Meeting Title: Brainforge Candidate Interview and Assessment Date: 2025-12-12 Meeting participants: Uttam Kumaran, Venkata Prasad Krupananda, Robert Tseng
WEBVTT
1 00:00:23.330 ⇒ 00:00:25.239 Uttam Kumaran: Hey, how are you?
2 00:00:25.240 ⇒ 00:00:27.070 Venkata Prasad Krupananda: I’m doing good, how are you? Hi, Roland.
3 00:00:28.230 ⇒ 00:00:30.579 Robert Tseng: Hey, is it Vank, or Venka? You go bank?
4 00:00:31.180 ⇒ 00:00:33.529 Robert Tseng: Vincada? Okay, Vinca, Venka, okay.
5 00:00:35.450 ⇒ 00:00:36.470 Uttam Kumaran: How’s everything?
6 00:00:36.700 ⇒ 00:00:41.130 Venkata Prasad Krupananda: Yeah, it’s going pretty well. Finally, I’m staying at one place right now.
7 00:00:41.600 ⇒ 00:00:44.230 Venkata Prasad Krupananda: Great. When my traveling is done.
8 00:00:44.470 ⇒ 00:00:45.430 Uttam Kumaran: Nice.
9 00:00:45.430 ⇒ 00:00:46.070 Venkata Prasad Krupananda: Awesome.
10 00:00:46.440 ⇒ 00:01:00.710 Uttam Kumaran: I appreciate you, dealing with the delay. We just had a lot of people join the company, so we did intros, and things just… the meeting just went a little bit long, but yeah, I mean, I would love… I don’t know, Robert, if you guys have met yet.
11 00:01:01.270 ⇒ 00:01:03.760 Uttam Kumaran: But, I would love, maybe.
12 00:01:04.030 ⇒ 00:01:14.229 Uttam Kumaran: I think maybe we just start with brief introductions, and then both of us, you know, I think Robert is the one that architected the exercise, so I think maybe, Robert, I’ll let you drive through…
13 00:01:14.300 ⇒ 00:01:24.169 Uttam Kumaran: you know, questions you have there, and then also, of course, like, want to make sure there’s time for you to ask us, you know, any questions. So if that’s a good agenda, then we can start there.
14 00:01:24.210 ⇒ 00:01:24.820 Robert Tseng: Yep.
15 00:01:24.820 ⇒ 00:01:25.720 Venkata Prasad Krupananda: Sounds good.
16 00:01:26.030 ⇒ 00:01:27.910 Venkata Prasad Krupananda: So, do you want to go first, Robert?
17 00:01:28.670 ⇒ 00:01:29.640 Robert Tseng: Yeah, sure.
18 00:01:29.970 ⇒ 00:01:35.870 Robert Tseng: So I run Brainforge with UTAM. I guess, like, this role that we’re hiring for is really to help on
19 00:01:36.080 ⇒ 00:01:41.769 Robert Tseng: the strategy analysis side. So, I guess, you know, for…
20 00:01:42.740 ⇒ 00:01:49.420 Robert Tseng: Most of our clients, we don’t only do engineering work, but we also need to, once we get
21 00:01:49.760 ⇒ 00:01:54.180 Robert Tseng: All the data… all the data kind of set up, when we’re…
22 00:01:54.450 ⇒ 00:02:07.880 Robert Tseng: continuing to work with our clients, it’s… we transition over to doing analysis, and so, I know this exercise was more of an operational analytics exercise, but yeah, I think, like, there’s really…
23 00:02:08.190 ⇒ 00:02:22.580 Robert Tseng: you know, marketing analytics is one big area where we do a lot of work in. There’s a financial analysis as well, which was not part of this exercise. And then ops analytics is not… is like… I think it’s more of, like, a neutral ground that everybody should
24 00:02:22.580 ⇒ 00:02:34.850 Robert Tseng: available to do first, and then you can kind of branch into either domain. So, yeah, I mean, I’ll talk to you more about kind of the assessment, and, I have some questions about it, but, yeah, that’s what I do here.
25 00:02:35.330 ⇒ 00:02:36.210 Venkata Prasad Krupananda: Okay.
26 00:02:36.420 ⇒ 00:02:46.310 Venkata Prasad Krupananda: So yeah, just to talk about me, just an intro. So, my full name goes Venkara Prasakriparanda, and I’ve been working as a, data analyst, for about
27 00:02:46.310 ⇒ 00:03:06.460 Venkata Prasad Krupananda: three and a half to four years now, and I’ve been in this field and focusing on digital analytics and experimentation and data-driven product strategy for clients like healthcare, education, and advertising. And most recently, I’ve been with McKinsey & Company, where I’ve been leading,
28 00:03:06.520 ⇒ 00:03:17.170 Venkata Prasad Krupananda: a lot of, A-B testing, functions, and funnel, and, retention analysis, and also building those data pipelines, and, using a lot of
29 00:03:17.410 ⇒ 00:03:31.229 Venkata Prasad Krupananda: intermediate SQL queries and a little bit of Python to support those digital transformation projects. And also, I work a lot with Amplitude and Mixpanel to track those,
30 00:03:31.600 ⇒ 00:03:33.560 Venkata Prasad Krupananda: User behavior, and…
31 00:03:33.560 ⇒ 00:03:51.919 Venkata Prasad Krupananda: translate that into, you know, actionable insights for product analytics and marketing teams. And before that, I was in the education sector at Texas A&M, where I focused on marketing analytics, segmentation of data sets, and also building Power BI dashboards and optimizing those digital outreach campaigns.
32 00:03:51.920 ⇒ 00:03:54.470 Venkata Prasad Krupananda: And so, yeah, that was…
33 00:03:54.470 ⇒ 00:04:14.030 Venkata Prasad Krupananda: I can say overall it was an intersection of data, product, and strategy, and that’s what drew me onto this particular role, especially the, you know, opportunity to work across different domains and put it together. So yeah, that might help me a lot in the future.
34 00:04:15.730 ⇒ 00:04:18.890 Robert Tseng: Cool. Yeah, I mean, I guess you want me to just jump into it, then?
35 00:04:19.320 ⇒ 00:04:19.890 Uttam Kumaran: Yeah, totally.
36 00:04:19.890 ⇒ 00:04:38.449 Robert Tseng: Okay, great. Yeah, I think… I think you’re, you know, the assessment is great. I think you… you know how to communicate well, and, like, I think you’re… the structured thinking is good, the technical chops are good, so I’m not gonna spend too much, like, kind of in the nitty-gritty of the analysis. I want to spend more time on the recommendation side, and kind of just thinking through
37 00:04:38.450 ⇒ 00:04:51.909 Robert Tseng: I like that you structure it, okay, there are some, like, short-term, medium-term, long-term kind of solutions, and I want to just make this a little bit more, like, realistic. So, yeah, I’d like to kind of just talk about, like, your recommendations on the near-term, like, scrappy moves, and…
38 00:04:51.910 ⇒ 00:04:57.230 Robert Tseng: Yeah, just kind of talk me through within this context, and then maybe I might kind of jump between
39 00:04:57.230 ⇒ 00:05:13.790 Robert Tseng: like, this exercise, and then also, like, your… your experiences, as I start to add, layer in more, complications into it, on kind of, like, what… what you were thinking when… when you were kind of coming up with what the near-term, kind of recommendations would be.
40 00:05:13.990 ⇒ 00:05:14.490 Robert Tseng: So…
41 00:05:14.490 ⇒ 00:05:24.330 Venkata Prasad Krupananda: Yeah, basically what I think of that was, those scrappy, ideas that I came up with. I think it’s better to focus on, what’s,
42 00:05:24.430 ⇒ 00:05:41.369 Venkata Prasad Krupananda: what’s happening in the short term. So, it’s better to focus on short-term aspect, and then, for example, there were, lower rates in, delivery data, and also, when compared to delivery and, the cloud performance, it was… it was… there was big, big,
43 00:05:41.370 ⇒ 00:05:45.909 Venkata Prasad Krupananda: difference when it comes to numbers. So, my basic approach was.
44 00:05:46.120 ⇒ 00:06:01.110 Venkata Prasad Krupananda: focus on the short-term problems, so that can help us, relieve the data and, depend on the data, and then go towards that long-term, goal. So that was what I was trying to, you know.
45 00:06:01.580 ⇒ 00:06:02.190 Venkata Prasad Krupananda: Communicate.
46 00:06:02.190 ⇒ 00:06:08.590 Robert Tseng: Yeah. So, I guess you kind of… you’re… as you’re deciding what are the short-term problems, right? You have these expect…
47 00:06:08.590 ⇒ 00:06:09.830 Venkata Prasad Krupananda: high risk. Yeah.
48 00:06:09.830 ⇒ 00:06:10.200 Robert Tseng: Hey, girl.
49 00:06:10.200 ⇒ 00:06:10.580 Venkata Prasad Krupananda: opportunities.
50 00:06:10.580 ⇒ 00:06:12.010 Robert Tseng: Yeah, kind of walk me through.
51 00:06:12.010 ⇒ 00:06:12.730 Venkata Prasad Krupananda: Hi, how are you?
52 00:06:12.730 ⇒ 00:06:18.369 Robert Tseng: defining risk, and, like, what, how do you opportunity… how do you size opportunities? Yeah.
53 00:06:18.370 ⇒ 00:06:27.039 Venkata Prasad Krupananda: So, when it comes to risks, I think, we can solve a number of risks by, retaining our customers.
54 00:06:27.270 ⇒ 00:06:33.839 Venkata Prasad Krupananda: That would be number one for especially, this assessment. And I think,
55 00:06:34.090 ⇒ 00:06:53.209 Venkata Prasad Krupananda: what I came across is mostly with the percentages that plays around, not even the numbers. So, let’s say the cloud performance is lower in… for a couple of months, and then the delivery data is lower when compared. I think retaining the customers is where we are lacking.
56 00:06:54.580 ⇒ 00:06:57.119 Robert Tseng: Got it. Okay, so yeah, you feel like…
57 00:06:57.590 ⇒ 00:07:09.219 Robert Tseng: Yeah, I mean, when you’re looking at high risk, and you default to, like… I mean, there’s this multiple ways to approach it, right? Like, you’re focused on trying to turn it around for existing customers, right? And, and trying to improve…
58 00:07:09.740 ⇒ 00:07:24.709 Robert Tseng: you’re… especially in operational… in this operational analysis, you’re… there’s… there’s maybe some internal process changes, you’re looking into delivery rates, you’re looking into whatever to try to make the experience better. But let’s take the other… other aside. Okay, let’s say that, like.
59 00:07:24.710 ⇒ 00:07:32.979 Robert Tseng: You lose those customers, and, like, how are you addressing the risk for moving forward outside of your existing customer base?
60 00:07:33.810 ⇒ 00:07:41.700 Venkata Prasad Krupananda: So I saw this probably in two ways. So the first was, Item level.
61 00:07:41.700 ⇒ 00:08:02.940 Venkata Prasad Krupananda: behavior. So over there, what I really felt was the item-level fields reveal patterns, and, you know, most of them were missing around, and most of them were substituted, and the found items which gave insights into merchant stocking quality and the complexity of certain categories. So there, what I learned was, as soon as an
62 00:08:02.940 ⇒ 00:08:18.729 Venkata Prasad Krupananda: order has one missing item, complaint probably jumps, significantly. And the next thing that I can talk about or think about is the merchant level, performance. So when it comes to… I had some notes on that.
63 00:08:19.540 ⇒ 00:08:43.289 Venkata Prasad Krupananda: Yeah, merchant-level performance. So, by aggregating merchants and the metrics that we find there by store, especially, I learned that performance varies substantially across merchants. And what stood out there is some stores had excellent delivery speed, but also, at the same time, they had poor item accuracy. So, others struggled with high,
64 00:08:43.289 ⇒ 00:08:56.699 Venkata Prasad Krupananda: you know, D2R suggesting, poor Dasher proximity or inefficient routing. And the next big thing was consumer experience, coming to the final one. And there, the complaint field allowed me to,
65 00:08:56.700 ⇒ 00:09:12.629 Venkata Prasad Krupananda: quantify those, customer impact of operational failures. So, over there, the key insights that I found out was missing items have stronger effect on complaints than, you know, than the delay or the lateness. So, these were the things.
66 00:09:12.960 ⇒ 00:09:25.410 Venkata Prasad Krupananda: That I would say that we can improve to retain the customers and also satisfy them, and that is what we will be, you know, making a good impact in the future.
67 00:09:25.900 ⇒ 00:09:29.490 Robert Tseng: Okay, yeah, I mean, I think I’ll just kind of final thoughts on this exercise, like.
68 00:09:30.340 ⇒ 00:09:47.320 Robert Tseng: Yeah, I like just kind of calling on things I think you did well. Like, one is, you have the triangulation of, like, okay, look, how do you, like, what is driving, like, higher complaint rates, right? And you’re able to go, and you’re exploring, okay, you’re saying, okay, there’s a correlation between missing items and complaint rates.
69 00:09:47.800 ⇒ 00:10:01.500 Robert Tseng: I think, yeah, oftentimes, like, our clients get so fixated on, like, this… what… whatever metric is going… is going down, like, I think that’s usually what happens, like, they bring in consultants because there is a disaster situation.
70 00:10:01.650 ⇒ 00:10:06.990 Robert Tseng: They don’t really know what’s going on, and they may have some bias on, like.
71 00:10:07.260 ⇒ 00:10:22.350 Robert Tseng: this is… they think they know what the problem is, but they don’t really know. And that’s why, like, as kind of the data team, data… data consultants, we kind of have to… we’re going up against that, often, and trying to kind of point them to something that they’re not looking… that they’re not usually looking at.
72 00:10:22.350 ⇒ 00:10:30.829 Robert Tseng: To, to explain, like, what was happening in their business, right? And so I think it’s in that situa… in these situations, it’s very important
73 00:10:30.830 ⇒ 00:10:37.499 Robert Tseng: To, yeah, like, to not state the obvious, right? You’re not going to a client and saying.
74 00:10:37.990 ⇒ 00:10:40.120 Robert Tseng: Oh yeah, like, complaint rates are going up.
75 00:10:40.120 ⇒ 00:10:40.950 Venkata Prasad Krupananda: Right.
76 00:10:40.950 ⇒ 00:10:47.950 Robert Tseng: But, like, you’re needing to kind of come in from a different angle in order to establish some, some degree of…
77 00:10:47.950 ⇒ 00:10:54.230 Venkata Prasad Krupananda: There, I think what I would think is, instead of stating the obvious, come up with possibilities, and
78 00:10:54.370 ⇒ 00:11:05.790 Venkata Prasad Krupananda: make them, you know, they shouldn’t just rely on one obvious thing. They should have, perspective on different possibilities. And that’s where I think we can come up with
79 00:11:05.920 ⇒ 00:11:13.290 Venkata Prasad Krupananda: operational, solutions or business goal solution, so I think, yeah, that’s where we can win.
80 00:11:13.450 ⇒ 00:11:17.509 Robert Tseng: Yeah, because they’re gonna know the context of their business more than we will, and so…
81 00:11:17.510 ⇒ 00:11:18.160 Venkata Prasad Krupananda: Yes.
82 00:11:18.160 ⇒ 00:11:22.229 Robert Tseng: If we just come to them kind of talking about the one thing that they already think they.
83 00:11:22.640 ⇒ 00:11:23.150 Venkata Prasad Krupananda: It’s got…
84 00:11:23.150 ⇒ 00:11:31.839 Robert Tseng: to make us look not great. Like, you were just… like, why are we paying you? Do you already knew that? And actually, I can tell you 10 more things about it that you didn’t know.
85 00:11:32.040 ⇒ 00:11:48.330 Robert Tseng: But if you’re able to kind of branch out very quickly and be able to connect other trends that they may not be paying attention to, that’s where you start to build that trust. And so I think that’s a really important factor to succeeding in this role with our clients.
86 00:11:48.330 ⇒ 00:11:53.940 Robert Tseng: is, you know, we’re going up against this a lot of the time.
87 00:11:53.940 ⇒ 00:12:07.899 Venkata Prasad Krupananda: And one more thing I just want to add on is, to validate those recommendations, I think what I feel is, for consumer messaging and assignment changes, I would run those A-B tests.
88 00:12:08.250 ⇒ 00:12:27.199 Venkata Prasad Krupananda: Okay. You know, we can come up with A-B tests for that. But for merchant playbooks, I would start off by, you know, with a small, pilot group, and compare, defect rates before and after. So for dash-up of positioning, I would, I would just, you know, run, localized experiments and also,
89 00:12:27.220 ⇒ 00:12:34.910 Venkata Prasad Krupananda: treated versus untreated zones, just experiment on those zones. So, I don’t know if I’m thinking right about this.
90 00:12:34.960 ⇒ 00:12:39.819 Venkata Prasad Krupananda: So, what do you think about my, perspective on this, what I said now?
91 00:12:40.050 ⇒ 00:12:58.379 Robert Tseng: Yeah, no, I think the experimentation approach is… I think that’s… from a data perspective, that’s the right approach, and so I want to also kind of qualify. Yeah, oftentimes, like, in real-world situation, our clients, they don’t know how to run good experiments, and we would run into the situation where we were recommending them, this is what you need to run.
92 00:12:58.450 ⇒ 00:13:05.480 Robert Tseng: But they don’t necessarily know how to… they don’t really execute. And so I think it’s also on us to be able to.
93 00:13:05.480 ⇒ 00:13:09.600 Venkata Prasad Krupananda: Simplify the experiments, make sure that they’re actually executing what we’re recommending, right?
94 00:13:09.600 ⇒ 00:13:11.760 Robert Tseng: And that’s probably a difference between
95 00:13:11.960 ⇒ 00:13:15.250 Robert Tseng: I don’t know, I guess you’ve worked for McKinsey, but at least for other consultancies.
96 00:13:15.250 ⇒ 00:13:15.700 Venkata Prasad Krupananda: I mean.
97 00:13:15.700 ⇒ 00:13:35.669 Robert Tseng: seen, oftentimes they just stop at the recommendation, but we actually are meant to be, like, operational partners to our clients as well. And, you know, we’re an extension of their team. We have, you know, we kind of have to push that second or third order level recommendation to make sure that they follow through on what we’re recommending, right?
98 00:13:35.670 ⇒ 00:13:36.150 Venkata Prasad Krupananda: Perfect.
99 00:13:36.170 ⇒ 00:13:55.820 Robert Tseng: You know, if I’m leading the engagement, and you’re the analyst on that… on that client, and the client adopts your recommendation, and there’s a, you know, obviously there’s a positive result, I want us to get credit for it, right? And I think that’s, that’s part of the… that’s part of the trade-off of putting… pushing out analysis as well, so…
100 00:13:56.200 ⇒ 00:14:09.609 Robert Tseng: just wanted to emphasize that we’re not just in the business of, like, running good analysis and sending the recommendation, but we, you know, we have to be creative with how do we actually get them to run these experiments. So let’s say, for example, like, the treatment size in
101 00:14:09.610 ⇒ 00:14:33.620 Robert Tseng: I mean, here, this is a big company, right? This is basically, like, modeled off of DoorDash, but, you know, oftentimes we’re working with growth stage platform companies. Maybe they’re… the way that they can roll out an experiment is very limited. It’s not gonna just be, like, they can code a change in a couple days and then, you know, push something out. Especially on the marketing side, it’s different. You can change
102 00:14:33.690 ⇒ 00:14:48.470 Robert Tseng: copy, you can change, you know, visuals very, very easily, but I’m talking about, like, a product-level change, right? And if you had to, you know, break down this, what, your recommendation, how would you roll… how would you roll that out for a team that doesn’t know how to run experiments?
103 00:14:48.990 ⇒ 00:14:52.020 Venkata Prasad Krupananda: For something to start off with.
104 00:14:52.020 ⇒ 00:15:11.560 Venkata Prasad Krupananda: Again, yeah, to just connect to what I said earlier, for a team that doesn’t know how to run experiments, I can say, individually, I can pitch in, and I can start off with experimenting, and then, if it’s for stakeholders, yeah, of course, communicate and show them what you’re doing, but if it’s for
105 00:15:11.560 ⇒ 00:15:13.639 Venkata Prasad Krupananda: The team that you’re working with.
106 00:15:13.640 ⇒ 00:15:17.610 Venkata Prasad Krupananda: take initiative, and this is how I look into it.
107 00:15:18.970 ⇒ 00:15:25.320 Robert Tseng: Okay, yeah, I think, very, very much about, like, like you said, if you need to be hands-on, you can go in and you.
108 00:15:25.320 ⇒ 00:15:25.710 Venkata Prasad Krupananda: Yes.
109 00:15:25.710 ⇒ 00:15:28.370 Robert Tseng: go, and you’d be scrappy, you run the experiment with them.
110 00:15:28.370 ⇒ 00:15:28.740 Venkata Prasad Krupananda: My use.
111 00:15:28.740 ⇒ 00:15:38.070 Robert Tseng: just, like, side-by-side. If you’re more of the thought partner, then, like, yeah, you can… you just… you just have to make sure that they understand your thinking, and that’s
112 00:15:38.470 ⇒ 00:15:53.410 Robert Tseng: That’s usually where we get kept at. We’re not really kind of coming in and jumping into their tooling and running the experiment for them, necessarily. But I think being able to, like, manage, like a strategic partner and make sure that
113 00:15:53.410 ⇒ 00:16:04.540 Robert Tseng: is where we are. And sometimes it’s… it’s… it’s tedious. It just means sending the follow-up every day, jumping on the calls over and over again. Yeah, like, I think that’s…
114 00:16:04.540 ⇒ 00:16:16.790 Robert Tseng: That’s a bias that we want for people who have a bias for action. Like, the job is not done just because you, you know, finish the ticket, you send the analysis, but gotta make sure that it’s adopted. So, I just wanted to…
115 00:16:16.790 ⇒ 00:16:25.089 Robert Tseng: you know, go down and make sure… just emphasize that that’s… that’s another… another piece to this job that I feel like is what would… that’s necessary to be successful.
116 00:16:25.090 ⇒ 00:16:49.430 Venkata Prasad Krupananda: Yeah, that is very true, because one biggest thing I took away from analytics is that critical communication is across all the three sides of marketplace when it comes to this particular assessment, so not just operational efficiency. So that… the data showed that many problems could be, you know, prevented and simplified or softened, but simply by setting those right expectations and customer dashes or merchants or
117 00:16:49.430 ⇒ 00:16:52.089 Venkata Prasad Krupananda: Whoever is on the entire project, so…
118 00:16:52.090 ⇒ 00:17:05.220 Venkata Prasad Krupananda: So yeah, communication was the key there, being direct, and also coming up with possibilities, not just one obvious thing that we talked earlier. So yeah, that’s what I think of when it comes to communication.
119 00:17:06.160 ⇒ 00:17:11.599 Robert Tseng: Okay, great. Yeah. Yeah, that’s all I want to say about this exercise.
120 00:17:12.230 ⇒ 00:17:23.659 Robert Tseng: Yeah, I mean, I guess, Utom, I can kind of turn over to you, see if you have any questions you want to ask him. Otherwise, I probably would just ask more questions from his experience. I also want, you know, you to have time to be able to ask us any questions as well.
121 00:17:24.670 ⇒ 00:17:26.050 Robert Tseng: Yeah, so I…
122 00:17:26.050 ⇒ 00:17:26.490 Venkata Prasad Krupananda: I feel…
123 00:17:26.490 ⇒ 00:17:28.160 Robert Tseng: feel good with this exercise. Yeah.
124 00:17:28.580 ⇒ 00:17:33.910 Uttam Kumaran: I guess one question I had is, like, sort of, how does… how do, how are you, how do you typically…
125 00:17:34.220 ⇒ 00:17:44.999 Uttam Kumaran: display, like, results like this, you know, to clients? Like, what are you used to? Is it typically, like, in memo format? Are you used to doing decks? Like, one of the big things I think we’re trying to…
126 00:17:46.160 ⇒ 00:17:57.300 Uttam Kumaran: escalate in terms of, like, the way we do analysis and strategy is just creating assets and things that really drive customers to think, but it’s also, like.
127 00:17:57.830 ⇒ 00:18:00.400 Uttam Kumaran: much better than, like, a screenshot, right? And so…
128 00:18:00.880 ⇒ 00:18:06.480 Uttam Kumaran: I don’t know, I’m just interested in, like, what you’ve seen at McKinsey, or, like, what you feel like you could take, or what you think could be different.
129 00:18:07.460 ⇒ 00:18:20.799 Venkata Prasad Krupananda: So yeah, when it comes to delivering, the standard is usually a mix of, you know, structured decks and also interactive dashboards, so the decks are used to tell the story and guide the client
130 00:18:20.800 ⇒ 00:18:44.530 Venkata Prasad Krupananda: through, the, so what, while the dashboards can also, be built, you know, mostly it’s Power BI dashboards and automated ones, and we let them explore the data themselves afterward, since it’s automated. So these… and also, honestly, we’ve been… I’ve found the most effective, things in combining both sometimes. So I usually start off with a
131 00:18:44.530 ⇒ 00:18:48.820 Venkata Prasad Krupananda: A simple narrative-style summary that highlights the…
132 00:18:49.240 ⇒ 00:19:10.030 Venkata Prasad Krupananda: key insights and, and, and, then link it to the, you know, live dashboards where, you know, stakeholder, stakeholders can drill, diver, or also, you know, dive deeper, and also into those specifics, metrics or, cohorts they’re looking for, and the balance keeps the analysis strategic and also
133 00:19:10.200 ⇒ 00:19:13.619 Venkata Prasad Krupananda: On the same page at the same time. And also.
134 00:19:13.740 ⇒ 00:19:30.889 Venkata Prasad Krupananda: It should, we make sure it aligns with our business goals and needs, like what we talked about previously, before starting off the project. We make sure it’s all aligned and it’s compliance-friendly. So yeah, these are the things that we do for deliverables.
135 00:19:31.390 ⇒ 00:19:32.400 Uttam Kumaran: Okay, okay.
136 00:19:34.700 ⇒ 00:19:40.349 Uttam Kumaran: Cool. Yeah, Robert, that’s probably all… that’s probably just the one thing coming out of this exercise I was just interested in, so…
137 00:19:40.640 ⇒ 00:19:45.809 Robert Tseng: Okay, yeah, no, that sounds good. Yeah, I guess, yeah, tell me more about…
138 00:19:45.810 ⇒ 00:20:03.249 Venkata Prasad Krupananda: exercise, I thought of… I thought of including the queries that I… the simple queries that I got online, and I did… I did take AI’s help with that. I ran a few queries on, RStudio, and… Yeah.
139 00:20:03.290 ⇒ 00:20:07.899 Venkata Prasad Krupananda: So, yeah, I did that, but I don’t know how sure I was to include that as well, so…
140 00:20:07.900 ⇒ 00:20:14.680 Robert Tseng: Okay, yeah, no, I mean, I don’t think it was that… I mean, I knew you used AI, but, like, it’s fine, like, it’s… we’re an AI company, so…
141 00:20:14.680 ⇒ 00:20:15.060 Venkata Prasad Krupananda: Yes.
142 00:20:15.060 ⇒ 00:20:21.409 Robert Tseng: It’s not like I was really, I mean, in a… I mean, sounds like you’re pretty full stack, like, I…
143 00:20:21.790 ⇒ 00:20:24.170 Robert Tseng: I don’t know tech… I don’t really have too many.
144 00:20:24.170 ⇒ 00:20:25.199 Venkata Prasad Krupananda: Yeah, tech,
145 00:20:25.200 ⇒ 00:20:27.649 Robert Tseng: reservations about your tech abilities, I’m sure.
146 00:20:27.650 ⇒ 00:20:31.430 Venkata Prasad Krupananda: I’m sure you’ll be fine. I mean, when we work with clients.
147 00:20:31.450 ⇒ 00:20:45.670 Robert Tseng: Yes, there are some situations, which I think this is part of what I want to ask you, like, one gap that we’re missing on our team is, people who feel comfortable jumping into a messy, like, situation. This is, like.
148 00:20:45.840 ⇒ 00:21:01.440 Robert Tseng: you know, maybe we jump into early-stage client situation, where no data warehouse set up yet, they just have a bunch of data kind of in random places. We’re gonna have… we just have to be very creative and scrappy, throwing things into whatever tool that you can to get something and earn the analysis out, right?
149 00:21:02.280 ⇒ 00:21:05.180 Robert Tseng: because I think for us, it’s…
150 00:21:06.080 ⇒ 00:21:21.079 Robert Tseng: I think I think the data engineering work, that obviously takes a little bit longer to kind of get up and running. I mean, once they… once we get the budget and the go-ahead for it, like, we could… we can… we could move very quickly, probably faster than most teams we’ve seen. But I think…
151 00:21:21.080 ⇒ 00:21:35.399 Robert Tseng: Yeah, they’re just… the early… getting that… those early insights out, just with what the client is willing to give us, I think that’s a… that’s… that’s a… that’s a gap that we have. So, I’d love to hear you kind of explain, you know, talk through some of your own experience.
152 00:21:35.400 ⇒ 00:21:38.000 Robert Tseng: Where, yeah, you had to just go and…
153 00:21:38.010 ⇒ 00:21:46.149 Robert Tseng: fish for the data yourself and work through, like, really messy data to get to things, without, like, even a SQL.
154 00:21:46.150 ⇒ 00:21:46.570 Venkata Prasad Krupananda: based.
155 00:21:46.570 ⇒ 00:21:48.460 Robert Tseng: environment. Yeah.
156 00:21:48.460 ⇒ 00:22:10.909 Venkata Prasad Krupananda: Yeah, so I’ve worked a lot with these kind of situations, so, there are, so yeah, so one thing that comes out of my mind was during, the time at Texas A&M in our adjacent sector, previously, so we were trying to analyze and, engagement data across multiple systems, Salesforce, Marketo, it was multiple CRM systems, and also
157 00:22:10.910 ⇒ 00:22:15.210 Venkata Prasad Krupananda: It was something that was run by the same department.
158 00:22:15.210 ⇒ 00:22:17.310 Venkata Prasad Krupananda: But different people.
159 00:22:17.310 ⇒ 00:22:42.239 Venkata Prasad Krupananda: It’s the same data, but the input style is different. So that was the problem there, and there was no centralized data warehouse as well, to clean, or the schema method was not something unique to work with. So what I did first was pull up all the raw data into RStudio, and use those simple queries to clean and transform those data as, like, you know, normalizing the
160 00:22:42.240 ⇒ 00:22:56.119 Venkata Prasad Krupananda: inconsistent data, formats, or duplicating, or, reduplicating the, records, and also, reallocating those mismatched IDs across the way, all the systems.
161 00:22:56.120 ⇒ 00:23:19.700 Venkata Prasad Krupananda: So once I had a semi-structured data set, I think I used a little bit of Python. I’m not an expert in Python, but I did take help of my teammate. I used Python Pandas to run exploratory analysis and also identify those key engagement patterns. And then, yeah, then comes the final stage of any project dashboard building. So.
162 00:23:20.040 ⇒ 00:23:33.000 Venkata Prasad Krupananda: we didn’t have a BI tool connected at that point, so I built up high-weighted SQL queries to, you know, automate that into a Power BI dashboard. So I think that is something that I can think of when it comes to
163 00:23:33.040 ⇒ 00:23:44.080 Venkata Prasad Krupananda: playing around with raw data, and then doing something from scratch, you know, because that was just data entered manually into Excel sheets.
164 00:23:44.080 ⇒ 00:23:53.229 Venkata Prasad Krupananda: So we converted that into a fully organized, structured, professional Power BI automated dashboard. So I can think of that.
165 00:23:53.420 ⇒ 00:23:58.010 Robert Tseng: Okay, great. No, I think that’s a great example, because, you know, once you deliver something like that.
166 00:23:58.060 ⇒ 00:24:02.230 Robert Tseng: You know, that allows me to go to the client and be like, look, if we can scale this up quickly.
167 00:24:02.230 ⇒ 00:24:21.309 Robert Tseng: across your team, across departments, this is… if you want something like this, let’s move it into a data warehouse, we’ll build out the actual data pipelining, we’ll give you a real BI tool. But, like, to be able to get to that point, to make the case to basically sell into a bigger service, we need to be able to, you know, at least do this, so…
168 00:24:21.310 ⇒ 00:24:39.270 Venkata Prasad Krupananda: It’s getting easier day by day. Building automated dashboards is getting easier with Google BigQuery as well, so that’s a big add-on there. Google BigQuery helps you do it without any waste of time, or without any more steps, like RStudio. It’s getting easy day by day.
169 00:24:39.640 ⇒ 00:24:47.450 Robert Tseng: Okay, great. I mean, now, I know we only have, like, maybe less than 15 minutes now, so I want to just give you some time to ask us any questions.
170 00:24:47.450 ⇒ 00:24:58.880 Venkata Prasad Krupananda: Oh, yeah, typically, just one question on the top of my mind. So typically, how would the first one month look like for me? And, would I be,
171 00:24:58.880 ⇒ 00:25:16.290 Venkata Prasad Krupananda: like, jumping right into a main project, or is it… is there a project that’s gonna start off, and then will I be included from stage one? Like, introduction, then whatever comes after that. So how does it work? Just the first month overview.
172 00:25:16.290 ⇒ 00:25:35.650 Robert Tseng: Yeah, I’ll take the first part, and then Utam can add on if he wants, but yeah, I think this is part of what we’re assessing as well with candidates. Like, can I… I mean, based on what you described, I would feel more comfortable putting you at, like, a stage… state… like, stage one, with a client. I mean, I think that’s really kind of the biggest gap for us, like, we have
173 00:25:35.780 ⇒ 00:25:44.310 Robert Tseng: folks that, like, once everything is all nice and running, like, they can come over and take over. But yeah, we just need…
174 00:25:44.310 ⇒ 00:25:56.680 Robert Tseng: We need people who are ready to go from day one. And so, you know, having somebody with, obviously, your McKinsey experience was, definitely stuck out to us, because, you know, working in some sort of consultancy, like.
175 00:25:56.680 ⇒ 00:25:57.130 Venkata Prasad Krupananda: Yes.
176 00:25:57.130 ⇒ 00:25:59.589 Robert Tseng: You kind of have to be able to
177 00:25:59.690 ⇒ 00:26:03.430 Robert Tseng: Like, be able to hit the ground running,
178 00:26:03.430 ⇒ 00:26:03.780 Venkata Prasad Krupananda: went quicker.
179 00:26:03.780 ⇒ 00:26:10.210 Robert Tseng: very little oversight, and so I think that’s definitely something we were… that caught our eye when we saw your background.
180 00:26:10.620 ⇒ 00:26:20.710 Robert Tseng: But yeah, assuming that’s the case, that yeah, we would find, I guess there’s… we could either start… I mean, I’m not entirely sure what our arrangement with you would be, but…
181 00:26:21.290 ⇒ 00:26:33.650 Robert Tseng: I think, typically, we will start part-time, and so, maybe we just put you on one client. We don’t really do 40 hours on a single client, so… but if we wanted to start you off full-time, then we would try to start you on at least two clients.
182 00:26:34.050 ⇒ 00:26:50.109 Robert Tseng: And you know, if that’s the case, they both wouldn’t be net new. It would be one net new client plus one existing client, so that you can kind of start off from day zero with a client, but then you can also kind of observe what it’s like to work within the Brainforge, like, system.
183 00:26:50.110 ⇒ 00:26:50.750 Venkata Prasad Krupananda: the team.
184 00:26:50.750 ⇒ 00:27:05.849 Robert Tseng: on a pretty established client, where you get to do more, kind of, as a secondary… as a… as a redundant secondary analyst. Maybe you’ll go over and you’ll take over that client eventually, but we would probably stagger it so you’re not only getting net new clients.
185 00:27:05.850 ⇒ 00:27:06.880 Venkata Prasad Krupananda: Okay, okay. Yeah.
186 00:27:07.230 ⇒ 00:27:07.889 Venkata Prasad Krupananda: Okay, that’s…
187 00:27:07.890 ⇒ 00:27:19.330 Uttam Kumaran: Yeah, I think… I think just a piece on that, like, we… we have a variety of clients, and we’re always growing. So, some clients that we’re expanding scope with, some clients that we’re starting, some clients that we’re sort of somewhere in the middle.
188 00:27:19.330 ⇒ 00:27:35.889 Uttam Kumaran: And so for us, it’s just, like, a little bit of risk tolerance. I would say you’re not coming on your first day and, like, okay, just go, like, own this. You know, instead, it’s more, come on, see the team, see the scope, probably spend a week or so, and, like, get your footing, and then as we, like, during the next.
189 00:27:35.890 ⇒ 00:27:42.090 Uttam Kumaran: presentation, we’ll start to loop you in and kind of give you, like, a layup, you know, to, like, take some work
190 00:27:42.090 ⇒ 00:27:47.229 Uttam Kumaran: show it to the client, and then you’ll start to go from there. I mean, we are interested in people that
191 00:27:47.230 ⇒ 00:28:03.270 Uttam Kumaran: have a desire to, like, own a relationship, drive towards finding a new scope for us to do and deliver, you know, great stuff for clients, but we also have a team. It’s not… I think the interesting thing about our company is, you know, different than a lot of consultancies, is it’s… you’re not operating in a silo, like.
192 00:28:03.270 ⇒ 00:28:15.399 Uttam Kumaran: you’ll meet everybody across both our data and AI team. We do stand-ups with sort of everybody across the different services, and so there’s a lot of, like, collaboration, even with people that you may not be on the same
193 00:28:15.400 ⇒ 00:28:26.210 Uttam Kumaran: client pod as. I mean, currently at the company, everybody is on typically, like, two to three clients right now. And so that’s just a function of our size, but also, as we start to grow accounts.
194 00:28:26.210 ⇒ 00:28:26.609 Venkata Prasad Krupananda: Of course.
195 00:28:26.610 ⇒ 00:28:34.159 Uttam Kumaran: people will start to be more fixed on maybe one or two, but we’re finding everybody’s super interested in the variety that we’re seeing.
196 00:28:34.270 ⇒ 00:28:41.850 Uttam Kumaran: But also, we’re just raising our standards for how we communicate, how we do decks, how we present work, how we do data.
197 00:28:41.850 ⇒ 00:28:45.099 Venkata Prasad Krupananda: So, yeah, that’s a little bit about, like, probably what the first.
198 00:28:45.100 ⇒ 00:28:57.860 Uttam Kumaran: like, week to 30 days is, and then I think after, like, you know, 2 to 4 weeks, given your comfort level, we would sort of bring you on to another client and start to… start to grow. And then we’re always checking in with everybody about
199 00:28:58.050 ⇒ 00:29:07.370 Uttam Kumaran: true capacity versus, like, if they’re starting to get slammed with… if a client is starting to grow and adjusting. So, we typically adjust on, like, a monthly basis, usually.
200 00:29:07.370 ⇒ 00:29:07.989 Venkata Prasad Krupananda: Thank you.
201 00:29:07.990 ⇒ 00:29:09.229 Uttam Kumaran: Where people are assigned.
202 00:29:09.440 ⇒ 00:29:22.930 Venkata Prasad Krupananda: Okay, okay, perfect. Good. Yeah, so I don’t have any, important questions as of now. That was one thing that I really wanted to know, how the first month looked like. But yeah, other than that, I think I’m good.
203 00:29:23.430 ⇒ 00:29:34.910 Robert Tseng: Okay. Yeah, tell me more about, I mean, your recruiting, like, kind of what… where are you at in your process, like, what are your expectations for, like, I guess, yeah, if we were to move forward with you, what would that look like from your side?
204 00:29:35.030 ⇒ 00:29:41.160 Venkata Prasad Krupananda: Like, right now, from the dates of my availability, or is it just where I’m.
205 00:29:41.160 ⇒ 00:29:47.990 Robert Tseng: Yeah, your availability, like, I mean, are you talking to other… I mean, you’re probably talking to other companies, so, like, you wanted to know, yeah.
206 00:29:47.990 ⇒ 00:30:02.889 Venkata Prasad Krupananda: Yes, I am… I am talking to other companies, I am interviewing as well, but keeping all that aside, this role is more matching with my background and my skills, so based on that, I think, my availability would be
207 00:30:02.890 ⇒ 00:30:17.119 Venkata Prasad Krupananda: by the end of December, because I’m wrapping up things with my last project, and it should have been done by this week, but next week, people are all off. They’re just gone in next week, and then, because of holidays, I think it’s getting delayed. So, by…
208 00:30:17.120 ⇒ 00:30:21.230 Venkata Prasad Krupananda: by 28th, Sunday, I think I should be good.
209 00:30:21.810 ⇒ 00:30:22.210 Robert Tseng: Okay.
210 00:30:22.210 ⇒ 00:30:25.510 Venkata Prasad Krupananda: Yes, so by… I think from Jan 1st, yes, I’ll be available.
211 00:30:25.800 ⇒ 00:30:31.050 Robert Tseng: Okay, yeah, I mean, that timing works well with us. Yeah.
212 00:30:31.150 ⇒ 00:30:32.239 Robert Tseng: I guess.
213 00:30:32.240 ⇒ 00:30:34.739 Venkata Prasad Krupananda: liability, and,
214 00:30:35.450 ⇒ 00:30:40.920 Venkata Prasad Krupananda: Yeah, when it comes to the process, are there gonna be any more rounds of interviews for me here?
215 00:30:40.960 ⇒ 00:30:41.730 Robert Tseng: Nope.
216 00:30:41.730 ⇒ 00:30:43.339 Uttam Kumaran: This would basically be… yeah.
217 00:30:43.780 ⇒ 00:30:47.919 Uttam Kumaran: This is basically it. I mean, happy to introduce you to more people if you’d like to… to chat more, but…
218 00:30:48.110 ⇒ 00:30:59.079 Uttam Kumaran: I feel like, yeah, this is usually our process. And for everybody, we tend to start small and sort of give you the opportunity to come on and do a couple things, meet us, and then we sort of grow from there.
219 00:30:59.080 ⇒ 00:31:11.190 Venkata Prasad Krupananda: Yeah, so yeah, Utam, yeah, if you think… if there’s somebody else that I can talk with, which can be useful for me and for them as well as an introductory, yeah, just go ahead and let me know. I’ll be available.
220 00:31:11.650 ⇒ 00:31:16.520 Venkata Prasad Krupananda: So yeah, next week I’ll be available, yes. Next week is… I’m fully available, so I can talk to anyone.
221 00:31:17.090 ⇒ 00:31:36.130 Uttam Kumaran: Okay, okay. So then, yeah, I think on our side, I’ll go ahead and start, you know, discussing with you over email about, like, contract process, let you kind of see a little bit about, of what that looks like from our side. And yeah, I do have one person I think you should chat with, just to say hi and see a different part of the business, like, on the AI side.
222 00:31:36.180 ⇒ 00:31:48.169 Uttam Kumaran: And then, yeah, I mean, I think overall, like, I feel like we’re… we’d be excited to work together. As I mentioned, like, we would… we typically start, you know, sort of in, like, a part-time capacity, but
223 00:31:48.170 ⇒ 00:32:01.139 Uttam Kumaran: we’ve had people start, like, in 3 days, we’re like, okay, like, let’s just, like, ramp up fully. So, we’re… this is not… this isn’t, like, gonna be, any way similar to any of the other,
224 00:32:01.560 ⇒ 00:32:08.119 Uttam Kumaran: normal places that you may be interviewing, in that we try to do things differently, not to just be weird.
225 00:32:08.120 ⇒ 00:32:08.470 Venkata Prasad Krupananda: Yeah, you bet.
226 00:32:08.470 ⇒ 00:32:24.340 Uttam Kumaran: like, that’s what’s really helped us grow and build, like, an awesome, awesome team, you know? So, in no way are we like, oh, just start part-time, and then we sort of drag things along. In fact, it’s mainly just to give everybody a sense of what it is like, and for you to
227 00:32:24.810 ⇒ 00:32:30.590 Uttam Kumaran: for you to also know that this is where you want to be. It’s a two-way relationship, right? So, that’s probably…
228 00:32:31.030 ⇒ 00:32:31.509 Venkata Prasad Krupananda: That’s what you do.
229 00:32:31.510 ⇒ 00:32:31.910 Uttam Kumaran: stay there.
230 00:32:31.910 ⇒ 00:32:35.209 Venkata Prasad Krupananda: That’s a good approach, though. I totally agree with that.
231 00:32:35.800 ⇒ 00:32:38.660 Uttam Kumaran: Perfect. Cool. Okay, great.
232 00:32:38.660 ⇒ 00:32:40.960 Venkata Prasad Krupananda: Yeah, thank you for your guys, and…
233 00:32:40.960 ⇒ 00:32:45.869 Uttam Kumaran: Yeah, I appreciate the time taking to do this, and yeah, feel free to email me with.
234 00:32:45.870 ⇒ 00:32:46.500 Venkata Prasad Krupananda: With any other bike.
235 00:32:47.050 ⇒ 00:32:55.599 Venkata Prasad Krupananda: Yeah, definitely. Yeah, perfect. And yeah, it was a great piece of conversation. And also, previously, I spoke to, Navesh Kumar, and then I spoke
236 00:32:55.790 ⇒ 00:32:56.850 Venkata Prasad Krupananda: who,
237 00:32:56.850 ⇒ 00:32:57.250 Uttam Kumaran: Amber?
238 00:32:57.250 ⇒ 00:32:58.060 Venkata Prasad Krupananda: looking at.
239 00:32:59.210 ⇒ 00:32:59.860 Uttam Kumaran: Amber?
240 00:32:59.860 ⇒ 00:33:02.740 Venkata Prasad Krupananda: Amber, yeah. So yeah, that was really good.
241 00:33:02.740 ⇒ 00:33:03.530 Uttam Kumaran: What was that?
242 00:33:03.560 ⇒ 00:33:13.659 Venkata Prasad Krupananda: Oh yeah, that went really good. We spoke a lot of product analytics, and what I did previously, and what she’s doing right now, and… Right. So it was a great piece of.
243 00:33:13.660 ⇒ 00:33:20.059 Uttam Kumaran: Yeah, she’s owning a lot of analysis work, and I think she will hopefully be handing some stuff off to you.
244 00:33:21.750 ⇒ 00:33:26.299 Uttam Kumaran: You know, which would be great, because she could… she deserves a little bit of a break, so…
245 00:33:26.300 ⇒ 00:33:32.550 Venkata Prasad Krupananda: Yeah, it sounded like she was on, you know, she was loaded with many things.
246 00:33:32.550 ⇒ 00:33:48.040 Uttam Kumaran: Yeah, it’s just… it’s just tricky in our business, like, as we start to do well, clients start asking for stuff, and we don’t really particularly like to hire… we don’t like to hire to fit, like, one thing we need. Like, when we hire, what we’re looking for is the aptitude to, like, learn.
247 00:33:48.040 ⇒ 00:33:53.199 Uttam Kumaran: And learn the next thing, because in 6 months, we may be doing different things, and so…
248 00:33:53.640 ⇒ 00:34:17.650 Uttam Kumaran: we’re looking for smart data people that can adapt, you know, and so Amber is one of the stars on our team that has shown a lot of, you know, that she’s able to adapt to the next challenge, but we now… because of her work now, we’ve started selling it to more people, and, like, there’s just so much analysis work. The other thing I’m very excited about is our analytics and, like, analysis team is very small. Like, our data team’s big.
249 00:34:17.739 ⇒ 00:34:34.179 Uttam Kumaran: But we have people across data engineering, analytics engineering, and, you know, in sort of BI strategy analysis. But that team that is sort of at the high level has really been mainly Robert and Amber so far. And so we’re looking to try to, like, build an actual pod of people that think about
250 00:34:34.179 ⇒ 00:34:38.799 Uttam Kumaran: not only, like, delivering for a client, but, like, how does Brainport do analysis work?
251 00:34:38.800 ⇒ 00:34:49.789 Uttam Kumaran: How do we do A-B testing? Like, what are our playbooks? And as a group, like, how do we think about that service? You know, and so that’s, like, I’m very, very excited to be able to, you know, add to that team, so…
252 00:34:50.409 ⇒ 00:34:54.719 Venkata Prasad Krupananda: Okay, wonderful then. Yeah, so I think, I think we’re done now.
253 00:34:55.190 ⇒ 00:34:56.790 Uttam Kumaran: Perfect. Okay.
254 00:34:56.790 ⇒ 00:34:57.530 Venkata Prasad Krupananda: Alright.
255 00:34:57.700 ⇒ 00:35:03.359 Venkata Prasad Krupananda: Yeah, thank you again, and then, yeah, I’ll definitely reach out if I have any questions, and yeah, I think I’m good for now.
256 00:35:03.550 ⇒ 00:35:04.510 Uttam Kumaran: Okay, alright.
257 00:35:04.510 ⇒ 00:35:05.680 Robert Tseng: Sounds good. Thank you both.
258 00:35:05.680 ⇒ 00:35:07.389 Venkata Prasad Krupananda: Thanks again, guys. Bye-bye.