, 10 tools in a month.

213 00:37:37.250 00:37:41.669 Awaish Kumar: And then I’ve… That in any other environment, so…

214 00:37:41.990 00:37:46.860 Awaish Kumar: I can participate with the AI team, I can collaborate with marketing, sales.

215 00:37:47.000 00:37:49.470 Awaish Kumar: So it is, like, although…

216 00:37:49.570 00:37:52.909 Awaish Kumar: I sit in data engineering, so that’s my core job.

217 00:37:53.370 00:37:54.279 Awaish Kumar: To support Thai.

218 00:37:54.880 00:37:57.710 Awaish Kumar: After… after that is done, like…

219 00:37:57.970 00:38:06.370 Awaish Kumar: like, I have all the time to support marketing team, to support sales team, to support AI team.

220 00:38:06.710 00:38:13.849 Awaish Kumar: And also, there are multiple clients coming in. I have a lot of flexibility in deciding, setting up data architecture.

221 00:38:14.180 00:38:15.420 Awaish Kumar: And,

222 00:38:16.080 00:38:22.589 Awaish Kumar: And learning about new tools that those new clients will come up with, and we might have to integrate, or…

223 00:38:22.860 00:38:28.580 Awaish Kumar: Yeah, on, on… like, deploy those tools for them. So, basically.

224 00:38:28.730 00:38:32.370 Awaish Kumar: So, yeah, being here, I learned a lot of different tools.

225 00:38:32.600 00:38:35.409 Awaish Kumar: From a lot of things, and…

226 00:38:36.700 00:38:42.660 Awaish Kumar: And yeah, being at a startup, it’s more like, also have a lot of ownership in defining

227 00:38:43.000 00:38:47.760 Awaish Kumar: Quality and, processes for doing… Data work.

228 00:38:47.870 00:38:48.640 Awaish Kumar: Pam.

229 00:38:48.970 00:38:56.150 Adegbite Ayoade: Yeah. Yeah, that makes a lot of sense. Linkedin tools, that’s amazing.

230 00:38:57.240 00:38:57.890 Awaish Kumar: Yeah.