Meeting Title: Brainforge Interview w- Awaish Date: 2026-03-24 Meeting participants: David Rose, Awaish Kumar
WEBVTT
1 00:00:10.010 ⇒ 00:00:12.289 Awaish Kumar: Hi, Devin, how are you?
2 00:00:13.640 ⇒ 00:00:14.530 David Rose: Hello!
3 00:00:16.020 ⇒ 00:00:17.000 Awaish Kumar: Oh, you’re doing?
4 00:00:17.470 ⇒ 00:00:18.610 David Rose: Good, how are you?
5 00:00:19.140 ⇒ 00:00:20.260 Awaish Kumar: I’m good as well.
6 00:00:22.540 ⇒ 00:00:27.189 Awaish Kumar: Yeah, so thank you for taking the time for this interview.
7 00:00:28.630 ⇒ 00:00:32.790 Awaish Kumar: And this… Session, we are just going to talk, like,
8 00:00:33.590 ⇒ 00:00:36.539 Awaish Kumar: We’re just going to get to know
9 00:00:36.790 ⇒ 00:00:42.849 Awaish Kumar: You better, and what you have been doing so far, and maybe walk over any one of your projects.
10 00:00:42.980 ⇒ 00:00:46.990 Awaish Kumar: And then, maybe if… and yeah, we can then maybe…
11 00:00:47.740 ⇒ 00:00:51.219 Awaish Kumar: Take some questions from you if you want to know anything about brain force.
12 00:00:51.920 ⇒ 00:00:54.170 Awaish Kumar: Sounds right.
13 00:00:55.460 ⇒ 00:00:56.030 David Rose: Great.
14 00:00:56.620 ⇒ 00:01:00.760 Awaish Kumar: Okay, can you… Introduce yourself.
15 00:01:01.670 ⇒ 00:01:19.090 David Rose: Yeah, so, I’m David. My, my background, so I guess we’ll start there. I went to… I went to the University of Minnesota for industrial and systems engineering, and so I went… I went right into that out of college and did,
16 00:01:19.250 ⇒ 00:01:35.990 David Rose: Industrial engineering and manufacturing engineering, and process engineering, so there’s a lot of, everything from examining, sort of, our output data and yields and modeling machine capacity, and, like, production line capacity.
17 00:01:36.240 ⇒ 00:01:46.389 David Rose: To, like, actually videotaping people and… and doing time studies on… on processes. And basically in that.
18 00:01:47.010 ⇒ 00:01:49.579 David Rose: When working in that space,
19 00:01:49.730 ⇒ 00:01:59.900 David Rose: I just sort of kept finding that my favorite projects that I worked on were ones where I got to do a bit of data analysis, or, you know.
20 00:02:00.120 ⇒ 00:02:08.909 David Rose: come up with something and make a beautiful visualization that really communicated something and changed the way we did things.
21 00:02:09.110 ⇒ 00:02:25.539 David Rose: So I tried to… basically just tried to transfer my career into that space. So I was working and doing those in medical device manufacturing, so pacemaker batteries and surgical systems, things like that.
22 00:02:25.730 ⇒ 00:02:30.689 David Rose: And then I… I transitioned to work for a cycling…
23 00:02:31.210 ⇒ 00:02:35.710 David Rose: wheel distribution company in their, in their warehouse.
24 00:02:35.800 ⇒ 00:02:48.239 David Rose: Because I just wanted to get my foot in the door, and then transitioned into, working on their, on their website, some of their internal analytics, because they were doing.
25 00:02:48.240 ⇒ 00:02:56.209 David Rose: quite a lot of, basically, the point that I joined the company, everything was, like, we’re doing all of our analytics and…
26 00:02:56.210 ⇒ 00:03:06.359 David Rose: in Excel, and on Monday, this person downloads all these, you know, downloads from here, here, here, puts it into their, puts it into their spreadsheet.
27 00:03:06.360 ⇒ 00:03:17.150 David Rose: finishes it, sends it out, and they’re getting to the point where the spreadsheet is slow, oh, now the spreadsheet’s ran out of rows, we can’t actually do this, we need something else. So…
28 00:03:17.480 ⇒ 00:03:22.660 David Rose: I started working, on a solution for them. They had an outside,
29 00:03:23.000 ⇒ 00:03:41.439 David Rose: consultant come and set up some extraction scripts and try to set them up, with the snowflake, with loading all the data they needed to Snowflake and starting a dbt project to get to their analytics that could run automatically instead of, you know, every Monday someone or…
30 00:03:41.460 ⇒ 00:04:00.159 David Rose: Pam does the spreadsheets, and if she’s gone, then we don’t have any analytics for the week. But, include… and then, so, in addition to that, analytics, some other, just classic ETL, and some other processes that I was able to automate, with Python.
31 00:04:00.240 ⇒ 00:04:08.450 David Rose: And from there, I moved to my current role, which is now, healthcare analytics. So my title is Data Engineer.
32 00:04:08.660 ⇒ 00:04:16.620 David Rose: But that’s… it would be more accurate to say I’m an analytics engineer. I’m basically building… building dbt models, we’re doing…
33 00:04:16.850 ⇒ 00:04:28.630 David Rose: in, you know, extracting some of the input data, and stuff, but that’s pretty light Python locally. There’s no sort of, like, higher infrastructure to, that we’re managing.
34 00:04:28.930 ⇒ 00:04:38.950 David Rose: To get what we need. And it’s mostly focused around my… I guess our niche is, like, population scale, healthcare analytics, so the…
35 00:04:38.950 ⇒ 00:04:51.030 David Rose: The basic, example is always breast cancer screening. So that might be, a state or, like, a hospital might say, what’s our screening rate? How do we compare to other people?
36 00:04:51.050 ⇒ 00:05:06.979 David Rose: And we take the clinical data, claims data, benefits data, and we’ll aggregate that into calculating, you know, here’s your 70% screening rate, here are the patients that should have had a screening but didn’t.
37 00:05:07.090 ⇒ 00:05:15.620 David Rose: And sort of, we’re in the middle of building out that, metrics suite, right now as my current project.
38 00:05:16.370 ⇒ 00:05:20.120 Awaish Kumar: Okay, and, like, what are you looking for in your next role?
39 00:05:20.750 ⇒ 00:05:37.360 David Rose: So there’s… there’s two things, really, I guess is… one, I’m very niche right now in my healthcare analytics, so I think it would be good to expand and be doing… I did a little bit of retail analytics and a little bit of production output in my previous jobs, but I wasn’t…
40 00:05:37.440 ⇒ 00:05:55.349 David Rose: as heavily in, using a real warehouse and SQL for that. There was a lot of Excel and, you know, or custom little Python Panda scripts that work for this one thing but aren’t going to be useful, it’s just something else. So the… that’s one, is, I guess expanding the scope of
41 00:05:55.410 ⇒ 00:06:01.920 David Rose: Of what I’m working in. And then the second one is, I’m… I’m sort of the…
42 00:06:02.120 ⇒ 00:06:11.820 David Rose: higher level at my company of analytics engineering, and there’s no one really for me to learn to, or learn from, and I be, you know…
43 00:06:11.820 ⇒ 00:06:24.360 David Rose: read a lot of documentation and ask AI a lot of questions about what the best way to do this or that would be, but, I don’t think those are sufficient to replace, so I think I’ve kind of reached a ceiling in my
44 00:06:24.390 ⇒ 00:06:26.579 David Rose: In my current role.
45 00:06:29.210 ⇒ 00:06:29.940 Awaish Kumar: Okay.
46 00:06:30.570 ⇒ 00:06:37.069 Awaish Kumar: And, so yeah, can we walk through one of your recent projects that you’ve delivered end-to-end?
47 00:06:39.470 ⇒ 00:06:50.980 Awaish Kumar: Delivered it end-to-end, and then you can talk about, like, what tech stack was used, what exactly were your responsibilities, was it a team effort, or a…
48 00:06:51.340 ⇒ 00:06:56.420 Awaish Kumar: Individual effort, and then… What was the final outcome?
49 00:06:58.450 ⇒ 00:07:05.820 David Rose: Sure, so the… basically, my previous main project at my current role, was to…
50 00:07:05.960 ⇒ 00:07:11.079 David Rose: Re-recalculate the analytics that, previous person had done.
51 00:07:11.180 ⇒ 00:07:13.979 David Rose: For a client, so they were… they were doing a…
52 00:07:14.240 ⇒ 00:07:19.749 David Rose: Four-year study or so on a number of behavioral health
53 00:07:19.930 ⇒ 00:07:27.699 David Rose: metrics. So we got the statewide data and, the existing sort of tech stack was
54 00:07:27.810 ⇒ 00:07:46.940 David Rose: here is… here’s the repository, here are the different outputs of, it’s reports that we’re sending them, the final results, but we’re also sending them all the code and all the, input steps. And my assignment was to take these and rerun them for the new calendar year.
55 00:07:47.040 ⇒ 00:08:00.019 David Rose: three more times, basically. We’d submitted the first one, and the issue I ran with, ran into was this was all written… this was a combination of SQL modeling and Redshift.
56 00:08:01.010 ⇒ 00:08:10.659 David Rose: sometimes you would download the data and run some Python models, and then re-add it back to Redshift. And when I was taking that over, it…
57 00:08:10.660 ⇒ 00:08:18.999 David Rose: became apparent that we needed to be using dbt. So that was basically the main thing I did on that.
58 00:08:19.000 ⇒ 00:08:30.700 David Rose: project was refactor everything that they had done, because they… the instructions that I had were, okay, you need to create this table, now create this view. Now that those are created.
59 00:08:30.880 ⇒ 00:08:45.099 David Rose: make this view, and now make this view, and go in and edit the dates of every single one so that it’s the new calendar year instead of the old calendar year, and you’re going to want to run these tests, and I just kind of stepped back and said, can I please
60 00:08:45.230 ⇒ 00:08:48.180 David Rose: Change this entire structure, because it’s…
61 00:08:49.350 ⇒ 00:09:00.169 David Rose: extremely time-consuming and not repeatable, so I basically refactored it into a DBG project with input variables for the time periods we needed.
62 00:09:00.260 ⇒ 00:09:12.789 David Rose: And some automatically test the… the sort of, like, ad hoc tests that we were doing at the end, so that basically the… the start was to refactor into a new
63 00:09:12.890 ⇒ 00:09:14.770 David Rose: dbt project set up.
64 00:09:14.870 ⇒ 00:09:18.330 David Rose: Fully automate it.
65 00:09:18.360 ⇒ 00:09:34.250 David Rose: regression test it so that my updated project was getting the exact same results as the old project. Then I went in and was able to just update the variables for the next year, rerun it, update the variables for the next year, rerun it. We did, you know, inspect the
66 00:09:34.250 ⇒ 00:09:45.720 David Rose: the data and the input schema as well to ensure it hadn’t drifted. But the end result was basically… took… the first time I repeated the analysis took
67 00:09:45.850 ⇒ 00:09:59.159 David Rose: probably a month just to figure out how to rerun the code. And after that, it was just a day of change to variable, rerun, change the variable, rerun for… for calculating the next years,
68 00:09:59.350 ⇒ 00:10:09.320 David Rose: So, saved a bunch of time on redoing the same metrics again in the next calendar year, and rewriting our reports and submitting everything.
69 00:10:09.480 ⇒ 00:10:13.829 David Rose: And the end result was, I guess we got it done quicker than…
70 00:10:14.010 ⇒ 00:10:27.650 David Rose: quicker than expected, which was nice, and actually ended up sending… sending over all the reports, all the data, all the code, to a S3 bucket for the client, and erased everything we were,
71 00:10:27.980 ⇒ 00:10:31.670 David Rose: I guess I’m forgetting the exact term, but in medical…
72 00:10:31.950 ⇒ 00:10:35.930 David Rose: Medical data, where a steward, or…
73 00:10:36.210 ⇒ 00:10:41.370 David Rose: Something where we, you know, we’re done with it, it all gets erased, but we pass them off all the…
74 00:10:41.580 ⇒ 00:10:44.740 David Rose: All the information they needed to be able to rerun.
75 00:10:44.910 ⇒ 00:10:54.379 David Rose: everything the next year and the next year in the same way if they want to, and have a test result tell them if the schema has drifted, or if there’s anything that they…
76 00:10:54.710 ⇒ 00:10:59.429 David Rose: Would signal that those analytics are no longer correct.
77 00:11:01.110 ⇒ 00:11:06.360 Awaish Kumar: Okay, so have you used, like, how much do you have experience with using dbt?
78 00:11:07.440 ⇒ 00:11:25.800 David Rose: It’s essentially the last year or so in my new role. When I started this analytics, I have a friend who’s an analytics engineer, and I was talking to him about the problems I was having, and he was like, you should be using this, and I read into it, and did dbt fundamentals, etc.
79 00:11:25.940 ⇒ 00:11:31.710 David Rose: Have, so basically, about a year of… Almost every day.
80 00:11:32.340 ⇒ 00:11:38.779 David Rose: I’m working in a repository with a dbt project. I’m not writing dbt models every single day, but
81 00:11:38.950 ⇒ 00:11:40.639 David Rose: Pretty much, almost.
82 00:11:41.990 ⇒ 00:11:45.360 Awaish Kumar: How familiar are you with DBTs?
83 00:11:45.830 ⇒ 00:11:46.890 Awaish Kumar: Concepts?
84 00:11:47.900 ⇒ 00:11:50.600 David Rose: I’d say very familiar,
85 00:11:50.650 ⇒ 00:12:09.360 David Rose: I, at least with the basic concepts, I’m not super familiar with, say, advanced things, like, snapshots, or, getting into very fancy, I guess, materialized, different, you know, types of, like, incremental materialization.
86 00:12:09.500 ⇒ 00:12:15.420 David Rose: But I think I do have a solid understanding of the basics, I would say.
87 00:12:16.690 ⇒ 00:12:17.470 Awaish Kumar: Okay.
88 00:12:17.750 ⇒ 00:12:26.769 Awaish Kumar: but… For example, if I run a cure… If I have a model.
89 00:12:26.910 ⇒ 00:12:34.079 Awaish Kumar: in DBT, And then after that model finishes, I just want to…
90 00:12:34.290 ⇒ 00:12:40.800 Awaish Kumar: run a SQL statement which adds an index on one of the columns.
91 00:12:41.750 ⇒ 00:12:42.420 Awaish Kumar: Wow.
92 00:12:42.900 ⇒ 00:12:44.100 Awaish Kumar: How would you do that?
93 00:12:44.870 ⇒ 00:12:47.880 David Rose: That I’m not familiar with, so I couldn’t tell you.
94 00:12:49.300 ⇒ 00:12:49.820 Awaish Kumar: Okay.
95 00:12:49.820 ⇒ 00:12:50.480 David Rose: Yep.
96 00:12:51.080 ⇒ 00:12:53.999 Awaish Kumar: And… are you familiar with dbt Seeds?
97 00:12:54.540 ⇒ 00:12:55.889 David Rose: Yes.
98 00:12:56.210 ⇒ 00:12:56.850 David Rose: Yep.
99 00:12:56.850 ⇒ 00:12:57.510 Awaish Kumar: Okay.
100 00:12:57.710 ⇒ 00:13:01.079 Awaish Kumar: And work? Like, why we should choose that?
101 00:13:01.780 ⇒ 00:13:21.019 David Rose: So those are for all of our reference data. So my, in my use case, I have some sample input outputs of golden data. We have a patient, all their data, and then their test results, and so those are seats for us, because we’re not… it’s not data that’s expected to change very much. We also have a set of
102 00:13:21.040 ⇒ 00:13:23.329 David Rose: What are called value sets.
103 00:13:23.510 ⇒ 00:13:38.010 David Rose: Which is, did a patient, have a certain procedure, or have a certain observation and surgery, and all of the medical codes? So that’s a… that’s a set that gets put out basically yearly, so that would be a seed for us as well, so…
104 00:13:38.220 ⇒ 00:13:41.740 David Rose: Seeds, yeah, reference… reference data,
105 00:13:41.740 ⇒ 00:13:45.440 Awaish Kumar: But, like, what is the benefit of using dbt seeds?
106 00:13:45.930 ⇒ 00:13:51.799 Awaish Kumar: like… what exactly you will be doing in… if you have to use dbt seats.
107 00:13:53.430 ⇒ 00:13:58.880 David Rose: I’m a little confused by this question, sorry. Oh, the benefit is…
108 00:14:00.490 ⇒ 00:14:06.399 Awaish Kumar: Like, the way you described it is more like you have reference data, which could be in a database table.
109 00:14:06.580 ⇒ 00:14:07.630 Awaish Kumar: Yeah. Right.
110 00:14:08.430 ⇒ 00:14:12.989 Awaish Kumar: But how that… how dbt relates to that table, that’s my question.
111 00:14:13.690 ⇒ 00:14:17.099 David Rose: Oh, I guess, I mean, I guess the benefit would be that it,
112 00:14:17.290 ⇒ 00:14:22.900 David Rose: dbt is loading those, and we’re able to keep those in a, change con…
113 00:14:23.200 ⇒ 00:14:28.030 David Rose: change-controlled environment, so it’s, you know, living on our GitHub repo, or if it’s a…
114 00:14:28.190 ⇒ 00:14:29.520 Awaish Kumar: is loading.
115 00:14:30.030 ⇒ 00:14:31.800 Awaish Kumar: That, that’s my question.
116 00:14:32.550 ⇒ 00:14:35.070 Awaish Kumar: How DBTeam’s loading that data?
117 00:14:35.960 ⇒ 00:14:42.469 David Rose: I believe it’s, ins… like, it’s just insert statements behind the scenes, but I’m not…
118 00:14:43.140 ⇒ 00:14:47.339 Awaish Kumar: Yeah, my question is just that, like, you provide some CERIC files, right?
119 00:14:47.340 ⇒ 00:14:47.730 David Rose: Hmm.
120 00:14:48.050 ⇒ 00:14:52.330 Awaish Kumar: Deputy Seeds folder, which it recognizes seeds.
121 00:14:52.490 ⇒ 00:14:57.300 Awaish Kumar: And then it… it obviously will generate some insert statements.
122 00:14:59.030 ⇒ 00:14:59.650 Awaish Kumar: Yeah.
123 00:15:00.700 ⇒ 00:15:04.999 Awaish Kumar: Like, that was my question, that it’s… it does not…
124 00:15:05.160 ⇒ 00:15:06.990 Awaish Kumar: It does not live in,
125 00:15:07.190 ⇒ 00:15:16.570 Awaish Kumar: in a table, we don’t actually get it directly into the table, we basically just get a file and put it in the dbt folder, dbt seeds folder.
126 00:15:16.690 ⇒ 00:15:18.699 Awaish Kumar: And it does the vision for us.
127 00:15:19.450 ⇒ 00:15:20.870 Awaish Kumar: Okay.
128 00:15:21.620 ⇒ 00:15:26.009 Awaish Kumar: We can move on. And are you familiar with the databases?
129 00:15:26.630 ⇒ 00:15:33.399 David Rose: Yes, I’m mostly just Redshift, and then, been using DuckDB for local
130 00:15:33.990 ⇒ 00:15:40.520 David Rose: local analytics, have never used Mother Duck, but basically just those two.
131 00:15:41.880 ⇒ 00:15:46.689 Awaish Kumar: So, for example, if given a table, which takes a really long time to carry.
132 00:15:47.240 ⇒ 00:15:51.490 Awaish Kumar: Because it is… it has… it is a really big table, has maybe…
133 00:15:51.610 ⇒ 00:15:53.600 Awaish Kumar: Billions of rows in it.
134 00:15:53.730 ⇒ 00:15:59.949 Awaish Kumar: And, it takes, like, for me to curate, it will take, like, 5 to 10 minutes to basically…
135 00:16:00.060 ⇒ 00:16:01.720 Awaish Kumar: Complete the results.
136 00:16:02.450 ⇒ 00:16:10.779 Awaish Kumar: Some… Optimization techniques that you can use to reduce the query time.
137 00:16:12.970 ⇒ 00:16:16.530 David Rose: I guess I under… I’ve never implemented it, but,
138 00:16:16.780 ⇒ 00:16:28.710 David Rose: would be partitioning, would be something that could help, or… I know, basically, the only thing I’ve done for… is removing order buys. We used to have a lot of,
139 00:16:28.990 ⇒ 00:16:34.900 David Rose: We had some models that were very slow, and it was because we were ordering a billion-row table.
140 00:16:35.010 ⇒ 00:16:40.499 David Rose: But I’ve… I’ve not gotten too much into that optimization myself, or looked at,
141 00:16:40.600 ⇒ 00:16:45.489 David Rose: partitioning. I know that we… we will probably be adding that to our project soon,
142 00:16:45.880 ⇒ 00:16:54.110 David Rose: do… with how our metrics are very, like, date-based and patient-based, but it’s not something I’ve personally, done too much of.
143 00:16:55.490 ⇒ 00:16:56.310 Awaish Kumar: Okay.
144 00:16:56.480 ⇒ 00:17:04.060 Awaish Kumar: Oh… Okay, and how do you normally communicate your… Findings with non-technical stakeholders.
145 00:17:04.970 ⇒ 00:17:06.819 David Rose: Generally…
146 00:17:06.990 ⇒ 00:17:18.469 David Rose: we’re doing PowerPoint presentations and, try to draw, like, you know, simple block diagrams. We have… we have a complicated architecture of,
147 00:17:19.060 ⇒ 00:17:25.760 David Rose: how all of our data flows into Redshift, and then how our analytics are, are,
148 00:17:25.869 ⇒ 00:17:31.660 David Rose: Output, but it generally, like a draw.io diagram.
149 00:17:31.860 ⇒ 00:17:38.519 David Rose: To communicate, and then try to answer any questions that someone more technical would have an answer for.
150 00:17:40.790 ⇒ 00:17:41.700 Awaish Kumar: Okay.
151 00:17:44.270 ⇒ 00:17:47.900 Awaish Kumar: And, like, if you have any disagreements, if you’re…
152 00:17:48.590 ⇒ 00:17:55.469 Awaish Kumar: stakeholders, disagrees with your findings, how do you pick your… your solution?
153 00:17:57.740 ⇒ 00:18:04.550 David Rose: That’s a good question. I’ve not really had that issue, but I would think
154 00:18:05.310 ⇒ 00:18:19.179 David Rose: We have to figure out if they have a particular reason why, and then go into the code and show them how a… you know, maybe it’s a definition about what an active customer or active patient is, or
155 00:18:19.720 ⇒ 00:18:28.960 David Rose: something along those lines. The analytics I work in are very prescriptive. This is exactly the definition, so that doesn’t come up
156 00:18:29.040 ⇒ 00:18:42.289 David Rose: Too often, but, I think I would probably want to write down their questions and go into the definitions of things is usually, I think, where those disagreements would come from. I know that did happen at my,
157 00:18:42.570 ⇒ 00:18:56.310 David Rose: retail company that, you know, we were defining active customers, or number of orders is that number of orders shipped, or number of orders that’s shipped and then not returned, things like that. You just kind of have to get into the semantics and…
158 00:18:56.410 ⇒ 00:18:59.059 David Rose: Agree on exactly how to define it.
159 00:18:59.540 ⇒ 00:19:05.239 David Rose: Usually, they’ll define it in English language, you have to define it in SQL, of course, which is…
160 00:19:05.400 ⇒ 00:19:09.500 David Rose: Can be a little nuanced in there, but, yeah.
161 00:19:10.890 ⇒ 00:19:12.970 Awaish Kumar: Okay, I think that’s it from my side.
162 00:19:13.240 ⇒ 00:19:15.880 Awaish Kumar: Yeah, I can answer if you have any questions.
163 00:19:17.920 ⇒ 00:19:26.510 David Rose: I guess… And the role would be analytics engineering, so I’m wondering, what’s the typical, like.
164 00:19:26.910 ⇒ 00:19:44.459 David Rose: time that you’re working with a specific client? Like, is it, you know, hey, you’re gonna work with this person, it’s gonna be the, you know, basically forever, or is it 30 days here, next year on this one, it’s gonna take probably 20 days, then do this one? I’m just curious what that looks like for you guys.
165 00:19:44.460 ⇒ 00:19:48.629 Awaish Kumar: It’s more like, you will be working it more than one client at a time.
166 00:19:49.710 ⇒ 00:20:05.350 Awaish Kumar: We don’t assign a person completely. Normally, it’s not the case, because we try to optimize the hours needed for each person on a specific client. So, maybe for a client, we need 20 hours of DE work, 20 hours of AE work.
167 00:20:05.470 ⇒ 00:20:16.669 Awaish Kumar: and 20 hours on data analyst, so that will be like that. So, you might be working more than, like, maybe on 2 clients or 3 clients, depend on how much,
168 00:20:16.780 ⇒ 00:20:24.670 Awaish Kumar: resourcing need there is for your work, right? If a client needs 10 hours of your analytics engineering work.
169 00:20:24.730 ⇒ 00:20:37.210 Awaish Kumar: you will be just assigned for that 10 hours for that client, and then your rest 30 hours will be assigned somewhere else. So at the same time, you might be working for more than 1, 2, or 3. That’s the max.
170 00:20:37.560 ⇒ 00:20:48.390 Awaish Kumar: And… After that, yeah, that’s… I think that’s basically it, and that’s how… Normally, like,
171 00:20:49.090 ⇒ 00:20:58.659 Awaish Kumar: Once you start to grow as a senior on a client, like, that’s… that’s when maybe, you spend maybe less hours, because once we have…
172 00:20:59.320 ⇒ 00:21:07.459 Awaish Kumar: created a strong foundation, like, then it reduces your AE work, then it’s mostly, like, data analysts, and then…
173 00:21:07.690 ⇒ 00:21:11.100 Awaish Kumar: Showing the findings to the stakeholders.
174 00:21:11.380 ⇒ 00:21:18.100 Awaish Kumar: And the data analytics engineers or data engineers are, like, kind of support. If needed, they will be there.
175 00:21:18.250 ⇒ 00:21:22.340 Awaish Kumar: For the modeling work, otherwise they move on to other clients.
176 00:21:22.810 ⇒ 00:21:23.400 David Rose: Sure.
177 00:21:24.340 ⇒ 00:21:25.270 David Rose: Okay.
178 00:21:25.620 ⇒ 00:21:35.620 David Rose: So I have two more questions. I guess the next one is, what… what’s your role at Brainforge, and how do you… do you like, you know, working there? Like, what’s your favorite part about…
179 00:21:35.780 ⇒ 00:21:38.110 David Rose: the company, or… or your role. Yeah.
180 00:21:38.110 ⇒ 00:21:41.330 Awaish Kumar: It’s kind of growing rapidly.
181 00:21:41.730 ⇒ 00:21:48.529 Awaish Kumar: Fast-moving company, use AI extensively at every… in every department.
182 00:21:48.870 ⇒ 00:21:56.790 Awaish Kumar: So that’s, like, something you… I like that we are adopting to the changes that are coming in the market.
183 00:21:57.540 ⇒ 00:22:01.019 Awaish Kumar: We… we are acceptable.
184 00:22:03.190 ⇒ 00:22:08.090 Awaish Kumar: basically, to use AI in your delivery work, like, and…
185 00:22:08.390 ⇒ 00:22:11.100 Awaish Kumar: And then, like, you get a lot of flexibility.
186 00:22:11.270 ⇒ 00:22:13.030 Awaish Kumar: So…
187 00:22:13.150 ⇒ 00:22:21.070 Awaish Kumar: How you want to do your work, but that comes with accountability, obviously, that you have to answer for your work.
188 00:22:21.200 ⇒ 00:22:34.490 Awaish Kumar: And then I’m working as a data engineering lead, so I’m kind of doing all the data engineering work, and helping with defining best practices for all the data engineering work.
189 00:22:34.850 ⇒ 00:22:36.600 Awaish Kumar: And apart from that.
190 00:22:36.890 ⇒ 00:22:44.550 Awaish Kumar: I’ve been here for a year, and it has grown rapidly since I joined, so I’m liking it here.
191 00:22:45.200 ⇒ 00:22:46.110 David Rose: Nice.
192 00:22:47.550 ⇒ 00:22:49.050 David Rose: And then, I think…
193 00:22:49.150 ⇒ 00:22:58.679 David Rose: I’ve learned this from watching some talks that the CEO has given on Brainforge, but it sounds like you guys are tool agnostic in
194 00:22:58.820 ⇒ 00:23:06.620 David Rose: You know, the warehouses that your clients use, etc, kind of just use specific, or… or are there, like, certain…
195 00:23:06.810 ⇒ 00:23:14.890 David Rose: Tools or warehouses that you’re, you know, you guys lean towards, or is it more just 100% what the client is?
196 00:23:14.890 ⇒ 00:23:19.400 Awaish Kumar: If the use case… it depends on the use cases, so it’s not…
197 00:23:19.630 ⇒ 00:23:24.560 Awaish Kumar: Normally our clients are, like, not, like… they don’t come up with any tools, right?
198 00:23:24.850 ⇒ 00:23:31.259 Awaish Kumar: Sometimes they… they do have contracts with, for example, we have IWS
199 00:23:31.420 ⇒ 00:23:36.530 Awaish Kumar: being used in the company, and we have contract with them, okay, then we don’t… we just want to use Redshift.
200 00:23:36.730 ⇒ 00:23:44.239 Awaish Kumar: But that happens very rarely. So some of our other clients are really open.
201 00:23:44.800 ⇒ 00:23:45.750 Awaish Kumar: 2…
202 00:23:46.480 ⇒ 00:23:58.119 Awaish Kumar: to get our feedback and recommendation from us that what are the best suited tools for their use case, and we recommend based on that. We don’t… and also, we don’t recommend based on
203 00:23:58.400 ⇒ 00:24:12.180 Awaish Kumar: What we are familiar with, or where we have relationships. But it also… it completely depends on… on their needs, and then we do the assessment, kind of created… create assessment docs.
204 00:24:13.980 ⇒ 00:24:28.140 Awaish Kumar: on… on the tools, like, for example, if we… if you have to… for example, if you’re looking for a warehouse, then I maybe compare, given the data size use… and the users, and the
205 00:24:28.310 ⇒ 00:24:38.580 Awaish Kumar: the volume… Maybe see 3-4 warehouses and compare them, like Snowflake, Redshift, BigQuery, Mother Duck.
206 00:24:38.850 ⇒ 00:24:46.250 Awaish Kumar: Let’s use, let’s create, like, a comparative analysis of all these tools.
207 00:24:47.160 ⇒ 00:24:55.580 Awaish Kumar: Keeping in mind the client’s use case, and then we should present it to the client that, okay, that’s how it’s going to look like.
208 00:24:55.680 ⇒ 00:24:59.039 Awaish Kumar: And what are you getting, within what price?
209 00:24:59.590 ⇒ 00:25:01.580 Awaish Kumar: And that’s how they make a decision.
210 00:25:02.440 ⇒ 00:25:03.060 David Rose: Okay?
211 00:25:05.210 ⇒ 00:25:09.230 David Rose: Alright, that’s everything I had written down, so… thank you very much.
212 00:25:10.030 ⇒ 00:25:14.890 Awaish Kumar: Okay, yeah, thank you very much, for taking the time for this interview.
213 00:25:16.450 ⇒ 00:25:21.709 Awaish Kumar: So, after I submit my feedback to our recruiter, Kayla, she will…
214 00:25:22.540 ⇒ 00:25:25.939 Awaish Kumar: Yeah, come back to you with the… with the next steps.
215 00:25:26.250 ⇒ 00:25:28.540 David Rose: Okay, great. Thank you very much.
216 00:25:29.540 ⇒ 00:25:30.130 Awaish Kumar: Right.