Meeting Title: Brainforge Data Engineering Interview Date: 2026-02-11 Meeting participants: Awaish Kumar, Alex Wilson


WEBVTT

1 00:00:07.460 00:00:08.280 Awaish Kumar: Hi.

2 00:00:09.260 00:00:10.310 Alex Wilson: Hello!

3 00:00:10.680 00:00:11.689 Alex Wilson: How are you?

4 00:00:12.080 00:00:13.509 Awaish Kumar: I’m good, how about you?

5 00:00:13.990 00:00:15.440 Alex Wilson: I am doing alright.

6 00:00:16.460 00:00:17.710 Awaish Kumar: How was the day?

7 00:00:19.790 00:00:20.589 Alex Wilson: Say again?

8 00:00:20.990 00:00:22.539 Awaish Kumar: How was your day?

9 00:00:23.180 00:00:26.150 Alex Wilson: Oh, it’s really good so far.

10 00:00:26.150 00:00:30.930 Awaish Kumar: Okay, great. Great to hear. So, yeah, I will just introduce myself.

11 00:00:32.060 00:00:34.859 Awaish Kumar: And a little bit about brain falls.

12 00:00:35.840 00:00:38.910 Awaish Kumar: And then, yeah, we can start with your introduction, and…

13 00:00:39.220 00:00:42.210 Awaish Kumar: We can learn more about your background.

14 00:00:42.890 00:00:43.660 Awaish Kumar: Okay.

15 00:00:43.840 00:00:50.950 Awaish Kumar: My name is Avish Kumar, and I am basically leading the data engineering services at Brainforge.

16 00:00:51.140 00:00:55.570 Awaish Kumar: I have, like, around 10 years of experience doing data engineering work.

17 00:00:56.150 00:01:00.760 Awaish Kumar: And, Yeah, and

18 00:01:01.350 00:01:07.449 Awaish Kumar: Working with growth-stage companies, other startups, and help them build the data foundations.

19 00:01:09.050 00:01:15.369 Awaish Kumar: And when it comes to the Brainforge, Brainforge basically is an AI and data consultancy service.

20 00:01:15.590 00:01:18.600 Awaish Kumar: Which operates, remotely.

21 00:01:19.210 00:01:23.960 Awaish Kumar: So, everybody here, it works, from across the world.

22 00:01:24.360 00:01:30.260 Awaish Kumar: And, yeah, the brain force works, you know, like, either full-time or in a contract.

23 00:01:30.650 00:01:34.759 Awaish Kumar: contracting, role, and,

24 00:01:35.220 00:01:41.449 Awaish Kumar: Yeah, apart from that, basically, yeah, like, mostly we have the…

25 00:01:41.780 00:01:45.909 Awaish Kumar: we have our clients from the, like, the US, and…

26 00:01:46.180 00:01:51.619 Awaish Kumar: We have medium to large-scale companies that we work for.

27 00:01:52.070 00:01:57.119 Awaish Kumar: That’s mainly about brain food, okay, and yeah, now let’s…

28 00:01:57.320 00:02:01.270 Awaish Kumar: Let’s talk about you, like, introduce… can you please introduce yourself?

29 00:02:01.270 00:02:01.870 Alex Wilson: Oh.

30 00:02:02.030 00:02:11.559 Alex Wilson: I am Alex Wilson. My background was mainly in banking. I worked at Wells Fargo for the last 23 years.

31 00:02:11.710 00:02:20.519 Alex Wilson: They are reorganizing their footprint, so they removed all the work they did in Colorado.

32 00:02:20.770 00:02:30.380 Alex Wilson: So, I was let go in April of last year, so I’m a bit rusty on technical talk and whatnot, because I haven’t really hung out with a lot of those folks over the last year.

33 00:02:30.580 00:02:36.600 Alex Wilson: But my main, role there was,

34 00:02:37.140 00:02:48.910 Alex Wilson: very much in the data, so… dealt with many different databases, a lot of ETL, I built an automation system to kind of control it. We had way too much work.

35 00:02:49.150 00:02:59.540 Alex Wilson: And dealing with fraud, I know the banks kind of move really slow, they’re still working on mainframe and whatnot, but with fraud, we had to move extremely fast, so…

36 00:02:59.910 00:03:01.719 Alex Wilson: That’s my background.

37 00:03:03.000 00:03:06.700 Awaish Kumar: Okay, yeah, so I would love to understand, like, if you can…

38 00:03:07.050 00:03:10.839 Awaish Kumar: Give me an example of a project where

39 00:03:11.250 00:03:14.440 Awaish Kumar: You did the work, like, end-to-end.

40 00:03:15.360 00:03:16.689 Awaish Kumar: That would be nice.

41 00:03:17.840 00:03:26.279 Alex Wilson: Project where I worked into, and one of the big things, so we did a lot with… with it being fraud, we did a lot with account takeover.

42 00:03:26.430 00:03:30.660 Alex Wilson: So, one of the keys to that was trying to figure out,

43 00:03:32.230 00:03:42.189 Alex Wilson: data that would be useful to use to identify that. So, there’s a lot of account takeover that happens with that. So we found a third-party tool

44 00:03:42.340 00:04:00.850 Alex Wilson: To… we realized that one of the methods of taking over accounts was, impersonating the customer, taking over, you know, their address, or their email, or their phone number. So we found a tool that actually gave a score to address changes. So if it was moving from a nice

45 00:04:01.470 00:04:14.349 Alex Wilson: ritzy area and moving to a warehouse, that’s suspicious, so it would get a high score. So the, the actual fraud team could actually investigate and call the customer and deal with stuff like that, so…

46 00:04:14.350 00:04:24.410 Alex Wilson: We then took that data to match it up with all kinds of other data, like if, there is a mid-range score, and…

47 00:04:24.450 00:04:36.310 Alex Wilson: There was also check orders, or we also got suspicious phone calls, and we matched up to all kinds of different banking data to help different teams and… across the bank.

48 00:04:36.620 00:04:41.750 Alex Wilson: So that was my… My project from start to finish.

49 00:04:42.730 00:04:51.020 Awaish Kumar: Yeah, I just want to understand, like, in that project, what exactly you did, what tools were used.

50 00:04:51.460 00:04:58.909 Awaish Kumar: Right? And… what was the process? How you plan… like, how you planned it, and how you executed it?

51 00:04:59.190 00:05:05.769 Awaish Kumar: And what, like, what exactly you executed, or was it, like, the team effort, or whatever?

52 00:05:06.350 00:05:19.499 Alex Wilson: It was definitely a team effort, so, that was within my automation, which helped out a lot, because with the automation, we were allowed to take different code samples and whatnot, and just toss it in and have it running.

53 00:05:19.580 00:05:30.830 Alex Wilson: Really quick, but the way we did that was meeting with the teams, so we had fraud servicing teams that were our experts on dealing with the customers and what the fraud looked like. So…

54 00:05:31.150 00:05:44.590 Alex Wilson: I would meet with them and get a full idea of, like, okay, what does the fraud look like? What kind of data can we tie into that? You know, the investigation. And then, we start to build some prototypes and give them some

55 00:05:44.880 00:05:48.300 Alex Wilson: Sample, reports to work.

56 00:05:48.540 00:06:01.340 Alex Wilson: And look at the data, and then report back to me to let me know, like, what’s working, what wasn’t working, and just reiterate and continue until we had a very valid product that was working really good for the teams.

57 00:06:02.680 00:06:10.570 Awaish Kumar: Yeah, like, my question still is the same, like, what tools were used, where data was coming from, what was the source.

58 00:06:11.680 00:06:19.700 Awaish Kumar: What pipeline looked like, how were it… it was working, how it was automated, what, like… like, things like that.

59 00:06:20.270 00:06:31.089 Alex Wilson: Okay, so our main system that we used was SaaS, and I know that sounds weird to use for automation, but it had access to all the different databases. With a bank that size.

60 00:06:31.090 00:06:40.660 Alex Wilson: we had data coming from many, many, many different systems of record. So, the main home equity data, which was the data that I worked with, was DB2.

61 00:06:40.950 00:06:45.069 Alex Wilson: We also had Oracle for some,

62 00:06:45.300 00:06:54.349 Alex Wilson: internal data that they… was built by a different team. We used SAS data sets. We also had data coming from

63 00:06:54.650 00:06:57.110 Alex Wilson: Teradata, and…

64 00:06:57.990 00:07:05.359 Alex Wilson: Microsoft SQL Server, which were other internal databases. So, it came from a lot of different, data sources.

65 00:07:07.150 00:07:25.429 Alex Wilson: we stored the data within that SaaS system, would go into SaaS data sets, was how we worked with it once it was within SAS. In that automation, I had to build all the orchestration along with it, so there was dependencies, you know.

66 00:07:25.480 00:07:35.249 Alex Wilson: You didn’t want to run a report on something where the data wasn’t available yet, so there were dependency checks for all the different systems of record that were coming in.

67 00:07:36.220 00:07:47.279 Alex Wilson: We also had time, what time of day, you know, the people were working and whatnot. A lot of those processes we would run overnight because it was automated, so we could have it by the time the employees came into work.

68 00:07:47.280 00:07:49.030 Awaish Kumar: So, all in SaaS.

69 00:07:50.210 00:07:57.909 Awaish Kumar: you were integrating to those sources through SAS, all the pipelines, like, the scheduling happens.

70 00:07:58.020 00:07:58.840 Awaish Kumar: There.

71 00:07:58.840 00:07:59.480 Alex Wilson: Okay.

72 00:07:59.900 00:08:07.090 Alex Wilson: Yes. And there was a couple different ways we did that. There was, we did have, .

73 00:08:09.480 00:08:17.780 Awaish Kumar: Okay, okay, I just have another question for you, like, while, you, like, you mentioned you were laid off, like.

74 00:08:18.140 00:08:22.059 Awaish Kumar: since April. So, like, have…

75 00:08:23.000 00:08:31.349 Awaish Kumar: Like, what have you tried, like, in the period after that? Like, if you have worked on any projects, any tools, anything?

76 00:08:32.020 00:08:44.149 Alex Wilson: So, I started by trying to… I’ve applied to a ton of jobs, so I’ve been doing interviews and just applying to a lot of work. In the meantime, while doing that, I’ve been working a lot with AI.

77 00:08:44.490 00:08:52.760 Alex Wilson: And I kind of was interested in the other position you had, the… Ai and automation.

78 00:08:53.340 00:09:00.139 Alex Wilson: Engineer. I’ve been doing a lot of work with that. I’ve gotten pretty good at, like, building full-stack applications.

79 00:09:00.280 00:09:08.520 Alex Wilson: And being able to maintain them and actually have them out. So my portfolio site is actually built… I built a site engine.

80 00:09:08.670 00:09:11.090 Alex Wilson: That actually can… I can…

81 00:09:11.250 00:09:18.949 Alex Wilson: add a domain directly into it, and then create websites on the fly, kind of like a, Bossinger, or…

82 00:09:19.130 00:09:21.390 Alex Wilson: Squarespace.

83 00:09:21.550 00:09:27.170 Alex Wilson: But it’s kind of purpose-built for kind of how I build. It’s built on Next.js, and…

84 00:09:27.170 00:09:28.210 Awaish Kumar: Well, that’s being…

85 00:09:28.210 00:09:28.929 Alex Wilson: We’re based?

86 00:09:30.390 00:09:31.260 Alex Wilson: Yes.

87 00:09:31.990 00:09:34.420 Awaish Kumar: I do. For the last year.

88 00:09:34.980 00:09:37.339 Awaish Kumar: Yeah, what you have tried.

89 00:09:37.750 00:09:40.640 Awaish Kumar: Whatever you have tried, like… Can you name.

90 00:09:40.640 00:09:48.650 Alex Wilson: So, the main one I’m using for that is, right now, Cloud Code. I also have done some work with, with,

91 00:09:49.220 00:09:56.669 Alex Wilson: Rag with, Gemini. I found Gemini and cloud services to be the easiest to work with that.

92 00:09:57.260 00:09:59.930 Alex Wilson: And then, of course, ChatGPT.

93 00:10:01.350 00:10:06.860 Awaish Kumar: Yeah, like, in the chat GPT, Cloud Code, there are multiple… models.

94 00:10:07.030 00:10:07.560 Awaish Kumar: Alright.

95 00:10:07.560 00:10:08.070 Alex Wilson: Yes.

96 00:10:08.350 00:10:16.649 Alex Wilson: And then also some… I have played with, like, Olama and… what is it, LMS, for local, so Meta.

97 00:10:16.800 00:10:19.379 Alex Wilson: I’ve worked with, just to experiment with.

98 00:10:21.430 00:10:31.410 Awaish Kumar: Okay, Yeah, so one more… yeah, like, actually, I have… I’m having this interview because…

99 00:10:31.760 00:10:34.319 Awaish Kumar: If you apply for some of the data roles, right?

100 00:10:34.610 00:10:40.790 Awaish Kumar: so, I would like to understand your interests, like, where you would like to

101 00:10:40.940 00:10:48.370 Awaish Kumar: Of where you would like to see yourself, like, in the data team, or in a AI team?

102 00:10:49.800 00:10:53.250 Alex Wilson: I kind of see the two going, hand in hand.

103 00:10:53.350 00:10:59.529 Alex Wilson: I don’t really believe you can have great AI without, good data pipelines.

104 00:10:59.980 00:11:05.660 Alex Wilson: But I am fascinated with AI, and I think it is…

105 00:11:05.660 00:11:14.259 Awaish Kumar: Depends… yeah, I want to understand what you’re looking for in your next role, because I… obviously, these both are linked together.

106 00:11:14.400 00:11:20.620 Awaish Kumar: But we have people who work with data, and we have people who work with AI, so…

107 00:11:21.120 00:11:26.739 Awaish Kumar: if you need data help, we just call on each other, like, if we need AI help or data help.

108 00:11:26.820 00:11:44.160 Awaish Kumar: So, I want to understand your interest, like, what you are looking for in your next role. If you want to be doing data, like, querying, doing SQL, query the data, do analysis, you can obviously use AI to speed up your workflows, okay? You can ask AI to write queries for you.

109 00:11:44.430 00:11:50.000 Awaish Kumar: That, but that is still a data work, like, you will be still doing analysis, looking at the data.

110 00:11:50.110 00:11:57.890 Awaish Kumar: Analyzing patterns, or maybe creating a report, or… Maybe we’ll… or even…

111 00:11:58.090 00:12:05.170 Awaish Kumar: Or maybe, like, building a, like, a model, a SQL model, which basically transforms the data.

112 00:12:05.830 00:12:13.930 Awaish Kumar: Or, like, is an AI, which you can just use whatever their team have worked on, And you just…

113 00:12:14.050 00:12:16.590 Awaish Kumar: Plug in and grab the data.

114 00:12:16.750 00:12:19.479 Awaish Kumar: And bill features on top of it, so…

115 00:12:20.290 00:12:23.290 Awaish Kumar: Like, what… where do you look for yourself?

116 00:12:24.750 00:12:31.980 Alex Wilson: Right now, I’m really kind of looking to transition, so I would really like to work in AI, but my expertise is definitely in data.

117 00:12:32.140 00:12:33.460 Awaish Kumar: Okay. So…

118 00:12:33.460 00:12:39.870 Alex Wilson: And talking to, Jody, Jody was the one that recommended this job. I guess he’s friends with

119 00:12:40.440 00:12:41.760 Alex Wilson: Upta?

120 00:12:42.160 00:12:43.550 Alex Wilson: Is that how you pronounce it?

121 00:12:45.030 00:12:46.440 Awaish Kumar: Sorry, who, Utong?

122 00:12:47.120 00:12:48.790 Alex Wilson: Utam? I’m sorry, Utam?

123 00:12:49.670 00:12:57.209 Alex Wilson: He’s the one that recommended, applying to this, and he recommended that I go, the direction of data.

124 00:12:57.710 00:13:02.870 Alex Wilson: So, that is my expertise, that’s what I’ve been doing for years, but I do really want to transition.

125 00:13:05.610 00:13:18.079 Alex Wilson: I kind of really feel like what I’m seeing from AI and whatnot, as I explore it, the stuff that I did historically, I know I could make a lot easier now. And I’m comfortable working with the data.

126 00:13:18.230 00:13:21.940 Alex Wilson: And I’m comfortable working with data with AI.

127 00:13:23.050 00:13:26.299 Awaish Kumar: In the data also, like, are you looking…

128 00:13:26.500 00:13:35.559 Awaish Kumar: More like being a data engineer, or a… Data analyst, or analytics engineer.

129 00:13:37.530 00:13:39.660 Alex Wilson: Probably a data engineer.

130 00:13:40.350 00:13:41.200 Awaish Kumar: Okay.

131 00:13:41.790 00:13:49.900 Awaish Kumar: And… so… Then I would like… To know, like, apart from The project you described.

132 00:13:50.230 00:13:52.530 Awaish Kumar: That was based on SARS.

133 00:13:52.950 00:13:58.389 Awaish Kumar: Are there any other tools, or… You have worked with, or…

134 00:13:59.700 00:14:02.130 Awaish Kumar: Any programming languages you have worked with?

135 00:14:02.880 00:14:06.539 Alex Wilson: the newer, like, you’re talking newer, like, Snowflake, and…

136 00:14:06.820 00:14:08.759 Alex Wilson: Airflow, that kind of stuff, or…

137 00:14:09.320 00:14:15.500 Awaish Kumar: Yeah, like, Snowflake is a data warehouse, but, like, you can mention any tools you have used.

138 00:14:16.650 00:14:20.380 Awaish Kumar: for For your data engineering projects.

139 00:14:20.930 00:14:24.320 Awaish Kumar: Apart from SaaS, which we rarely use.

140 00:14:24.750 00:14:25.580 Alex Wilson: Right.

141 00:14:26.340 00:14:43.959 Alex Wilson: That was primarily what we used. We did use it to do a lot of… actually, a lot of it was SQL on the back end to do a lot of the ETL that we were doing. So there were some databases, some Oracle databases that were built by people that didn’t really know

142 00:14:44.010 00:14:47.460 Alex Wilson: How to engineer data and whatnot, so we would have to do…

143 00:14:47.610 00:14:58.089 Alex Wilson: tons of, joining and whatnot in order to get just a basic data set. So we did a ton of ETL in order for that data to work well within the system.

144 00:14:58.290 00:14:59.980 Alex Wilson: So that’s…

145 00:15:00.090 00:15:10.289 Alex Wilson: why I include that. Just working with all those different sorts of databases, translating that data into usable forms within

146 00:15:10.790 00:15:12.300 Alex Wilson: the bigger picture.

147 00:15:12.400 00:15:13.530 Awaish Kumar: So we could…

148 00:15:13.530 00:15:19.639 Alex Wilson: Link phone calls to address changes to transactions.

149 00:15:19.900 00:15:25.190 Alex Wilson: There was tons of different data that we would link to within the fraud space.

150 00:15:27.360 00:15:39.030 Awaish Kumar: I understand, but I… like, you… you work with data which is in Oracle, maybe not formatted the way you want it, so how did you change it? Like, were you writing SQL?

151 00:15:39.160 00:15:41.370 Awaish Kumar: For transformation, we’re using.

152 00:15:41.370 00:15:42.050 Alex Wilson: No.

153 00:15:42.050 00:15:50.130 Awaish Kumar: I don’t know how were you using SAS, like, was… are you scheduling SQL on SAS, or it’s more like… was it more like,

154 00:15:50.530 00:15:55.180 Awaish Kumar: Microsoft, SPSS, so where you just drag and drop.

155 00:15:55.780 00:15:58.139 Awaish Kumar: Transformation, so how was that?

156 00:15:59.070 00:16:11.579 Alex Wilson: We had one database that actually was an ATESP, that was completely on Microsoft SQL Server. For that, we would use SSRS and SSIS.

157 00:16:12.260 00:16:14.629 Alex Wilson: A lot of, like you said,

158 00:16:14.990 00:16:19.969 Alex Wilson: Point, or click and drag, we did use some SAS,

159 00:16:20.040 00:16:35.850 Alex Wilson: For a lot of the analytics that we would do on top of, but for the most part, the translation we were doing was SQL, like, massive queries, and we would do step-by-step, of course, you can do data steps in SAS, so you could do one step, two-step, or you could do…

160 00:16:35.850 00:16:41.630 Alex Wilson: the width clauses within your SQL, which I would do a lot. I was primarily a SQL person.

161 00:16:41.990 00:16:49.319 Alex Wilson: My background was heavily in SQL. The team I was on had SaaS, so a lot of the SaaS I did was proc SQL.

162 00:16:50.100 00:16:51.070 Alex Wilson: You know…

163 00:16:51.070 00:16:52.270 Awaish Kumar: Just steps.

164 00:16:54.220 00:16:57.740 Awaish Kumar: So, how to deal with data in, like…

165 00:16:58.750 00:17:02.920 Awaish Kumar: from various sources, like the files from FTP server.

166 00:17:03.800 00:17:09.659 Awaish Kumar: S3, like, the files are scattered around, the data is scattered.

167 00:17:09.790 00:17:10.480 Awaish Kumar: Oh.

168 00:17:10.640 00:17:14.679 Awaish Kumar: like, were you using similar tools like SSIS to…

169 00:17:15.050 00:17:18.349 Awaish Kumar: fully through some connector to an FTP server, or…

170 00:17:19.890 00:17:33.480 Alex Wilson: We had a few systems that needed that, and some of those used, like, flat files and whatnot to transfer. The original address changes we had were straight flat files that we would have to bring over, through FTP,

171 00:17:33.720 00:17:39.380 Alex Wilson: And eventually changed that over to, What’s the tool called?

172 00:17:40.940 00:17:42.720 Alex Wilson: I forget what the tool was called.

173 00:17:48.250 00:17:53.679 Alex Wilson: I can’t remember what the tool was, but we had to build a whole process for that, for moving those files, and…

174 00:17:53.680 00:17:55.589 Awaish Kumar: Then extracting it.

175 00:17:55.590 00:17:56.820 Alex Wilson: the flat files.

176 00:17:58.410 00:18:03.600 Awaish Kumar: So, how would you handle, disagreements,

177 00:18:04.220 00:18:09.080 Awaish Kumar: For… if you’re working on a… Data pipeline and,

178 00:18:10.370 00:18:19.040 Awaish Kumar: And if you have disagreement with your… For example… You can see… Colleagues, or, like.

179 00:18:19.370 00:18:22.370 Awaish Kumar: So, how you architecture it, or…

180 00:18:22.770 00:18:26.899 Awaish Kumar: Architect it, or how would you execute it, like.

181 00:18:27.510 00:18:32.000 Awaish Kumar: If there are disagreements, what is your way to handle that?

182 00:18:33.110 00:18:35.110 Alex Wilson: Typically, whenever I…

183 00:18:35.670 00:18:43.199 Alex Wilson: come across, disagreements, the first thing I try to do is just understand where the concern’s coming from.

184 00:18:44.430 00:18:48.549 Alex Wilson: I feel like a lot of times disagreements come about from…

185 00:18:49.220 00:18:56.140 Alex Wilson: making a stand versus trying to understand, right? So, the first thing I try to do is just understand, and…

186 00:18:56.250 00:18:59.980 Alex Wilson: take it from the approach that maybe I’m not fully hearing what…

187 00:19:00.300 00:19:07.029 Alex Wilson: They’re saying maybe they’re not fully hearing what I’m saying, and maybe my approach is not very good.

188 00:19:07.260 00:19:14.100 Alex Wilson: You know, sometimes that can be an issue if I’m, you know, bullheaded and saying, we have to do this this way.

189 00:19:14.260 00:19:17.210 Alex Wilson: Perhaps I should just take a beat and listen and…

190 00:19:17.570 00:19:21.370 Alex Wilson: Usually I’ve found with a lot of data,

191 00:19:22.260 00:19:41.759 Alex Wilson: if you have access to it and whatnot, and it’s an issue, a disagreement on how something should be done, it’s usually not that hard to test it. It’s usually not hard to actually run a process and just say, okay, this is what we get if we do it this way. Does this look alright? What are the issues? And point out, like.

192 00:19:42.160 00:19:47.320 Alex Wilson: here is an issue I find with doing it this way, this is what We might want to avoid?

193 00:19:48.480 00:20:03.290 Awaish Kumar: Okay, my last question is that, for example, if you join us, and then we tie you to some of… one of our clients, for example, if there’s a new client which comes in, there is,

194 00:20:03.710 00:20:07.070 Awaish Kumar: There is… data is scattered across different sources.

195 00:20:07.290 00:20:08.650 Awaish Kumar: It looks messy.

196 00:20:08.810 00:20:10.270 Awaish Kumar: are not advised.

197 00:20:10.470 00:20:15.170 Awaish Kumar: They don’t have the right definitions yet.

198 00:20:15.780 00:20:18.980 Awaish Kumar: So how would you then… approach.

199 00:20:19.100 00:20:23.170 Awaish Kumar: Working for that client, how would you… Planet.

200 00:20:24.630 00:20:34.260 Alex Wilson: Do you have, like, a current, like, a tech stack that you’re using, or any standardizations across, or…

201 00:20:34.260 00:20:37.270 Awaish Kumar: No, we don’t. Like, we…

202 00:20:37.870 00:20:52.200 Awaish Kumar: Like, obviously, we have different tax stacks for different clients, and each client is really different, so based on their use case, we recommend what to use and what not to use.

203 00:20:53.010 00:20:57.789 Awaish Kumar: Unless they have… Any fixed, like,

204 00:20:59.560 00:21:03.529 Awaish Kumar: culture of using something, or, like, if they are,

205 00:21:03.910 00:21:11.369 Awaish Kumar: tied in, like, in a contract or something. We try to push towards best possible tools and technologies.

206 00:21:11.750 00:21:16.530 Awaish Kumar: But now it’s more of a question for you if you come on.

207 00:21:17.320 00:21:20.300 Awaish Kumar: Or for a… working for a client, which is…

208 00:21:20.560 00:21:28.250 Awaish Kumar: Data is scattered across different sources, it’s messy, it is unorganized, not clean, no documentation.

209 00:21:29.090 00:21:36.029 Awaish Kumar: What, like, what will be your process? Like, what will you do in the first 30 days for that client?

210 00:21:37.110 00:21:45.799 Alex Wilson: The very first thing that I would do is try to, lock down, what their system of record is.

211 00:21:46.060 00:21:48.350 Alex Wilson: What the main source they believe

212 00:21:49.380 00:21:59.629 Alex Wilson: is their main source of records, and then, obviously, with the other records, start trying to figure out what they have in common, how you can actually map that out.

213 00:21:59.820 00:22:07.000 Alex Wilson: And start building a data dictionary for what the stakeholders can actually use.

214 00:22:07.250 00:22:11.329 Alex Wilson: That we could deliver.

215 00:22:11.430 00:22:25.490 Alex Wilson: That would be the very first step, and hopefully we have a good working relationship with them so we can actually have those conversations, and it’s not just a, here, do it for us, we’re paying you to do all the work.

216 00:22:25.780 00:22:32.200 Alex Wilson: I’d much rather work with teams and, you know, fully understand.

217 00:22:34.190 00:22:44.940 Awaish Kumar: And then… What if… Your client pushes you to do something, Like, will you push back?

218 00:22:45.050 00:22:52.080 Awaish Kumar: They can see… Have you ever pushed back on your stakeholders or your clients?

219 00:22:53.680 00:23:03.319 Alex Wilson: I have, respectfully. I don’t think I’m a very, confrontational type of personality.

220 00:23:03.430 00:23:07.200 Alex Wilson: I’m usually pretty calm, so… but I will…

221 00:23:07.620 00:23:12.190 Alex Wilson: push back if something is wrong or dangerous about fraud. We had…

222 00:23:12.470 00:23:17.279 Alex Wilson: Had a lot of, stuff that could be potentially bad, like,

223 00:23:17.710 00:23:25.820 Alex Wilson: Sharing data with third parties and whatnot that shouldn’t be shared, or, you know… customers…

224 00:23:26.160 00:23:34.629 Alex Wilson: trying to make false payments. So you… you do have to push back, but there is a way that you could do it that I feel is not confrontational.

225 00:23:35.730 00:23:42.920 Alex Wilson: So, I wouldn’t approach it as confrontational, but I would, like, if I feel like a company…

226 00:23:43.460 00:23:50.440 Alex Wilson: would do better… approaching something differently, I would try to have those conversations and try to explain

227 00:23:50.680 00:23:54.299 Alex Wilson: as clearly as I could, why…

228 00:23:55.080 00:23:58.909 Alex Wilson: It might be a better idea to approach something in a different way.

229 00:23:59.640 00:24:06.729 Awaish Kumar: Okay, yeah, I think that’s it from my side. I will leave, like, last 2 minutes for you, if you have any questions.

230 00:24:08.010 00:24:11.970 Alex Wilson: I think I asked a few,

231 00:24:12.680 00:24:22.580 Alex Wilson: Again, with the… between the data engineering and the AI automation, how much crossover is there between the engineering and that work?

232 00:24:25.430 00:24:33.250 Awaish Kumar: Right now, like, the crossover is, like,

233 00:24:33.820 00:24:43.089 Awaish Kumar: you can say, like, I don’t think, I, maybe if I should, if I should ask, like, what do you mean by crossover?

234 00:24:43.590 00:24:49.059 Awaish Kumar: Because, like, the data team does all the data work, and the AI team works on…

235 00:24:49.200 00:24:54.540 Awaish Kumar: Or for the clients where… for the AI projects. Normally.

236 00:24:55.750 00:25:07.160 Awaish Kumar: If there is any data, like, data engineering work required, they will come back to data team, okay, we need help with building data warehouse for the client.

237 00:25:07.570 00:25:10.769 Awaish Kumar: Okay, let’s meet, let’s figure out what would be the way.

238 00:25:11.420 00:25:21.990 Awaish Kumar: And then, not really it’s like that, so… AI teams… Need data help?

239 00:25:22.310 00:25:23.280 Awaish Kumar: Alright.

240 00:25:24.280 00:25:30.459 Awaish Kumar: And generally, like, when data teams, lovely, like,

241 00:25:30.670 00:25:33.600 Awaish Kumar: need help on the i side is basically

242 00:25:33.820 00:25:43.040 Awaish Kumar: when clients want chat with AI kind of features, which we have some tools, like BI tools, like Omni, which provide that

243 00:25:43.180 00:25:46.579 Awaish Kumar: Kind of feature, so we normally try to use that, like.

244 00:25:46.920 00:25:53.910 Awaish Kumar: But if the scope increases, we completely kind of hand over, like, Okay, libs.

245 00:25:54.110 00:25:59.129 Awaish Kumar: make it a separate contract with our AI team, or it’s a separate service.

246 00:25:59.520 00:26:03.250 Awaish Kumar: let’s, like, scope it out differently. So that’s how we…

247 00:26:04.040 00:26:09.520 Awaish Kumar: like, do that. We don’t normally work in a single contract for all the services.

248 00:26:10.890 00:26:11.830 Alex Wilson: That makes sense.

249 00:26:14.420 00:26:20.339 Alex Wilson: And I think that’s it. I know we’re at the top of the hour. Thank you for your time.

250 00:26:21.010 00:26:24.310 Awaish Kumar: Yeah, that’s… that’s… There was,

251 00:26:26.110 00:26:35.390 Awaish Kumar: I think that was it from me also, like, so from Rico, from our operations team, we’ll reach out with the next steps.

252 00:26:36.050 00:26:40.670 Awaish Kumar: And, once I leave the feedback, And I think,

253 00:26:41.110 00:26:43.069 Awaish Kumar: It should be in the…

254 00:26:43.440 00:26:49.190 Awaish Kumar: maybe in a week’s time, he’s going to reach out, right? Okay.

255 00:26:49.880 00:26:50.440 Awaish Kumar: Thank you.

256 00:26:50.440 00:26:51.230 Alex Wilson: Thank you much.

257 00:26:51.440 00:26:52.619 Alex Wilson: You have a good day.