Meeting Title: Brainforge Data Engineer Interview Date: 2026-02-10 Meeting participants: Selenge Tulga, Awaish Kumar
WEBVTT
1 00:00:13.820 ⇒ 00:00:14.930 Selenge Tulga: Can you hear me?
2 00:00:16.190 ⇒ 00:00:17.620 Awaish Kumar: Yes, I can’t.
3 00:00:17.620 ⇒ 00:00:18.980 Selenge Tulga: Oh, hi, Ambush!
4 00:00:19.490 ⇒ 00:00:20.050 Awaish Kumar: Hi.
5 00:00:21.150 ⇒ 00:00:25.060 Selenge Tulga: Is it… am I pronouncing correctly, Ayubash?
6 00:00:25.710 ⇒ 00:00:26.520 Awaish Kumar: Ofish.
7 00:00:27.050 ⇒ 00:00:29.539 Selenge Tulga: Avish, okay. Got it, got it.
8 00:00:29.540 ⇒ 00:00:30.240 Awaish Kumar: Today?
9 00:00:30.350 ⇒ 00:00:31.160 Awaish Kumar: Excellent.
10 00:00:32.170 ⇒ 00:00:33.779 Awaish Kumar: How do you pronounce your name?
11 00:00:34.130 ⇒ 00:00:35.260 Selenge Tulga: A ceiling.
12 00:00:35.790 ⇒ 00:00:36.740 Awaish Kumar: swing, okay.
13 00:00:36.740 ⇒ 00:00:39.159 Selenge Tulga: More, like, silently selling.
14 00:00:39.560 ⇒ 00:00:43.559 Awaish Kumar: So, like, yeah, nice to meet you, Zlake, how are you doing?
15 00:00:43.950 ⇒ 00:00:49.270 Selenge Tulga: Oh, I’m doing good, perfect, yeah, it is job hunting thing, you know, right?
16 00:00:49.270 ⇒ 00:00:50.870 Awaish Kumar: It is sometimes…
17 00:00:51.140 ⇒ 00:00:55.860 Selenge Tulga: it feels so… I have… sometimes I have an anxiety.
18 00:00:56.080 ⇒ 00:01:00.060 Selenge Tulga: But, yeah, but it’s good, yeah.
19 00:01:01.950 ⇒ 00:01:03.860 Awaish Kumar: Yeah, how was your day today?
20 00:01:04.290 ⇒ 00:01:07.910 Selenge Tulga: It is pretty good now, it’s,
21 00:01:08.040 ⇒ 00:01:17.709 Selenge Tulga: 2PM in the… in Austin, Texas, yeah, it is the… spend time for the prepping and applying, yeah, as a usual stay.
22 00:01:18.380 ⇒ 00:01:20.220 Awaish Kumar: Okay, so…
23 00:01:20.220 ⇒ 00:01:21.369 Selenge Tulga: Where are your living?
24 00:01:22.130 ⇒ 00:01:24.050 Awaish Kumar: With UAE.
25 00:01:26.830 ⇒ 00:01:28.299 Awaish Kumar: Where are you?
26 00:01:28.920 ⇒ 00:01:31.780 Selenge Tulga: Oh, I am in Austin, Texas, in the US.
27 00:01:32.610 ⇒ 00:01:32.930 Awaish Kumar: Okay.
28 00:01:32.930 ⇒ 00:01:33.360 Selenge Tulga: What?
29 00:01:33.700 ⇒ 00:01:34.890 Selenge Tulga: Time is it now?
30 00:01:36.120 ⇒ 00:01:38.340 Awaish Kumar: It’s, 12, 130.
31 00:01:38.610 ⇒ 00:01:39.090 Awaish Kumar: Yeah.
32 00:01:39.090 ⇒ 00:01:42.489 Selenge Tulga: Oh… Oh, it’s midnight for you.
33 00:01:42.720 ⇒ 00:01:43.350 Awaish Kumar: Yeah.
34 00:01:44.670 ⇒ 00:01:51.700 Awaish Kumar: Yeah, that’s… okay. My name is Avish Kawar. I lead Terrajury.
35 00:01:53.040 ⇒ 00:01:54.699 Awaish Kumar: It’s Ed Bread Forge.
36 00:01:56.650 ⇒ 00:01:57.470 Awaish Kumar: So…
37 00:01:57.910 ⇒ 00:02:05.050 Awaish Kumar: like, I have both had around 8 to 10 years of experience working as a data engineer at multiple companies, startups.
38 00:02:06.310 ⇒ 00:02:11.400 Awaish Kumar: I have built companies, build the foundations for the data.
39 00:02:11.640 ⇒ 00:02:15.690 Awaish Kumar: So… That’s… that’s about me.
40 00:02:15.800 ⇒ 00:02:27.219 Awaish Kumar: And, for Brain Forge, if I can give you a little bit of introduction, it is a data IDI consultancy services company, and who provides services to…
41 00:02:27.620 ⇒ 00:02:32.200 Awaish Kumar: To clients, mostly with 100 million revenue.
42 00:02:32.390 ⇒ 00:02:41.710 Awaish Kumar: We have a team of 30 people from across departments. Also, like, maybe 10 to 15 in the engineering team.
43 00:02:43.100 ⇒ 00:02:52.769 Awaish Kumar: And… Which is… consists of engineers, analysts, analytics engineers, and for data engineers.
44 00:02:54.240 ⇒ 00:02:57.449 Awaish Kumar: We also have some packet engineers here.
45 00:02:58.240 ⇒ 00:03:01.010 Awaish Kumar: But yeah, that’s mostly it.
46 00:03:01.410 ⇒ 00:03:12.729 Awaish Kumar: And, we focus on data consultancy services, also provide AI consultancy services, which is mainly, maybe to the same customer, maybe some
47 00:03:12.910 ⇒ 00:03:17.730 Awaish Kumar: some customers which only want AI services, like chatbot and things like that.
48 00:03:18.040 ⇒ 00:03:23.419 Awaish Kumar: So that’s about brain folds. So, yeah, you can start with your introduction now.
49 00:03:23.830 ⇒ 00:03:36.790 Selenge Tulga: Okay, my name is Selene, and I am data engineer with over 7 years of industry experience, and I started my career as a software engineer at the Roseley Company in Mongolia.
50 00:03:37.200 ⇒ 00:03:53.059 Selenge Tulga: And I think which gave me a strong foundation in the building systems, and as my needs seems for the scalable data solution, I transitioned into data engineering, and over the past several years, I focused on building,
51 00:03:53.060 ⇒ 00:04:10.200 Selenge Tulga: production-grade data platforms and end-to-end owning, ingesting transformation, and data modeling, and downstream analytics. It can be the web app, or the Power BI dashboards and the Tableau. It depends on the company, right? And,
52 00:04:10.520 ⇒ 00:04:25.459 Selenge Tulga: In my, most recent role at the… at a consulting company, it is called the Superman Automation Group, and which is, more focused on the analytic field, and mostly Power BI, and I work with multiple clients, and…
53 00:04:25.460 ⇒ 00:04:33.580 Selenge Tulga: They have a similar issue, like, separate operations systems, they need to unify those systems to centralize, and
54 00:04:34.170 ⇒ 00:04:52.709 Selenge Tulga: turning, reliable data solutions, and now I’m looking for the company, where I can real ownership and contribute meaningfully, because my experience is mostly, before that, in a railway company, we have 16,000 employees, but we have just 6 software engineers.
55 00:04:52.720 ⇒ 00:04:55.499 Selenge Tulga: And the one system dis… system internet, and…
56 00:04:55.700 ⇒ 00:05:00.930 Selenge Tulga: My whole career, I’m working on mostly, like, a founding engineer.
57 00:05:01.470 ⇒ 00:05:07.749 Selenge Tulga: to solve the data solutions, to messy data into business decisions, yeah.
58 00:05:08.430 ⇒ 00:05:09.460 Awaish Kumar: Okay.
59 00:05:09.850 ⇒ 00:05:17.439 Awaish Kumar: As you mentioned, you’re an RF, like, founding engineer.
60 00:05:19.260 ⇒ 00:05:24.320 Awaish Kumar: I would love to understand how… if you can give me an example of…
61 00:05:25.070 ⇒ 00:05:33.369 Awaish Kumar: Where you have… led the foundation for the data project, and if you can… Walk me through the…
62 00:05:33.660 ⇒ 00:05:37.880 Awaish Kumar: How did you set up the infrastructure? How you made the decisions?
63 00:05:39.200 ⇒ 00:05:42.390 Awaish Kumar: And finally, how… how it got set up.
64 00:05:43.210 ⇒ 00:05:56.520 Selenge Tulga: Okay, I, talked about my last project. It is also in my, it is also very fascinating. Okay, my last project in a Zuckerman Automation Group, we are just concepting company, and…
65 00:05:57.280 ⇒ 00:06:04.159 Selenge Tulga: It is, customer was a door construction company, and they don’t have any,
66 00:06:04.350 ⇒ 00:06:24.699 Selenge Tulga: they don’t have any data engineering or the software team, because they just need to order the door and install them, and they all… the systems in just, they are using Intuit for the timesheet, and another one is called, FieldWise CRM. It is… they can track all the works
67 00:06:25.090 ⇒ 00:06:29.180 Selenge Tulga: And for the, progresses, and
68 00:06:29.560 ⇒ 00:06:32.500 Selenge Tulga: Rest of the all of data ended just in the Google Sheets.
69 00:06:32.820 ⇒ 00:06:39.280 Selenge Tulga: Because it is the financial data, and the… If some, projects,
70 00:06:39.560 ⇒ 00:06:42.500 Selenge Tulga: Didn’t start, they just need to,
71 00:06:42.730 ⇒ 00:06:46.269 Selenge Tulga: the track on a Google Sheet, not on the field by CRM. It is.
72 00:06:46.760 ⇒ 00:07:05.849 Selenge Tulga: it is so hard to track, and they wanted to ask, okay, we need to build the dashboard or the app system, but we just need to see the… all this data in one place. And in a consulting company, there are just three people, and my CEO, and me, and also one,
73 00:07:06.210 ⇒ 00:07:14.159 Selenge Tulga: the HR, more like, HR and the customer service, that they need to talk with their customers, and…
74 00:07:16.540 ⇒ 00:07:22.149 Selenge Tulga: And, we need to discuss, and okay, first we need to get all this data.
75 00:07:22.340 ⇒ 00:07:25.079 Selenge Tulga: And there’s… where is the data sources? And…
76 00:07:25.600 ⇒ 00:07:37.469 Selenge Tulga: after that, I need to understand, okay, what are you wanting, guys? What do you want? Because they just said, okay, we just need to see the data in a one…
77 00:07:37.610 ⇒ 00:07:44.480 Selenge Tulga: the dashboard, But I just ask, okay, what KPI do you need? And what,
78 00:07:44.940 ⇒ 00:07:52.970 Selenge Tulga: the functions to unit, because, if they need to change anything in this field via a CRM, they can… they need to…
79 00:07:52.970 ⇒ 00:08:05.160 Selenge Tulga: also, update the Google Sheet and the timesheet. And they said, okay, we need… I think it’s… we discussed almost for two weeks to just… to find out what they want.
80 00:08:05.640 ⇒ 00:08:25.099 Selenge Tulga: And, okay, this is the data source, and this is the… the KPI, they want, and also they wanted to see the data like a Gantz chart, because now the previous, Google Sheet was what looks like a Gantz chart, and they are just selling… sales of the call, then they can see the,
81 00:08:25.370 ⇒ 00:08:35.210 Selenge Tulga: the timelines, and where is the gap, and where is the work scheduled. And okay, I just, okay, I understand. And, first I did,
82 00:08:35.520 ⇒ 00:08:47.840 Selenge Tulga: the data ingesting, using their APIs, and, luckily, the Intuit cookbook and also this field via CRM and also, Google Sheet have an open API.
83 00:08:48.360 ⇒ 00:08:56.679 Selenge Tulga: And the FieldWire CRM has a webhook, which is great. We can do incremental loading. And, for the…
84 00:08:57.070 ⇒ 00:09:00.710 Selenge Tulga: For the cookbook, it is also… we have…
85 00:09:01.180 ⇒ 00:09:05.429 Selenge Tulga: Yeah, we have a webhook, and we just need to,
86 00:09:05.680 ⇒ 00:09:15.360 Selenge Tulga: the track where data changes, and most… the problem was, most painful problem was, integrating Google Sheets, because
87 00:09:15.540 ⇒ 00:09:22.919 Selenge Tulga: Okay, we can see, okay, this dollars amount and the cleaning data is, painful. Also, the most,
88 00:09:23.150 ⇒ 00:09:26.170 Selenge Tulga: Hardest part is, it is…
89 00:09:26.420 ⇒ 00:09:34.729 Selenge Tulga: They don’t have any webhook for the, booshit, and there was no reliable change tracking, and the partial,
90 00:09:35.070 ⇒ 00:09:42.610 Selenge Tulga: updates can cause the silence inconsistencies, and I just did some trade-off, and
91 00:09:42.820 ⇒ 00:09:56.319 Selenge Tulga: for the Google Sheet, there is no incremental loading, we just need to truncate and load, because it is… I think this is just around 800 projects, but it is easy to truncate, and
92 00:09:56.680 ⇒ 00:10:02.770 Selenge Tulga: load the data again. And, one key input was, building…
93 00:10:03.070 ⇒ 00:10:16.669 Selenge Tulga: getting data from this Gant-style worksheet, worksheet, and, I need to build, Python scripts to get this metadata, and what is a color change, yeah.
94 00:10:17.060 ⇒ 00:10:35.790 Selenge Tulga: And, and after the ingestion, I uploaded all data into Snowflake for the raw… the mob thing, and, using a dbt, I can create, okay, it’s the raw data and a transformation, intermediate layer, and also we build the data marks, and…
95 00:10:36.130 ⇒ 00:10:49.899 Selenge Tulga: the check with, the DBT and the Snowflakes, data validation and the freshness check, and result was, we replaced the 15… the manual spreadsheets.
96 00:10:51.340 ⇒ 00:10:53.660 Selenge Tulga: I hope it’s not so long.
97 00:10:55.140 ⇒ 00:10:57.619 Awaish Kumar: Yeah, it was great.
98 00:10:58.540 ⇒ 00:11:00.099 Awaish Kumar: What was that how…
99 00:11:01.680 ⇒ 00:11:12.459 Awaish Kumar: how did you decide, you know, all those tools? Like, you explained, you talked to the people, did a discovery, come up with KPIs, come up with sources.
100 00:11:13.120 ⇒ 00:11:23.940 Awaish Kumar: how did you then build decisions for the tools? Like, why you chose Snowflake? Why did… why not Piccurity? Why not Redshift? Why not Mother Duck?
101 00:11:24.930 ⇒ 00:11:26.260 Selenge Tulga: Yeah.
102 00:11:26.490 ⇒ 00:11:39.329 Selenge Tulga: That’s… yeah, that’s true. I also explored the model tech, and it was very grateful, and I think AWS… I also have experience with some AWS experiences, and I think AWS is also
103 00:11:39.870 ⇒ 00:11:56.490 Selenge Tulga: too big for this project, and, because it is, more like analytic project, and we don’t need to do the real-time processing, and, it’s more like a batch processing, and I didn’t… I think Snowflake is the good for,
104 00:11:56.810 ⇒ 00:11:59.759 Selenge Tulga: Especially for staging kids.
105 00:11:59.940 ⇒ 00:12:02.399 Selenge Tulga: If data is too small, like.
106 00:12:02.500 ⇒ 00:12:09.959 Awaish Kumar: Why don’t just… DuckDB, or… Like, why we treated Snowflake there?
107 00:12:11.930 ⇒ 00:12:14.300 Selenge Tulga: Yeah, for…
108 00:12:14.490 ⇒ 00:12:26.099 Selenge Tulga: The Google Sheet data was a small part. When it comes to this FieldVire CRM, it is… I think it’s around… the historical data is around the 1TB, and
109 00:12:26.610 ⇒ 00:12:37.780 Selenge Tulga: And also for the data governance and the data quality check, I think, the snowflake is a great way.
110 00:12:38.200 ⇒ 00:12:40.770 Awaish Kumar: You mentioned that you explored Mother Duck.
111 00:12:41.190 ⇒ 00:12:46.849 Awaish Kumar: So, seeing the size of the data, why not Mother Duck?
112 00:12:48.080 ⇒ 00:12:49.920 Selenge Tulga: Because…
113 00:12:50.250 ⇒ 00:13:01.120 Selenge Tulga: I’m also thinking, because I just explored a mother duck for, in my local computer, and I am not sure about, okay, how it works with my
114 00:13:01.680 ⇒ 00:13:05.749 Selenge Tulga: the production grade data, right? And…
115 00:13:05.750 ⇒ 00:13:08.250 Awaish Kumar: is a cloud database, right?
116 00:13:08.660 ⇒ 00:13:09.210 Selenge Tulga: Yeah.
117 00:13:09.210 ⇒ 00:13:13.240 Awaish Kumar: Like, you can… in local computers, you can use DuckDB, but…
118 00:13:13.570 ⇒ 00:13:16.600 Awaish Kumar: When you talk about mother duck, it is, like,
119 00:13:17.030 ⇒ 00:13:21.309 Awaish Kumar: tier of your house, which is a cloud similar to Snowflake.
120 00:13:21.820 ⇒ 00:13:22.950 Awaish Kumar: weekend, whatever.
121 00:13:23.530 ⇒ 00:13:26.530 Awaish Kumar: It’s a… it’s a data warehouse, right?
122 00:13:26.700 ⇒ 00:13:27.070 Selenge Tulga: Yeah.
123 00:13:27.410 ⇒ 00:13:32.380 Awaish Kumar: It can handle, like, Like, 1, 2 terabytes of data, very easily.
124 00:13:33.480 ⇒ 00:13:35.220 Selenge Tulga: Oh, okay, yeah.
125 00:13:35.680 ⇒ 00:13:43.419 Selenge Tulga: Because I… Honestly, I just say I was more familiar with the snowflake at this time, and…
126 00:13:43.920 ⇒ 00:13:54.059 Selenge Tulga: And, yeah, this is the reason I chose the… because, okay, I am more confident that the Snowflake and the dbt, how it works, and I choose the Snowflake, yeah, at this time.
127 00:13:54.840 ⇒ 00:14:03.140 Awaish Kumar: Okay, understood. So… If you are currently working, then why did you…
128 00:14:03.420 ⇒ 00:14:05.749 Awaish Kumar: Why are you looking for a new job?
129 00:14:06.030 ⇒ 00:14:19.460 Selenge Tulga: Oh, it is because I laid off them, last month, because it was a company issue. They turned, down to the data teams, yeah, because now they are just focusing on the Power BI projects.
130 00:14:22.840 ⇒ 00:14:25.760 Awaish Kumar: Okay, and…
131 00:14:29.250 ⇒ 00:14:34.870 Awaish Kumar: So… And, like, what are you looking for in your next role?
132 00:14:35.720 ⇒ 00:14:46.050 Selenge Tulga: I’m looking for the next role. It is… I think it’s my perspective, but some people can think differently, because,
133 00:14:46.440 ⇒ 00:14:49.069 Selenge Tulga: Most of the pupils,
134 00:14:49.340 ⇒ 00:14:56.169 Selenge Tulga: the dreams are okay working in the big tech, right? And you can be one of these thousands of engineers, you can…
135 00:14:56.360 ⇒ 00:15:05.769 Selenge Tulga: the bills, the great things, and the people. But for me, it’s my, I think it’s just my perspective, because,
136 00:15:05.960 ⇒ 00:15:09.730 Selenge Tulga: I think during this, 7 years, I know I work with
137 00:15:10.330 ⇒ 00:15:19.760 Selenge Tulga: I… including my internship, I work with the four different companies, and which is, Always the…
138 00:15:19.980 ⇒ 00:15:36.869 Selenge Tulga: they take steps to change how people work, and in a railway company, we replace it, the manual, the PHP, data entry into, data automating. It is… can replace these, boring manual jobs, and…
139 00:15:36.870 ⇒ 00:15:41.889 Selenge Tulga: Now I am looking for the company at the same value, like, okay, in my,
140 00:15:42.120 ⇒ 00:15:55.919 Selenge Tulga: I wanted to take an entrepreneurship, and I wanted to see how people lives change, and, I also, talked with Yudham, and he shared, we are working on multiple declines, and
141 00:15:56.260 ⇒ 00:16:04.100 Selenge Tulga: We are, solving customers’ painful points, and this is the reason why I wanted to work with Brainforge.
142 00:16:06.510 ⇒ 00:16:09.290 Awaish Kumar: Yeah, but… My question was…
143 00:16:09.840 ⇒ 00:16:12.979 Awaish Kumar: What are you looking for in your text role?
144 00:16:13.620 ⇒ 00:16:16.809 Selenge Tulga: Yeah, I am, more looking for the,
145 00:16:16.960 ⇒ 00:16:22.100 Selenge Tulga: The analytic data engineering roles, and…
146 00:16:22.400 ⇒ 00:16:30.290 Selenge Tulga: can… the data warehouse and debit snowflake, I can have an ownership, and… .
147 00:16:30.520 ⇒ 00:16:31.810 Awaish Kumar: If I can leave.
148 00:16:32.020 ⇒ 00:16:33.349 Awaish Kumar: my question.
149 00:16:33.650 ⇒ 00:16:37.799 Awaish Kumar: Like, where you would like to be in the next 5 years?
150 00:16:38.220 ⇒ 00:16:44.020 Selenge Tulga: For the next 5 years, I…
151 00:16:44.320 ⇒ 00:16:56.769 Selenge Tulga: I think it’s, 3 weeks ago, I, in a Austin, Texas, data meeting, and it was the same question, and I’m just thinking, I met it
152 00:16:57.060 ⇒ 00:17:15.949 Selenge Tulga: the, Meta’s ex-data director, and he asked me the same questions. I was saying, what would you want to do the next 5 years or the 10 years? And I just, okay, I want to… because most of the big companies already, solve their data problem. They are just,
153 00:17:16.130 ⇒ 00:17:22.530 Selenge Tulga: Pursuing different challenges, but there is a lot of the things to do in a smaller and a,
154 00:17:22.530 ⇒ 00:17:37.280 Selenge Tulga: smaller and or the medium-sized companies. I want to change this company’s problem, and I want to someone be, okay, if there is anything, data challenges, and Celine is, the person who can solve that problem, yeah.
155 00:17:38.560 ⇒ 00:17:43.049 Selenge Tulga: That is my goal for… They’re being a reliable data engineer.
156 00:17:45.460 ⇒ 00:17:47.050 Awaish Kumar: That you already are, right?
157 00:17:49.040 ⇒ 00:17:51.759 Awaish Kumar: My next question is then, like,
158 00:17:54.490 ⇒ 00:17:57.590 Awaish Kumar: So, if you are,
159 00:17:59.050 ⇒ 00:18:03.989 Awaish Kumar: My next question would be that, like, as a…
160 00:18:05.690 ⇒ 00:18:07.820 Awaish Kumar: Like, if you would like to…
161 00:18:08.190 ⇒ 00:18:11.159 Awaish Kumar: If you have an option to choose your carrier.
162 00:18:11.320 ⇒ 00:18:18.010 Awaish Kumar: to continue being in IC, in an IC role, or if you would like to… Have an opportunity to…
163 00:18:19.300 ⇒ 00:18:22.980 Awaish Kumar: take a different path, being an IC, or…
164 00:18:23.590 ⇒ 00:18:26.579 Awaish Kumar: Some kind of a lead role, or tech lead role.
165 00:18:27.080 ⇒ 00:18:29.960 Awaish Kumar: Like, which path would you like to choose?
166 00:18:31.590 ⇒ 00:18:32.740 Selenge Tulga: Mmm…
167 00:18:35.130 ⇒ 00:18:44.639 Selenge Tulga: I think it is… I am more, like, technical… I am mostly still with the technical roles, and maybe technical leads, and…
168 00:18:45.540 ⇒ 00:18:53.539 Awaish Kumar: Well, I mean, there are roles where you become an IC, and you continue to be an IC, like, as an individual contributor.
169 00:18:53.740 ⇒ 00:18:59.710 Awaish Kumar: You write code, Solve problems, deserve solutions.
170 00:18:59.890 ⇒ 00:19:07.140 Awaish Kumar: But that’s, like, the path of being an IC. Then there’s a path of being a… being into leadership roles, like, you…
171 00:19:07.480 ⇒ 00:19:11.560 Awaish Kumar: You spend some time doing… Doing hands-on.
172 00:19:11.770 ⇒ 00:19:22.640 Awaish Kumar: Programming, but, like, majority of times, then maybe you sometimes go to taking interviews, talking to your juniors, talking to peers.
173 00:19:22.820 ⇒ 00:19:28.750 Awaish Kumar: reviewing their work, things like that. So, like, which career path would you like to choose?
174 00:19:29.560 ⇒ 00:19:36.659 Selenge Tulga: I am more, like, still, staying with, individual contributor, and…
175 00:19:37.090 ⇒ 00:19:40.190 Selenge Tulga: The more, like, not like a more,
176 00:19:40.640 ⇒ 00:19:49.060 Selenge Tulga: the managing role, I just wanted some more technical, and the leading IT, maybe, mentoring juniors, and yeah.
177 00:19:49.460 ⇒ 00:19:50.969 Selenge Tulga: That’s the path I choose.
178 00:19:51.470 ⇒ 00:19:56.050 Awaish Kumar: I think that’s it from my side, if…
179 00:19:57.580 ⇒ 00:20:00.080 Awaish Kumar: Some time for you to ask any questions.
180 00:20:00.500 ⇒ 00:20:19.460 Selenge Tulga: Okay, it is just, for me, just interesting, because I see your, the LinkedIn, and you are working a lot of the years, and I… do you think, how do you feel AI today? Because you have experience before AI and after the AI, right? And…
181 00:20:19.710 ⇒ 00:20:25.990 Selenge Tulga: What difference have you noticed, and what has changed most in your day-to-day work?
182 00:20:26.800 ⇒ 00:20:27.750 Awaish Kumar: Yeah.
183 00:20:30.140 ⇒ 00:20:33.560 Awaish Kumar: What has… what has changed? Like, a lot of things have changed.
184 00:20:36.070 ⇒ 00:20:38.260 Awaish Kumar: Before and after AI, I think.
185 00:20:39.030 ⇒ 00:20:42.609 Awaish Kumar: Every… now everyone is a TEDx engineer.
186 00:20:42.610 ⇒ 00:20:43.530 Selenge Tulga: Yeah.
187 00:20:44.510 ⇒ 00:20:49.800 Awaish Kumar: So… If… if you’re a… like, now it’s more like…
188 00:20:50.760 ⇒ 00:20:53.910 Awaish Kumar: Being in a strategic role, or a…
189 00:20:54.140 ⇒ 00:20:59.620 Awaish Kumar: Architect role, is what… what is really needed right now.
190 00:21:01.330 ⇒ 00:21:16.909 Awaish Kumar: So you can see a problem and architect a solution. That’s what is needed from a person, from a human. After you have an architecture, or after you have designed a solution, it’s really… now have become, it’s really…
191 00:21:17.040 ⇒ 00:21:19.950 Awaish Kumar: Easy to execute, right?
192 00:21:20.950 ⇒ 00:21:31.940 Awaish Kumar: have a solution, once you have captured all the edge cases, once you have the business domain knowledge. If you write all of this down, then…
193 00:21:32.080 ⇒ 00:21:34.770 Awaish Kumar: And… and hand over this plan to any…
194 00:21:35.240 ⇒ 00:21:38.610 Awaish Kumar: AI agent, and it can execute.
195 00:21:39.020 ⇒ 00:21:40.749 Awaish Kumar: Institution has become really…
196 00:21:41.050 ⇒ 00:21:50.970 Awaish Kumar: fast, so that’s why I said everybody has a big optics interior, but that comes with a good architect, if you plan to
197 00:21:51.650 ⇒ 00:22:02.080 Awaish Kumar: really well. You can have a really good product, but otherwise, you are going to spend more time than prompting than actually writing code.
198 00:22:02.080 ⇒ 00:22:02.820 Selenge Tulga: Yeah.
199 00:22:03.790 ⇒ 00:22:11.449 Awaish Kumar: Yeah, it’s all that, Like, the… like, the…
200 00:22:12.950 ⇒ 00:22:19.109 Awaish Kumar: the skills you needed to learn to become a good engineer has, maybe has changed, like, you don’t need to learn.
201 00:22:19.110 ⇒ 00:22:19.560 Selenge Tulga: Hmm.
202 00:22:20.060 ⇒ 00:22:25.110 Awaish Kumar: Don’t remember syntax, you don’t have to… Where we learn programming.
203 00:22:25.550 ⇒ 00:22:29.570 Awaish Kumar: If you understand the… how language works.
204 00:22:30.170 ⇒ 00:22:30.730 Selenge Tulga: Bye.
205 00:22:30.790 ⇒ 00:22:36.750 Awaish Kumar: enough, but you have to know how the software works. As an engineer, you…
206 00:22:36.960 ⇒ 00:22:41.909 Awaish Kumar: You have to design a solution that’s much more important.
207 00:22:44.060 ⇒ 00:22:55.999 Selenge Tulga: Yeah, because now writing good SQL queries, maybe… I just… I just, 5 years ago, I just, okay, I can… I can do the great SQL and the processors and the queries, but now it’s not at 20.
208 00:22:56.530 ⇒ 00:22:57.720 Awaish Kumar: Yeah, it happened.
209 00:22:58.160 ⇒ 00:23:00.669 Awaish Kumar: before AI, it is like, you have to write
210 00:23:01.510 ⇒ 00:23:04.490 Awaish Kumar: If you are writing a script, you know, they…
211 00:23:04.850 ⇒ 00:23:12.090 Awaish Kumar: That’s great, right? You have done so much. But nowadays, you can write that same script in an hour.
212 00:23:14.220 ⇒ 00:23:17.650 Awaish Kumar: Only thing is that if you understand what needs to be done.
213 00:23:18.410 ⇒ 00:23:26.320 Awaish Kumar: if your script is doing what is expected, if you can test it, that’s all right. And you can do all of that with AI, so…
214 00:23:26.930 ⇒ 00:23:35.779 Awaish Kumar: That’s the basic change, right? You have to just figure out the ways to use it.
215 00:23:37.860 ⇒ 00:23:42.049 Selenge Tulga: How does your, typical days look like? It is… are you…
216 00:23:42.370 ⇒ 00:23:44.000 Awaish Kumar: It is more like…
217 00:23:44.380 ⇒ 00:23:47.029 Selenge Tulga: The talking customers are more like coding.
218 00:23:48.780 ⇒ 00:23:51.220 Awaish Kumar: It’s a mix of both.
219 00:23:52.220 ⇒ 00:23:59.479 Awaish Kumar: So I… So normally, we start our days with a stand-up, talk about clients.
220 00:23:59.610 ⇒ 00:24:05.580 Awaish Kumar: then we go back to work. And,
221 00:24:06.940 ⇒ 00:24:20.950 Awaish Kumar: Normally, we try to avoid any meetings, like, all the communication can be a sync. You have to… you can write in the Notion, you can send Zoom looms, or Zoom clips,
222 00:24:21.380 ⇒ 00:24:25.429 Awaish Kumar: For us in communication, Between the team members.
223 00:24:25.580 ⇒ 00:24:33.849 Awaish Kumar: But yeah, there are some meetings which are required, like the standards that is required. Then there are some…
224 00:24:34.740 ⇒ 00:24:39.430 Awaish Kumar: Company meeting, sometime, that maybe some… It’s a quiet one.
225 00:24:39.740 ⇒ 00:24:46.280 Awaish Kumar: But, then there are client meetings, the peop… the client you are working for, is…
226 00:24:46.480 ⇒ 00:24:48.680 Awaish Kumar: Is required, so… so, like.
227 00:24:48.680 ⇒ 00:24:49.330 Selenge Tulga: Hmm.
228 00:24:49.330 ⇒ 00:24:56.840 Awaish Kumar: It’s required for you to meet with them in your weekly check-ins or something like that, so… Normally,
229 00:24:57.660 ⇒ 00:25:01.319 Awaish Kumar: someone who is an IC…
230 00:25:01.480 ⇒ 00:25:13.250 Awaish Kumar: Every day, he might just have some stand… one stand-up, and then spend his time doing programming, or working tickets, and maybe once a week.
231 00:25:13.300 ⇒ 00:25:33.090 Awaish Kumar: he’s going to meet with his client. We are more a sync, so we send regular daily updates to internal channels, to external channels, what we are doing, what we have done this, today, what we have done. Like, we send out the updates, what we are ex…
232 00:25:33.300 ⇒ 00:25:37.880 Awaish Kumar: Expecting to finish today, and at the end of day, like, you can tell.
233 00:25:38.000 ⇒ 00:25:40.819 Awaish Kumar: You can say what has been done today.
234 00:25:40.820 ⇒ 00:25:41.770 Selenge Tulga: No,
235 00:25:41.910 ⇒ 00:25:46.129 Awaish Kumar: So, probably, we do all the Asan communication, but…
236 00:25:46.660 ⇒ 00:25:50.760 Awaish Kumar: Yeah, there are some meetings, and then there are…
237 00:25:51.140 ⇒ 00:25:54.369 Awaish Kumar: some hands-on work. So, like, I do all of it.
238 00:25:55.090 ⇒ 00:25:58.350 Awaish Kumar: meet with clients, I also meet with… have…
239 00:25:58.630 ⇒ 00:26:00.629 Awaish Kumar: I lead the stand-up, and also…
240 00:26:01.610 ⇒ 00:26:09.110 Awaish Kumar: do supply work, so… we use AI a lot, so it’s not a…
241 00:26:09.110 ⇒ 00:26:09.720 Selenge Tulga: Sure.
242 00:26:09.720 ⇒ 00:26:17.980 Awaish Kumar: Something, Like… Like, it’s something, like, everybody uses in the company.
243 00:26:18.640 ⇒ 00:26:22.140 Awaish Kumar: Everybody has access to all the AI tools,
244 00:26:22.280 ⇒ 00:26:26.139 Awaish Kumar: And if you prefer something, like, I think Utam is really…
245 00:26:26.360 ⇒ 00:26:34.510 Awaish Kumar: encouraging of using AI to improve your workflows, so I can get you anything if you… if you think it’s useful.
246 00:26:34.670 ⇒ 00:26:41.560 Awaish Kumar: But the midpoint, major thing is that as long as you are delivering your work reliably,
247 00:26:42.230 ⇒ 00:26:46.520 Awaish Kumar: The tools you are asking for make sense, then you get… you have it.
248 00:26:46.520 ⇒ 00:26:47.310 Selenge Tulga: Mmm.
249 00:26:47.490 ⇒ 00:26:48.160 Awaish Kumar: Oh.
250 00:26:49.020 ⇒ 00:26:58.269 Awaish Kumar: And we, at the company, everybody, almost everyone, doesn’t matter if it’s an engineer or not, uses AI to improve their workflows.
251 00:26:58.420 ⇒ 00:27:02.680 Awaish Kumar: And, yeah, that’s… that’s basically how we work here.
252 00:27:03.300 ⇒ 00:27:16.950 Selenge Tulga: That’s very cool. Okay, my last question, I think I have tons of questions to ask, okay. It’s… you said it is, we are now 10x engineering with an AI, right? We can do this all the… even the Apache
253 00:27:17.130 ⇒ 00:27:25.390 Selenge Tulga: It is, I think it’s 7 years ago, or 6 years ago, Apache Flink, and the Kaka was, it was really hard to learn, right?
254 00:27:25.560 ⇒ 00:27:35.119 Selenge Tulga: And we just need to read all the documents, and to see the stack overflows, all the comments, and to see… and… but now, we can ask the right questions.
255 00:27:35.170 ⇒ 00:27:46.339 Selenge Tulga: ask questions from the AI, and do you think now, what is the most challenging data problems you have, you are facing right now?
256 00:27:50.160 ⇒ 00:27:54.300 Awaish Kumar: In the data world, there are always two problems.
257 00:27:54.300 ⇒ 00:27:54.950 Selenge Tulga: Thank you.
258 00:27:54.950 ⇒ 00:28:02.960 Awaish Kumar: If it is… it’s AI, then it’s a directs, but you are always able to do the work that you are doing now with AI.
259 00:28:03.130 ⇒ 00:28:04.720 Awaish Kumar: So, nothing deal.
260 00:28:05.540 ⇒ 00:28:15.390 Awaish Kumar: So, the… I… what I think, there were two problems from the beginning, which… which are really, hard to solve, or maybe, like, people…
261 00:28:15.960 ⇒ 00:28:18.830 Awaish Kumar: Spend much, like, more time.
262 00:28:19.030 ⇒ 00:28:23.199 Awaish Kumar: solving those, because that’s really important. One is data quality.
263 00:28:25.230 ⇒ 00:28:28.429 Awaish Kumar: Like, AI hasn’t done much there yet.
264 00:28:28.950 ⇒ 00:28:32.289 Awaish Kumar: In terms of monitoring that.
265 00:28:32.410 ⇒ 00:28:37.830 Awaish Kumar: So… Like, now there are a lot of tools in the market, you can add…
266 00:28:38.830 ⇒ 00:28:44.949 Awaish Kumar: You can add, like, checks, or you can add custom checks, you have some
267 00:28:45.520 ⇒ 00:28:50.499 Awaish Kumar: AI running on your tables to look for outliers.
268 00:28:51.090 ⇒ 00:29:03.230 Awaish Kumar: But still, it’s not that mature, I would say. We have tried some tools, I think. What happens, like, you get a spam, and then you…
269 00:29:03.440 ⇒ 00:29:07.259 Awaish Kumar: lose track of it, because, like, some… some…
270 00:29:08.170 ⇒ 00:29:15.360 Awaish Kumar: Some values are expected, and then… but you still get this spam notification that, okay, it’s outlier, but it’s not.
271 00:29:15.650 ⇒ 00:29:28.559 Awaish Kumar: hence, like, you… you lose that interest, like, in the long run, people just, they’ll start to develop more, and don’t see the messages, because it might seem
272 00:29:29.070 ⇒ 00:29:38.010 Awaish Kumar: That it’s just, it’s, just a footboard, spam message, and that’s why,
273 00:29:39.910 ⇒ 00:29:44.659 Awaish Kumar: Data quality is really, like, the important thing in the data world.
274 00:29:45.940 ⇒ 00:29:51.480 Awaish Kumar: It has been challenging to… To actually accomplish that.
275 00:29:55.560 ⇒ 00:30:01.600 Awaish Kumar: Unless you spend, like, even dedicated people to do just data monitoring.
276 00:30:01.790 ⇒ 00:30:05.050 Awaish Kumar: It’s really hard to basically…
277 00:30:05.430 ⇒ 00:30:15.439 Awaish Kumar: speed it up, right? I have been at the companies where we had actual data QA people who could just sit there and do data QA.
278 00:30:15.860 ⇒ 00:30:19.509 Awaish Kumar: like, if I push a bottle, there are people sitting there.
279 00:30:20.250 ⇒ 00:30:24.620 Awaish Kumar: They monitor, like, okay, After pushing the change.
280 00:30:25.370 ⇒ 00:30:33.160 Awaish Kumar: Is there something… like, the… isn’t some… is there something which got broken, right? So there are people…
281 00:30:33.390 ⇒ 00:30:42.799 Awaish Kumar: the new models, the old models, and things like that, and their job is to basically QA and see if everything is working fine, or the data looks good.
282 00:30:43.070 ⇒ 00:30:43.790 Awaish Kumar: Wow.
283 00:30:45.040 ⇒ 00:30:45.910 Awaish Kumar: Duh.
284 00:30:46.390 ⇒ 00:30:50.970 Awaish Kumar: That’s what is challenging. Second, And I haven’t seen…
285 00:30:51.780 ⇒ 00:30:54.829 Awaish Kumar: It’s seen it like being mature.
286 00:30:55.250 ⇒ 00:31:00.220 Awaish Kumar: At the same pace as other stuff has.
287 00:31:00.860 ⇒ 00:31:07.430 Awaish Kumar: Second thing, getting, knowledge, right? Domain knowledge, right?
288 00:31:08.400 ⇒ 00:31:10.339 Awaish Kumar: Gathering that, like…
289 00:31:10.510 ⇒ 00:31:18.750 Awaish Kumar: maybe 10 people who joined the companies, 9 of them left, and all the knowledge lived with them. You might… you have, like, very little…
290 00:31:19.260 ⇒ 00:31:27.140 Awaish Kumar: knowledge of the data, of the systems, the things. So right now, like, I think we are in a phase where this is happening.
291 00:31:27.630 ⇒ 00:31:28.150 Selenge Tulga: people.
292 00:31:28.150 ⇒ 00:31:30.119 Awaish Kumar: I’m starting to gather…
293 00:31:30.930 ⇒ 00:31:40.410 Awaish Kumar: what’s going on, like, but… and we are starting to basically… we have started in our company to capture everything, to record everything.
294 00:31:40.460 ⇒ 00:31:54.810 Awaish Kumar: To get transcripts of every meeting and everything, so the data lives… the knowledge lives with us. If I talk to a person from my… from one of my clients for an year, and then he leaves, right?
295 00:31:54.900 ⇒ 00:32:04.319 Awaish Kumar: But I still have the records of each of BTGAT, and I have that knowledge. I could easily pass… pass it on to…
296 00:32:04.430 ⇒ 00:32:06.239 Awaish Kumar: Someone else who joins.
297 00:32:06.350 ⇒ 00:32:07.240 Awaish Kumar: the D.
298 00:32:07.930 ⇒ 00:32:12.400 Awaish Kumar: I can pass it out to my clients. So…
299 00:32:12.880 ⇒ 00:32:19.079 Awaish Kumar: But that, historically, it has not been there, and that’s why we… To have knowledge gaps.
300 00:32:19.850 ⇒ 00:32:23.409 Selenge Tulga: Yeah, okay, that’s great.
301 00:32:25.890 ⇒ 00:32:27.599 Awaish Kumar: I think we are on time.
302 00:32:27.600 ⇒ 00:32:28.960 Selenge Tulga: Yeah, yeah.
303 00:32:29.120 ⇒ 00:32:30.250 Awaish Kumar: I’ll get to you.
304 00:32:30.920 ⇒ 00:32:34.860 Awaish Kumar: I think… Rico from our team.
305 00:32:35.710 ⇒ 00:32:40.789 Awaish Kumar: is going to contact you about the next steps. I will submit my feedback, and after that.
306 00:32:41.060 ⇒ 00:32:42.910 Awaish Kumar: He’s going to come back on.
307 00:32:43.390 ⇒ 00:32:44.640 Awaish Kumar: Back to you. Okay.
308 00:32:44.640 ⇒ 00:32:49.070 Selenge Tulga: Yeah, thank you, Aybosh, and yeah, it’s very late for you.
309 00:32:49.370 ⇒ 00:32:54.310 Selenge Tulga: Thank you for taking my time today. Yeah, it was very insightful. Thank you.
310 00:32:54.480 ⇒ 00:32:55.320 Selenge Tulga: Bye-bye.