Meeting Title: Data Engineer Interview (Zoran Selinger) Date: 2025-08-05 Meeting participants: Zoran Selinger, Awaish Kumar
WEBVTT
1 00:01:09.490 ⇒ 00:01:10.440 Zoran Selinger: Hello!
2 00:01:16.040 ⇒ 00:01:16.730 Awaish Kumar: Hello!
3 00:01:16.880 ⇒ 00:01:18.149 Zoran Selinger: How are you doing?
4 00:01:18.580 ⇒ 00:01:21.030 Zoran Selinger: Yeah, very good, very good. How are you.
5 00:01:21.610 ⇒ 00:01:25.799 Awaish Kumar: I’m good as well. How to pronounce your name, Zoram.
6 00:01:26.180 ⇒ 00:01:31.959 Zoran Selinger: That’s fine. Yes, yes, yeah, that’s excellent, and you are a wife.
7 00:01:32.910 ⇒ 00:01:38.779 Zoran Selinger: Yes, Avesh Kumar, and like, where are you located?
8 00:01:39.500 ⇒ 00:01:40.869 Zoran Selinger: I’m in Croatia.
9 00:01:41.560 ⇒ 00:01:44.220 Zoran Selinger: Croatia is next to Italy. I’m not sure if you.
10 00:01:44.220 ⇒ 00:01:45.400 Awaish Kumar: Yeah, I know it.
11 00:01:45.670 ⇒ 00:01:46.170 Zoran Selinger: Yeah.
12 00:01:47.260 ⇒ 00:01:49.279 Awaish Kumar: I have colleagues from Russia.
13 00:01:49.440 ⇒ 00:01:50.259 Zoran Selinger: You do.
14 00:01:51.100 ⇒ 00:01:52.700 Awaish Kumar: I had. I want.
15 00:01:52.700 ⇒ 00:01:53.430 Zoran Selinger: You can. You can.
16 00:01:53.430 ⇒ 00:01:59.049 Awaish Kumar: I’ve worked in Denmark, and there I had leads from Krotia, working with us.
17 00:01:59.460 ⇒ 00:02:02.615 Zoran Selinger: Oh, okay, cool, cool. How’d you find them?
18 00:02:03.010 ⇒ 00:02:04.690 Awaish Kumar: Yeah, never. Nice.
19 00:02:04.690 ⇒ 00:02:05.636 Awaish Kumar: Yeah, okay.
20 00:02:06.110 ⇒ 00:02:06.710 Awaish Kumar: Yeah.
21 00:02:07.110 ⇒ 00:02:08.389 Zoran Selinger: Yeah.
22 00:02:09.060 ⇒ 00:02:14.320 Awaish Kumar: Okay, so my name is Avesh Kumar, and I’m an engineering manager
23 00:02:15.090 ⇒ 00:02:17.369 Awaish Kumar: here at Brain Forge.
24 00:02:18.530 ⇒ 00:02:22.410 Awaish Kumar: So basically, you you have already.
25 00:02:22.800 ⇒ 00:02:27.939 Awaish Kumar: had a conversation with autumn. So he he must have given a lot of a lot of context. What
26 00:02:28.150 ⇒ 00:02:29.729 Awaish Kumar: what brain force does.
27 00:02:29.730 ⇒ 00:02:30.250 Zoran Selinger: Yes.
28 00:02:30.820 ⇒ 00:02:39.639 Awaish Kumar: And if you have any questions you can ask, otherwise we can move ahead with with getting to know more about you and your experiences.
29 00:02:39.940 ⇒ 00:03:02.609 Zoran Selinger: Sure, sure. So I I talked to Tom last week, and generally kind of we he wanted me to see kind of the sample of the work you you guys do what kind of what kind of work it is? What are the tickets like? Right? So that’s why he pointed me to you. So about me.
30 00:03:02.770 ⇒ 00:03:14.169 Zoran Selinger: I’m Zoran. I’ve worked I mean, I have a a freelancing digital career for for almost 15 years now started in digital marketing.
31 00:03:14.663 ⇒ 00:03:26.470 Zoran Selinger: Done that like, I’ve managed campaigns and and all that. But I was always always into into technology as much as possible. So I was always trying to get
32 00:03:26.810 ⇒ 00:03:31.790 Zoran Selinger: to do as many of the technical things around that as I could.
33 00:03:32.150 ⇒ 00:03:54.869 Zoran Selinger: Naturally, that kinda led me to to analytics engineering? so I I will always be the one to set up set up tracking and and reporting and and all that jazz. And recently I’ve been doing a lot of like custom integration work for clients as well. Usually build stuff in aws.
34 00:03:56.740 ⇒ 00:04:26.140 Zoran Selinger: you know. For some custom stuff that like a normally detection. Things like that. For campaigns and all that. Generally speaking, just seen. Seen a lot of things Tom seems to. Seems to think that like a solution architect might be a good fit for for myself between that and maybe maybe actually doing some engineering work, which is sounds interesting to me.
35 00:04:26.260 ⇒ 00:04:32.119 Zoran Selinger: And yeah, I would love to see how that looks like. In in Brain Forge.
36 00:04:35.970 ⇒ 00:04:38.930 Awaish Kumar: Yeah. So you have been working as a
37 00:04:39.140 ⇒ 00:04:42.460 Awaish Kumar: fill out. So as an analytics like.
38 00:04:42.890 ⇒ 00:04:50.658 Awaish Kumar: I see your title as cloud Engineer and analytics, engineer. So where you do, you see yourself like
39 00:04:51.260 ⇒ 00:05:05.340 Awaish Kumar: like, what real time has been? Much of your time has been spent on doing modeling work or doing more of data, engineering work, building pipelines, setting up infrastructure and things like that.
40 00:05:05.560 ⇒ 00:05:30.019 Zoran Selinger: So it’s so the the official, the official title that that I have. That I’ve done most of my works work over the last, probably 4 or 5 years is the back end developer. But most of it. It’s almost all for marketing agencies. So it’s mostly integrate, like in api integration work
41 00:05:30.790 ⇒ 00:05:37.910 Zoran Selinger: and and data pipelines. Right? So it’s mostly it’s mostly those things I like to build.
42 00:05:38.670 ⇒ 00:05:39.889 Zoran Selinger: Yeah, go ahead.
43 00:05:40.070 ⇒ 00:05:50.690 Awaish Kumar: So you have done like like the data warehousing part. How’s that? Right? Apart from just injection like, have you done any transformation modeling on top of it?
44 00:05:50.690 ⇒ 00:05:56.370 Zoran Selinger: So. No, no. So that’s that’s bi team. They deal. They deal with all that. Yeah.
45 00:05:57.950 ⇒ 00:06:07.670 Awaish Kumar: Yeah. So what I’m getting is that you have been mostly working as a back end developer, but like, maybe do writing some Apis
46 00:06:08.080 ⇒ 00:06:15.220 Awaish Kumar: ingesting data connecting to some warehouses, getting data, and maybe using some
47 00:06:15.550 ⇒ 00:06:22.159 Awaish Kumar: Apis like custom writing some custom scripts to get data from some Apis and loading it to.
48 00:06:22.160 ⇒ 00:06:49.469 Zoran Selinger: There’s a lot of that. There’s a lot of that. So most of those Apis are, of course, advertising platforms. Crm stuff like that. So there would be like automation work like updating audiences automatically for for those marketing platforms. Then some things like on anomaly detection. In kind of real time anomaly detection stuff like that. That’s I mean, that’s very interesting work. I’m.
49 00:06:49.470 ⇒ 00:06:56.069 Awaish Kumar: Like, how like, how you are doing that injections like I want to understand more of
50 00:06:56.550 ⇒ 00:07:04.670 Awaish Kumar: to the road from like the work you have been doing so
51 00:07:05.020 ⇒ 00:07:08.760 Awaish Kumar: that in like right now, we have a lot of tools, a lot of
52 00:07:08.990 ⇒ 00:07:11.699 Awaish Kumar: cloud services to do these ingestions.
53 00:07:12.505 ⇒ 00:07:15.930 Awaish Kumar: So like, I see this?
54 00:07:17.020 ⇒ 00:07:17.740 Awaish Kumar: Like
55 00:07:17.870 ⇒ 00:07:39.540 Awaish Kumar: people. Some some companies are looking to hire a data engineer plus back end engineer kind of person who can get some in just some data for them, and then also can write some Apis. So their platforms can basically use these Apis to serve the data on on the front end side.
56 00:07:39.680 ⇒ 00:07:44.649 Awaish Kumar: And basically then just the data needs to little bit of
57 00:07:44.910 ⇒ 00:07:56.180 Awaish Kumar: transformations and and things like that. So this that role is more like back end engineering plus data engineering and kind of that role. So ha! Have you been doing that.
58 00:07:56.370 ⇒ 00:08:08.740 Zoran Selinger: Yeah, I’ve been. I’ve been, I’ve been doing. I’ve been doing that. For example, the the anomaly detection thing that that’s kinda a recent recent project that I’ve done is so
59 00:08:09.710 ⇒ 00:08:15.390 Zoran Selinger: this is built on on Aws. Lambdas, okay. Some.
60 00:08:16.330 ⇒ 00:08:21.890 Awaish Kumar: Anomaly. Detection is more like in the statistical project.
61 00:08:21.890 ⇒ 00:08:40.247 Zoran Selinger: Yes, so. But the the whole process involves ingestion of the data, you know, managing the data. So do it in big bigquery, right? And then a little bit of modeling, querying for more, for for anomalies. And then, you know, sending slack notification stuff like that.
62 00:08:41.099 ⇒ 00:08:47.570 Zoran Selinger: So there’s a little bit of little bit of everything in that project right? And I’ve done.
63 00:08:50.420 ⇒ 00:08:53.779 Awaish Kumar: How your experience looks like in in like.
64 00:08:54.640 ⇒ 00:08:59.469 Awaish Kumar: So if I would say, if you if you want to rate yourself
65 00:08:59.926 ⇒ 00:09:06.349 Awaish Kumar: on a like data engineer out of 10, as a data engineer and as a back end developer.
66 00:09:06.490 ⇒ 00:09:08.349 Awaish Kumar: how would you rate yourself.
67 00:09:08.870 ⇒ 00:09:12.120 Zoran Selinger: So my, my.
68 00:09:13.650 ⇒ 00:09:31.490 Zoran Selinger: the most precise thing I would call myself is the analytics engineer. So this is this is the the most of my work that I’ve ever done is in setting up analytics. So tracking for the clients for the systems, this is the most thing I’ve I’ve done.
69 00:09:32.280 ⇒ 00:09:34.460 Zoran Selinger: So it’s mostly it’s mostly.
70 00:09:34.460 ⇒ 00:09:39.201 Awaish Kumar: I I just want to see your interest and your
71 00:09:40.810 ⇒ 00:09:43.910 Awaish Kumar: like the strong technical
72 00:09:45.600 ⇒ 00:10:00.679 Awaish Kumar: like the capabilities like like as a back end engineer. You you are mostly mainly like they are good with building this full Apis, and things like that. Scalable, secure kind of like that’s knowledge
73 00:10:00.890 ⇒ 00:10:06.810 Awaish Kumar: is is this, with like as a back end, as a back end developer. You have more of experience on that.
74 00:10:06.960 ⇒ 00:10:09.519 Awaish Kumar: And like, for example, people working for.
75 00:10:09.520 ⇒ 00:10:14.099 Zoran Selinger: It’s more with data. I would still say, it’s more, it’s more data related work. Yes.
76 00:10:14.100 ⇒ 00:10:22.149 Awaish Kumar: Okay, so so like building, writing restful Apis, like, you have limited experience with that right.
77 00:10:22.720 ⇒ 00:10:34.939 Zoran Selinger: I mean, I’ve been. I’ve been kind of managing the the back end for a for a big agency, for like 4 or 5 years now.
78 00:10:34.940 ⇒ 00:10:39.703 Awaish Kumar: And how? How’s that like experience like, for example, was that,
79 00:10:41.120 ⇒ 00:10:44.149 Awaish Kumar: can I want to understand the architecture like, for example.
80 00:10:44.150 ⇒ 00:10:45.530 Zoran Selinger: I’ll tell you. I’ll tell you.
81 00:10:45.530 ⇒ 00:10:46.100 Awaish Kumar: Right
82 00:10:46.673 ⇒ 00:10:59.720 Awaish Kumar: like, there are gonna be some flask or some fast Api which can serve some some things, and also then there are some more like complex architectures, micro service, architecture, things like that.
83 00:11:03.841 ⇒ 00:11:20.170 Zoran Selinger: So in that that particular, for that particular client everything is, is cloud functions, everything. So, both in Lambdas and on Gcp. As well, so I have basically managing 2 stacks at the moment. Yeah, yeah.
84 00:11:20.998 ⇒ 00:11:38.329 Zoran Selinger: so everything is a micro service. Initially, I was, I was using the serverless framework for it, I mean for most of that. Time I was using the serverless framework to manage lambdas it worked really well now, with with version 4 and kinda
85 00:11:38.830 ⇒ 00:11:48.049 Zoran Selinger: I’m currently switching to aws. SDK, for most of that work. For example, yeah, that’s I’m currently working on that just transitioning.
86 00:11:48.050 ⇒ 00:11:51.520 Awaish Kumar: What are what are your? What is your interest?
87 00:11:51.820 ⇒ 00:11:53.810 Awaish Kumar: If you would like to? More
88 00:11:54.460 ⇒ 00:11:57.790 Awaish Kumar: like you have, you are more interested towards
89 00:11:58.380 ⇒ 00:12:19.279 Awaish Kumar: back in engineering or data engineering or ae work. A work is more like for us, like when we define analytics, engineering is more like doing modeling like the not just ingesting the data, but then working with the data to to transform it, to write from that as well.
90 00:12:19.713 ⇒ 00:12:34.669 Awaish Kumar: And like, for example, use Dbt and create some marks and then finally, maybe some experience with Vi tools like tableau or something. So how you what is your interest like. I want to understand that
91 00:12:35.390 ⇒ 00:12:40.850 Awaish Kumar: in like next 2 years what you would like. Which direction would you like to take?
92 00:12:41.330 ⇒ 00:12:47.170 Zoran Selinger: This is the action I would like to take. I would like to be closer to AI. This is my.
93 00:12:47.790 ⇒ 00:12:51.394 Zoran Selinger: this is my biggest thing right now.
94 00:12:52.389 ⇒ 00:13:21.929 Zoran Selinger: I kind of started being an enthusiast there. And this is where I want to learn, and I want to take what I know for now, so far, and kinda learn how to inject AI into that. And of course, actually do some do some great work. So whatever’s closest there when I was talking to Tom, just talking about my experience and how I like to approach work, he told me. Like
95 00:13:21.930 ⇒ 00:13:30.150 Zoran Selinger: he. He’s he’s looking for people close to the to the solutions architecture, right? And just with
96 00:13:30.600 ⇒ 00:13:35.219 Zoran Selinger: with experience that I have, it just sounds like that’s that’s it.
97 00:13:35.340 ⇒ 00:13:41.189 Zoran Selinger: Basically in in that one. I might not directly
98 00:13:41.410 ⇒ 00:13:44.350 Zoran Selinger: work on every little task. Right?
99 00:13:44.852 ⇒ 00:13:47.669 Zoran Selinger: Might not do a lot of engineering work
100 00:13:48.180 ⇒ 00:13:55.429 Zoran Selinger: directly. Yeah, but con conceptually, yes. In terms of what kind of tasks.
101 00:13:55.880 ⇒ 00:14:03.439 Zoran Selinger: I think. Yeah. Still, like AI data engineering, closer to that would be would be great, would be great.
102 00:14:04.490 ⇒ 00:14:16.939 Awaish Kumar: Yeah. Like, like, as you mentioned, we are looking for some sort of solution architects which which basically can lead the teams and and provide the solutions. But our hands on like they, they know
103 00:14:17.170 ⇒ 00:14:21.760 Awaish Kumar: the the work, they know what they are doing, and can fill in.
104 00:14:22.790 ⇒ 00:14:37.120 Zoran Selinger: I would love. I would like to do that as well like I I told him. Even if I I don’t want it to just be like meeting and document writing. I would actually like to do some engineering work.
105 00:14:37.780 ⇒ 00:14:38.700 Zoran Selinger: but.
106 00:14:38.700 ⇒ 00:14:39.220 Awaish Kumar: Okay.
107 00:14:39.220 ⇒ 00:14:42.829 Zoran Selinger: Basically. I wanna see how the
108 00:14:43.010 ⇒ 00:14:53.299 Zoran Selinger: what are some samples? How that? How does that look? In in in for this, he said you, you will be able to kind of show me. Give me a sample of how that.
109 00:14:54.020 ⇒ 00:14:56.520 Zoran Selinger: How does that look in in Brainforge?
110 00:14:56.520 ⇒ 00:15:02.549 Awaish Kumar: So like for samples, I I would say like, I’m not sure like that.
111 00:15:03.670 ⇒ 00:15:05.329 Zoran Selinger: I mean, I don’t really.
112 00:15:05.920 ⇒ 00:15:10.199 Awaish Kumar: I I don’t have the like tickets right now, but I can give you an overview of what.
113 00:15:10.200 ⇒ 00:15:12.540 Zoran Selinger: Yeah, that’s fine. Yeah, that’s fine.
114 00:15:13.560 ⇒ 00:15:14.273 Awaish Kumar: So the
115 00:15:15.050 ⇒ 00:15:32.550 Awaish Kumar: like, as I, as I mentioned, like, Brain Forge is an AI and data consult consultancy so basically, we have all, yeah, so we have, we are providing data consultancies plus AI services. So
116 00:15:33.260 ⇒ 00:15:37.193 Awaish Kumar: in the data side, you have like 3 different
117 00:15:38.279 ⇒ 00:15:46.549 Awaish Kumar: focus teams, we have data engineering, we have data analytics, engineering. And then we have data analyst
118 00:15:48.056 ⇒ 00:15:56.659 Awaish Kumar: position. So basically the kind of work we get is for example. Normally, you
119 00:15:57.681 ⇒ 00:16:22.969 Awaish Kumar: so it depends on clients. If the the if you have some client big clients like you can get okay, this is our data. This is our warehouse. And then, like, build some reports for it. Right? So basically so somebody, some clients, they don’t even have anything. And they just have some data, or they can give you an overview of, for example, we have some marketing tools.
120 00:16:23.100 ⇒ 00:16:52.490 Awaish Kumar: We have some like Google ads, Facebook ads, marketing data, some customer service data coming from Zendesk customer, I/O, things like that. And then we want to. And then we have some sales platforms like data is we are making running our sales platform like on on shopify or Amazon. So basically, this is the kind of data or someone, some even have their own custom. Crm.
121 00:16:52.570 ⇒ 00:16:54.920 Awaish Kumar: kind of like the ERM.
122 00:16:55.440 ⇒ 00:17:01.669 Awaish Kumar: System, where they all they they run their own sales related complex.
123 00:17:01.670 ⇒ 00:17:03.640 Zoran Selinger: I work with some custom? Crms, yeah.
124 00:17:03.640 ⇒ 00:17:26.910 Awaish Kumar: So basically so from customers to customer, these tools can vary, the things can vary. But the both, the what what we at the end. What we are doing is trying to ingest data from all these tools. So we are like and and we are not bound by any tools. If that that’s not coming from any client. For example, if I
125 00:17:27.650 ⇒ 00:17:35.990 Awaish Kumar: if I, if something is, comes, comes up, and I find a great tool to solve that problem in a in a reasonable manner, like in terms of pricing.
126 00:17:36.070 ⇒ 00:17:59.110 Awaish Kumar: I I can adopt. With that. We are flexible about that. If that makes sense, we should do that. If it’s a best practice, or if it is something new in the market, or and and makes complete sense in our use cases. So we adopted those tools, get the data and then then comes the the analytics engineering role.
127 00:17:59.180 ⇒ 00:18:19.699 Awaish Kumar: So that 1st part ingesting data writing, maybe. They have. They are using some tools. We don’t find a built in solution. So we do write. We have our own dexter pipelines. We we write our own custom pipelines to ingest data from those sources. So we have de work
128 00:18:20.045 ⇒ 00:18:29.760 Awaish Kumar: supporting our clients, then we have ae work, and most of it is ae work, because, what we are doing is like, the customer comes in. For example.
129 00:18:30.000 ⇒ 00:18:32.660 Awaish Kumar: I want to do some product analytics.
130 00:18:33.030 ⇒ 00:18:34.540 Awaish Kumar: Here’s my product.
131 00:18:34.680 ⇒ 00:18:42.120 Awaish Kumar: And that’s what I’m doing. And now I want to mayor this like, for example, I want to, Mayor.
132 00:18:42.655 ⇒ 00:18:47.794 Awaish Kumar: I have come customer segments like in this country in some regions
133 00:18:49.217 ⇒ 00:18:57.872 Awaish Kumar: customers who buy product Xyz want to measure or like their how they are doing
134 00:18:58.710 ⇒ 00:19:01.399 Awaish Kumar: on our platform like it’s a
135 00:19:01.998 ⇒ 00:19:07.499 Awaish Kumar: we want to analyze the people who end up clicking on some
136 00:19:07.720 ⇒ 00:19:18.579 Awaish Kumar: blogs. They end up making some some sales at the end of it right? So we want to measure, or maybe in the product analytics, we also have some requirements like, we have
137 00:19:18.720 ⇒ 00:19:31.560 Awaish Kumar: flow A on our website, we have flow. B. We don’t know what we have what the conversion looks like from these 2 flows. So things like that. And then there comes the brain forge. So
138 00:19:32.110 ⇒ 00:19:56.429 Awaish Kumar: right now, for example, that client which I just described does not even have. They have the the product where they’re selling some some soft. So they don’t even have the product analytics set up like we, we need to set up our own. What kind of events you want to capture. What kind of data want to capture. So we define the requirements, get the approval from client, then
139 00:19:56.570 ⇒ 00:20:07.120 Awaish Kumar: decide on the tools like what tools we need to use to get those data. And then, finally, when the data is in, we that that part is like, we need solution, architect basically
140 00:20:07.230 ⇒ 00:20:29.309 Awaish Kumar: design that solution like what the customer based on the customer needs. We define like, we need this tool. We need to set up these events. This data is going to come from here. And then we need some, a work, and we need to some build, some modeling models, some marks and things like that. And then we need a data analyst further down the roads to
141 00:20:29.570 ⇒ 00:20:32.550 Awaish Kumar: basically build some dashboard or yeah.
142 00:20:32.550 ⇒ 00:20:39.329 Awaish Kumar: show metrics for the to the client, right? How? How? So? That’s what basically we do on the data side.
143 00:20:39.930 ⇒ 00:20:51.340 Zoran Selinger: So out of everything that you just described. The most things I’ve done so far is the scenario where clients are missing tracking
144 00:20:52.260 ⇒ 00:20:56.369 Zoran Selinger: when when we are missing data. And we wanna implement something
145 00:20:56.650 ⇒ 00:21:14.619 Zoran Selinger: and start actually ingesting start collecting and ingesting data in some shape or form. Obviously, I’ve done. I’ve done local reports and and stuff like that, so that that part is is also fine, right? But most of it it was. It was it was.
146 00:21:14.620 ⇒ 00:21:20.669 Awaish Kumar: So there are some in in terms of missing data. There are 2 things like one is that
147 00:21:21.152 ⇒ 00:21:24.050 Awaish Kumar: they just don’t. They are getting data
148 00:21:25.226 ⇒ 00:21:29.920 Awaish Kumar: from their sales platform in the warehouse. But they haven’t yet.
149 00:21:30.732 ⇒ 00:21:32.450 Awaish Kumar: Automated their marketing.
150 00:21:32.660 ⇒ 00:21:33.750 Awaish Kumar: Yeah. Yeah.
151 00:21:33.750 ⇒ 00:21:40.030 Awaish Kumar: So we help them get the marketing data in the warehouse and then combine marketing data with sales data
152 00:21:40.240 ⇒ 00:21:47.090 Awaish Kumar: to show the how the campaigns are doing the performance of the campaign. So things like that. But the second thing is then
153 00:21:48.020 ⇒ 00:21:54.600 Awaish Kumar: in the product, analytic analytics side, we have the setting up the the tools like the development
154 00:21:54.970 ⇒ 00:21:56.240 Awaish Kumar: itself. Right?
155 00:21:59.000 ⇒ 00:22:16.770 Awaish Kumar: So you have to set up some Gtm. Like the client needs a event data which is not being captured right now. So we need might need to go in into the Gtm and set up those events. And then the data starts coming to Big carry. And then we can run call confirmation.
156 00:22:16.970 ⇒ 00:22:18.755 Awaish Kumar: So basically, that’s
157 00:22:20.340 ⇒ 00:22:24.420 Awaish Kumar: So what you have done like. Have you done some gtm or.
158 00:22:24.420 ⇒ 00:22:28.770 Zoran Selinger: I’ve done hundreds, hundreds of Gtm projects
159 00:22:29.080 ⇒ 00:22:33.450 Zoran Selinger: for big, small, everything, custom, server, side, everything.
160 00:22:34.100 ⇒ 00:22:37.800 Zoran Selinger: everything. Gtm, Google analytics. I’ve done.
161 00:22:39.710 ⇒ 00:22:51.497 Zoran Selinger: I mean, even like like, I said, look reports and just just today I was I was setting up avin for for
162 00:22:52.380 ⇒ 00:22:54.100 Zoran Selinger: our client.
163 00:22:54.745 ⇒ 00:23:05.150 Zoran Selinger: We’re just setting up avian. They have both the client and server side part that needs to be set up and consent, and all that stuff.
164 00:23:06.770 ⇒ 00:23:07.969 Awaish Kumar: Okay. So
165 00:23:09.661 ⇒ 00:23:23.399 Awaish Kumar: Hi, so, okay, yeah, I want, yeah, I haven’t got this distracted, distracted a little bit I wanted to get to know more about your projects. In terms of data engineering and the data
166 00:23:23.710 ⇒ 00:23:26.000 Awaish Kumar: analytics engineering.
167 00:23:26.110 ⇒ 00:23:27.190 Awaish Kumar: So
168 00:23:28.404 ⇒ 00:23:36.645 Awaish Kumar: what I want wanted to know? Like, yeah, you you showed your interest is more like in AI. So now we are touching a lot of different
169 00:23:38.659 ⇒ 00:24:06.379 Awaish Kumar: roles like data, engineering analytics, engineering, AI engineering. So what exactly like, I would like to understand like, so if you come in in the brain force, we we do, we have data engineering work. We have analytics, engineering work. We have AI engineering work. But we want to make sure that, like, for example, if we hire you as an analytics engineer we like. We want you to do that like you. You might get some work on the AI side as well
170 00:24:06.710 ⇒ 00:24:07.550 Zoran Selinger: Sounds, great.
171 00:24:07.550 ⇒ 00:24:13.769 Awaish Kumar: To learn that, but, like the the main focus, would be to ensure that the current needs
172 00:24:14.000 ⇒ 00:24:18.060 Awaish Kumar: we have for clients like they are just like they are met.
173 00:24:19.580 ⇒ 00:24:20.120 Zoran Selinger: Yeah.
174 00:24:21.350 ⇒ 00:24:30.600 Zoran Selinger: So what does a typical solution architect do in brain forge? So what is there a a
175 00:24:31.380 ⇒ 00:24:35.080 Zoran Selinger: do? They do an engineering work a lot.
176 00:24:38.430 ⇒ 00:24:53.599 Awaish Kumar: yeah, like, it depends. Really. Right? Now, we we don’t have much, many solution, we have lead developers. So lead developers are basically kind of doing the solution architect work along with the
177 00:24:54.160 ⇒ 00:25:03.430 Awaish Kumar: also developing as part of yeah. But like, it really depends on on the fun.
178 00:25:04.680 ⇒ 00:25:32.370 Awaish Kumar: on the on the clients and the team like, if we have enough resources, you know, the lead is the solution. Architect is going to propose some solutions and responsible for driving the solution. Using the resources need lead developers or analytic engineers or analyst data analysts. Right? So that’s the more main focus. But so far we have a lead engineer who is also primarily developing doing the development work as well.
179 00:25:35.570 ⇒ 00:25:36.700 Zoran Selinger: Listen.
180 00:25:38.490 ⇒ 00:25:42.310 Zoran Selinger: Best case is I would.
181 00:25:44.580 ⇒ 00:25:48.120 Zoran Selinger: I will do a lot of analytics work.
182 00:25:48.790 ⇒ 00:25:54.450 Zoran Selinger: But I really wanna learn about AI. So this is this is this is.
183 00:25:54.450 ⇒ 00:25:57.080 Awaish Kumar: Yeah, you are going to get that right. It’s not A.
184 00:25:57.240 ⇒ 00:26:08.039 Awaish Kumar: So AI is is more about like how much you are going to want to take on and learn, and things like that. Right? So we have AI work. And
185 00:26:08.350 ⇒ 00:26:21.870 Awaish Kumar: so in the company. I we have some internal you have some external clients, but it really depends on who should you have done your ae work, and then you’re still available, and then
186 00:26:22.030 ⇒ 00:26:32.030 Awaish Kumar: you can do that the AI engineering work. So that’s completely fine work for yeah, that’s okay. If
187 00:26:32.780 ⇒ 00:26:40.179 Awaish Kumar: at Brainforge you will always be encouraged to use the eye in your work, and if you want to
188 00:26:40.380 ⇒ 00:26:46.066 Awaish Kumar: pursue as a carrier, AI! There’s there’s no blockers on that right.
189 00:26:46.540 ⇒ 00:26:59.880 Zoran Selinger: Great. I mean, I use a AI on a daily basis. I I use cursor in my work, of course, and and and all of that. So that’s that. That is part of my my workflow. But
190 00:27:00.250 ⇒ 00:27:05.539 Zoran Selinger: rag in general Reg is something like I would.
191 00:27:05.640 ⇒ 00:27:17.220 Zoran Selinger: I’m really curious about how how that process works. I’m really curious. I really wanna see it. And of course, I wanna I wanna I wanna be a part of of that process as well.
192 00:27:17.880 ⇒ 00:27:23.779 Zoran Selinger: So if we have like, I see you, you do projects like that. So that’s that is very interesting to me.
193 00:27:24.580 ⇒ 00:27:25.300 Zoran Selinger: Right.
194 00:27:25.300 ⇒ 00:27:35.449 Awaish Kumar: And okay, apart from that, how like, what kind of the tech stack you have worked with, for example, you mentioned big Carry. But what else you use.
195 00:27:35.450 ⇒ 00:27:37.249 Zoran Selinger: I mean. Sorry. Can you say again.
196 00:27:37.250 ⇒ 00:27:41.110 Awaish Kumar: Tech stack like what programming language you have for.
197 00:27:41.110 ⇒ 00:27:44.310 Zoran Selinger: Of course. So yeah, it’s Javascript. Yeah, I
198 00:27:44.410 ⇒ 00:27:48.880 Zoran Selinger: I I hear hear you you guys do python. Right? You.
199 00:27:51.100 ⇒ 00:27:56.299 Awaish Kumar: You mentioned. You have been writing ingestation pipelines, or were you writing them in Javascript?
200 00:27:57.900 ⇒ 00:28:03.330 Zoran Selinger: No. So this is basically, this is pure, pure
201 00:28:04.750 ⇒ 00:28:24.860 Zoran Selinger: back end work building Apis ingestion in this. In this last last com over the last few years. Just bi does this, and they use Dbt, right? I don’t. I don’t deal with with big data ingestion in that call in in that for that client at all.
202 00:28:26.070 ⇒ 00:28:31.200 Awaish Kumar: No, no. And you mentioned about ingesting ingesting the data right?
203 00:28:31.800 ⇒ 00:28:32.940 Zoran Selinger: Yeah, but that that’s.
204 00:28:32.940 ⇒ 00:28:33.540 Awaish Kumar: So both.
205 00:28:33.860 ⇒ 00:28:46.429 Zoran Selinger: Yeah. But that’s just that’s just a little bit of Api work. And the workflow that that I built does not deal with with big data sets, or they’re most like web hooks.
206 00:28:46.430 ⇒ 00:28:50.429 Awaish Kumar: Yeah, that that is something like, that’s what did engineers do right?
207 00:28:50.430 ⇒ 00:28:51.000 Zoran Selinger: It’s.
208 00:28:51.370 ⇒ 00:28:58.800 Awaish Kumar: They they write those scripts or to ingest the data and build the data pipeline, basically. And the the data pipeline.
209 00:28:59.030 ⇒ 00:29:03.419 Awaish Kumar: So Javascript is is for front end visualizations.
210 00:29:04.060 ⇒ 00:29:15.307 Zoran Selinger: Oh, not. I’m not. I mean Node. We I use node on on the back end. So so I’m using using Javascript to do all that work. So I’m using.
211 00:29:15.670 ⇒ 00:29:20.239 Awaish Kumar: Yeah, I want to understand. I want to understand your workflows. That’s what
212 00:29:20.430 ⇒ 00:29:33.590 Awaish Kumar: that’s what I mean. Like when I when I asked about if you are doing data, engineering or data like the more like if you’re writing Javascript building front end and they’re calling some vacant Apis
213 00:29:33.690 ⇒ 00:29:35.410 Awaish Kumar: and then
214 00:29:35.960 ⇒ 00:29:46.539 Awaish Kumar: and back end is doing some something and returns data and you show it on front end. That’s kind of role of a back end engineer or a or maybe full stack engineer.
215 00:29:47.420 ⇒ 00:30:10.699 Awaish Kumar: The data engineer does like more of the working with big carry working with warehouses, big data ingesting pipeline data coming from Google Ads made different these platforms. And then it goes to Big Carry. Then they are data from Crm, or maybe to the Crm tools and things like that. That’s the data engineering part.
216 00:30:11.160 ⇒ 00:30:12.840 Zoran Selinger: Yeah. So been
217 00:30:13.020 ⇒ 00:30:21.290 Zoran Selinger: doing a lot of dot that I I mean, I call that mostly integration work right? Because I’m I’m
218 00:30:21.710 ⇒ 00:30:26.089 Zoran Selinger: pulling data from from one place, shaping it
219 00:30:26.210 ⇒ 00:30:29.390 Zoran Selinger: to work with with the next platform right.
220 00:30:29.390 ⇒ 00:30:38.380 Awaish Kumar: Yeah, but that that like they can developers do that right. Exactly. The they pull data from some database, and then they might need to
221 00:30:38.650 ⇒ 00:30:40.883 Awaish Kumar: do little bit of transformation.
222 00:30:41.330 ⇒ 00:30:46.720 Zoran Selinger: Yeah, you just business logic of some kind. And then, yeah.
223 00:30:46.720 ⇒ 00:30:51.460 Awaish Kumar: But that is to satisfy how they want to show on the front end, like
224 00:30:51.984 ⇒ 00:30:57.499 Awaish Kumar: in in what format or things like that like for that. They do some transformation right?
225 00:30:57.980 ⇒ 00:31:05.605 Zoran Selinger: There’s a lot of specific marketing scenarios. It’s not just visualization or notification, like
226 00:31:06.150 ⇒ 00:31:06.920 Awaish Kumar: I don’t.
227 00:31:06.920 ⇒ 00:31:09.350 Zoran Selinger: Like offline conversions are a big thing.
228 00:31:09.350 ⇒ 00:31:13.019 Awaish Kumar: There are Tours, for example, which are showing the
229 00:31:13.848 ⇒ 00:31:23.389 Awaish Kumar: analytics for Amazon, for example, for any plan like you can plug and play to your Amazon account, and they just pop up everything for you, and that’s like
230 00:31:23.940 ⇒ 00:31:37.369 Awaish Kumar: in the Javascript. They would they? They need to do some aggregations and some transformation to basically to be able to utilize that right? I understand that part and I get like, that’s
231 00:31:37.620 ⇒ 00:31:39.539 Awaish Kumar: what I get from, you know.
232 00:31:41.380 ⇒ 00:31:51.209 Awaish Kumar: from you like. That’s why you have been mentioning about analytics engineering like. So, how have have you been using the Dbt.
233 00:31:52.270 ⇒ 00:31:58.489 Zoran Selinger: Very little because I don’t do data modeling at all. Currently, in my work.
234 00:31:59.160 ⇒ 00:31:59.690 Awaish Kumar: Yes.
235 00:31:59.690 ⇒ 00:32:01.070 Zoran Selinger: I do?
236 00:32:01.070 ⇒ 00:32:05.880 Awaish Kumar: That I’m so our analytics engineers. They do a lot of Dbt.
237 00:32:05.880 ⇒ 00:32:22.340 Zoran Selinger: Yeah, I mean the the company I work for. right now they do. Everything’s dbt as well. Everything’s dbt, so I know what it is. I’ve I’ve done a little bit of like Cicd pipeline work where I I would run
238 00:32:22.680 ⇒ 00:32:26.560 Zoran Selinger: a little bit of Dbt Cli commands
239 00:32:26.670 ⇒ 00:32:35.440 Zoran Selinger: to complete them. But that’s just very, very basic usage. I don’t. I don’t do it day to day at all.
240 00:32:36.020 ⇒ 00:32:36.960 Zoran Selinger: It’s purely.
241 00:32:38.633 ⇒ 00:32:51.730 Awaish Kumar: How do you see yourself in like in the next 5 years? Like I, I understand you mentioned about learning move moving towards AI. But did you want to continue being I in IC roles, or
242 00:32:52.140 ⇒ 00:33:00.030 Awaish Kumar: the staff level, like the roles stuff engineer, or A,
243 00:33:00.680 ⇒ 00:33:03.189 Awaish Kumar: or more like a managerial positions?
244 00:33:04.403 ⇒ 00:33:11.939 Zoran Selinger: So I’m interested in in technology. That’s what I’m interested in. And I told told Utam
245 00:33:12.900 ⇒ 00:33:14.259 Zoran Selinger: I’ve done
246 00:33:14.510 ⇒ 00:33:35.010 Zoran Selinger: so many different things. Since I started. I’ve been freelancing most of that time, usually had pretty wide roles. I’ve I’ve done anything, everything there is in digital marketing to do from writing a B testing cr, like, email, all the channels and I, just.
247 00:33:35.090 ⇒ 00:33:48.050 Zoran Selinger: I like solving problems. Whatever in A enables me to do that this is where I want to be. So that sounds like solution, architect, position, or something adjacent to that.
248 00:33:48.320 ⇒ 00:33:49.920 Awaish Kumar: Yeah, that’s okay. That’s.
249 00:33:49.920 ⇒ 00:33:50.999 Zoran Selinger: That’s exactly that.
250 00:33:51.000 ⇒ 00:33:51.830 Awaish Kumar: That’s not my.
251 00:33:52.090 ⇒ 00:34:07.399 Awaish Kumar: but that’s not my question, for for now, right, I understand you want to be solution. Architect. We have a role for that. Utah already discussed things with you. But I’m I want to understand how you want to move forward like from here until the next 5 years
252 00:34:07.750 ⇒ 00:34:21.049 Awaish Kumar: next 5 years you want don’t want to be like solution, architect for your whole life. Right? You want to grow to some senior levels. So which path you want to take, like more like towards being.
253 00:34:22.310 ⇒ 00:34:39.469 Zoran Selinger: I mean, obviously the the biggest, the biggest position there is that I see is CTO. That’s net, I mean. That’s of course, that is the answer. A CTO would be the the ultimate goal here.
254 00:34:41.110 ⇒ 00:34:41.900 Awaish Kumar: Okay, fine.
255 00:34:42.529 ⇒ 00:34:43.289 Awaish Kumar: It’s cool.
256 00:34:43.819 ⇒ 00:34:52.549 Awaish Kumar: So I would understand from there, like you want to, we focus more on gosh, for example.
257 00:34:52.669 ⇒ 00:34:57.919 Awaish Kumar: more on engineering online being hands on close the door.
258 00:34:57.920 ⇒ 00:35:09.080 Zoran Selinger: I would like to be. I would like to be hands on. This is cause I really like to do the work as well, I like being in the code right? I like
259 00:35:09.280 ⇒ 00:35:10.969 Zoran Selinger: getting my hands dirty.
260 00:35:11.360 ⇒ 00:35:14.089 Awaish Kumar: Are you familiar with Python?
261 00:35:14.260 ⇒ 00:35:15.510 Awaish Kumar: Little better.
262 00:35:15.660 ⇒ 00:35:27.030 Zoran Selinger: When I when I initially started I I started with with python. Obviously. Then I joined this company. Their stack was was js, so I switched
263 00:35:28.690 ⇒ 00:35:35.350 Zoran Selinger: I have no strong opinions of or preference towards Javascript.
264 00:35:35.480 ⇒ 00:35:42.196 Zoran Selinger: I just obviously we need to get up to speed. I’ve done some like tutorials.
265 00:35:42.570 ⇒ 00:35:43.520 Awaish Kumar: But what about.
266 00:35:43.520 ⇒ 00:35:47.340 Zoran Selinger: And and those things like that right before I’ve done it.
267 00:35:47.340 ⇒ 00:35:50.839 Awaish Kumar: What other languages you have experience with, apart from Javascript.
268 00:35:52.060 ⇒ 00:35:57.820 Zoran Selinger: Apart from, I mean a little bit of a python. I do. I’ve done cool stuff with r
269 00:36:01.450 ⇒ 00:36:04.519 Awaish Kumar: Like, have you studied computer science or.
270 00:36:04.913 ⇒ 00:36:10.035 Zoran Selinger: Yeah, I I had a I have a like information management
271 00:36:11.106 ⇒ 00:36:13.740 Zoran Selinger: degree. Like, master’s in.
272 00:36:13.740 ⇒ 00:36:14.070 Awaish Kumar: So.
273 00:36:14.070 ⇒ 00:36:15.090 Zoran Selinger: Management.
274 00:36:15.750 ⇒ 00:36:21.149 Awaish Kumar: Yeah. So like, how like, as you mentioned, you have little bit of
275 00:36:21.570 ⇒ 00:36:24.210 Awaish Kumar: knowledge of Python, some knowledge of.
276 00:36:24.210 ⇒ 00:36:26.960 Zoran Selinger: Every everything was at the job.
277 00:36:27.760 ⇒ 00:36:36.650 Awaish Kumar: Not just. Yeah. But so once you see the difference between, for example, Python, and
278 00:36:38.484 ⇒ 00:36:42.440 Awaish Kumar: like some other, for example, Nodejs or Javascript.
279 00:36:43.620 ⇒ 00:36:53.710 Zoran Selinger: The diff I mean python is, is just like a high level that is very versatile, right?
280 00:36:54.110 ⇒ 00:36:56.499 Zoran Selinger: High level language that is very versatile.
281 00:36:57.245 ⇒ 00:37:00.880 Zoran Selinger: Is very good for for data.
282 00:37:01.535 ⇒ 00:37:04.964 Zoran Selinger: I’m I’m aware of that. I just chose when I was doing
283 00:37:05.830 ⇒ 00:37:09.530 Zoran Selinger: more data work I I chose. R.
284 00:37:11.520 ⇒ 00:37:13.120 Zoran Selinger: Do you know about R.
285 00:37:13.880 ⇒ 00:37:14.570 Awaish Kumar: Yeah, yeah.
286 00:37:14.570 ⇒ 00:37:22.220 Zoran Selinger: Yeah, I chose R. But python is, I see python is now preferred, is mostly preferred for for data work
287 00:37:23.011 ⇒ 00:37:25.179 Zoran Selinger: for data engineering and.
288 00:37:25.180 ⇒ 00:37:33.089 Awaish Kumar: How like, for example, how, for example, I see you have a lot of experience with with Javascript on there, right.
289 00:37:33.980 ⇒ 00:37:34.600 Zoran Selinger: Yeah.
290 00:37:36.030 ⇒ 00:37:40.230 Awaish Kumar: Like how memory, like management, works in Javascript.
291 00:37:42.670 ⇒ 00:37:45.580 Zoran Selinger: How how memory, memory, management, works.
292 00:37:45.580 ⇒ 00:37:58.069 Awaish Kumar: Like, yeah, in in programming language, we create variables, we create lists, things like that. So like. And they are that that process have limited memory
293 00:37:58.240 ⇒ 00:37:59.310 Awaish Kumar: assigned.
294 00:37:59.310 ⇒ 00:37:59.740 Zoran Selinger: Yes.
295 00:37:59.740 ⇒ 00:38:02.570 Awaish Kumar: How they manage the memory.
296 00:38:02.990 ⇒ 00:38:09.370 Zoran Selinger: I mean, we do not manage, especially especially in micro services. I mean, we we don’t manage.
297 00:38:09.370 ⇒ 00:38:10.590 Awaish Kumar: We don’t manage like.
298 00:38:10.590 ⇒ 00:38:11.290 Zoran Selinger: That’s good.
299 00:38:11.290 ⇒ 00:38:17.090 Awaish Kumar: The language Javascript, as a language manages the memory
300 00:38:17.400 ⇒ 00:38:27.210 Awaish Kumar: like, if you define variable. And then yeah, what? What is it like? How how does that happen like? Do you know anything about.
301 00:38:28.380 ⇒ 00:38:33.919 Zoran Selinger: No, it’s I don’t have experience with with low level programming languages. Okay.
302 00:38:33.920 ⇒ 00:38:34.710 Awaish Kumar: No, no, I’m not.
303 00:38:34.710 ⇒ 00:38:41.040 Zoran Selinger: In terms of how memory works behind the scenes. I know very little about.
304 00:38:42.180 ⇒ 00:38:42.950 Awaish Kumar: Forbid!
305 00:38:42.950 ⇒ 00:38:43.380 Zoran Selinger: Yeah.
306 00:38:43.380 ⇒ 00:38:45.750 Awaish Kumar: I just want to understand more like how the
307 00:38:46.020 ⇒ 00:38:49.510 Awaish Kumar: because if you understand how memory works, you can better
308 00:38:49.730 ⇒ 00:38:52.890 Awaish Kumar: that report. So that’s the only point here.
309 00:38:53.530 ⇒ 00:38:57.290 Awaish Kumar: So you are. I’m not asking you to write assembly code, or or I don’t want.
310 00:38:59.010 ⇒ 00:39:02.920 Awaish Kumar: I need people who who can, who work only with the High Level
311 00:39:03.540 ⇒ 00:39:16.260 Awaish Kumar: programming languages. So we are. We don’t need anyone working with the low level languages. But we just need people. They understand how the language is working itself. So it’s it’s just good
312 00:39:16.500 ⇒ 00:39:20.360 Awaish Kumar: to know. And and it helps writing better. Code. That’s it.
313 00:39:20.786 ⇒ 00:39:22.490 Zoran Selinger: 100% 100%.
314 00:39:23.380 ⇒ 00:39:31.770 Zoran Selinger: Yeah, just the system that I I work in is just microservices.
315 00:39:32.900 ⇒ 00:39:36.639 Zoran Selinger: Functions are kind of small use. Cases are simple.
316 00:39:36.640 ⇒ 00:39:37.500 Awaish Kumar: But my.
317 00:39:37.500 ⇒ 00:39:40.569 Zoran Selinger: And optimization.
318 00:39:41.940 ⇒ 00:39:45.510 Zoran Selinger: Optimization is O-, obviously very useful. But.
319 00:39:45.510 ⇒ 00:39:50.830 Awaish Kumar: The micro services you are saying. Do they collaborate with each other, or.
320 00:39:51.090 ⇒ 00:40:14.739 Zoran Selinger: I mean, obviously we do with Sqs and and and Sns and all all. Of course they do communicate with each other. I mean, I have controllers and workers separately. They communicate via an Sqs. Right? Then the worker pulls, pulls the Sqs. Messages. They trigger it, they pull it. They process it. Basically every event
321 00:40:15.050 ⇒ 00:40:20.210 Zoran Selinger: runs its own, its own lambda, runtime of lambda function right.
322 00:40:20.210 ⇒ 00:40:20.980 Awaish Kumar: Good.
323 00:40:20.980 ⇒ 00:40:21.520 Zoran Selinger: Yeah.
324 00:40:23.230 ⇒ 00:40:31.289 Awaish Kumar: So I mean, and what has been your role in writing that micro service like, are you be like, are you one?
325 00:40:32.322 ⇒ 00:40:34.620 Awaish Kumar: Implementing that as a developer or.
326 00:40:34.620 ⇒ 00:40:36.519 Zoran Selinger: Yes, everything, everything.
327 00:40:36.800 ⇒ 00:40:38.369 Zoran Selinger: I do everything.
328 00:40:39.010 ⇒ 00:40:41.929 Zoran Selinger: So I just, I get a, I get a use case.
329 00:40:42.070 ⇒ 00:40:47.849 Zoran Selinger: We got a use case. Let’s call it. Okay, we need, we need automated audiences.
330 00:40:48.420 ⇒ 00:40:55.800 Zoran Selinger: So we need a day daily updates of the audiences on 5 different platforms, Google ads, Facebook ads and all that.
331 00:40:55.940 ⇒ 00:41:03.520 Zoran Selinger: So I create a an Api. This is one. This is, for example, one endpoint, just
332 00:41:04.250 ⇒ 00:41:14.610 Zoran Selinger: one endpoint. You get a request that day, so that request contains a query for bigquery and Sq. Like bigquery, SQL.
333 00:41:15.560 ⇒ 00:41:16.070 Awaish Kumar: Okay.
334 00:41:16.070 ⇒ 00:41:19.259 Zoran Selinger: Is the audience Id, which platform it is.
335 00:41:19.460 ⇒ 00:41:23.089 Zoran Selinger: and where we wanna remove or add people. Right?
336 00:41:23.580 ⇒ 00:41:24.590 Zoran Selinger: That’s it.
337 00:41:25.700 ⇒ 00:41:27.070 Awaish Kumar: So, then.
338 00:41:27.654 ⇒ 00:41:29.989 Zoran Selinger: Code, everything, right.
339 00:41:29.990 ⇒ 00:41:33.969 Awaish Kumar: So. Do you know anything about warehouses like, for example, Big Carry?
340 00:41:35.360 ⇒ 00:41:36.970 Awaish Kumar: Are you familiar with that?
341 00:41:37.200 ⇒ 00:41:43.820 Awaish Kumar: I mean you mentioned I I do write a a little little bit of like basic SQL.
342 00:41:44.707 ⇒ 00:41:47.030 Zoran Selinger: Now things have been working on.
343 00:41:48.020 ⇒ 00:41:50.391 Awaish Kumar: I wanted to understand more about like,
344 00:41:52.190 ⇒ 00:41:54.299 Awaish Kumar: I want to deep dive into
345 00:41:55.520 ⇒ 00:42:08.630 Awaish Kumar: bigquery like how you? How would you, for example, architect the the data database, for example, tomorrow, if if we hire you, and then a client comes in
346 00:42:08.860 ⇒ 00:42:12.590 Awaish Kumar: and he don’t have a warehouse yet. For example.
347 00:42:12.880 ⇒ 00:42:15.502 Awaish Kumar: he has some operations running on
348 00:42:16.310 ⇒ 00:42:35.320 Awaish Kumar: on a custom Emr system. They they have in house system. They are maybe using postgres as their database to run their operations. That’s like they. They have engineering team, but for that on purpose. Now we are hired for for supporting their data work. So
349 00:42:35.520 ⇒ 00:42:41.250 Awaish Kumar: and you are you as a solution architect is assigned to basically understand
350 00:42:42.250 ⇒ 00:42:50.378 Awaish Kumar: what the data looks like, how they are going to when the sales happen, what kind of data comes in like, how does data products, data, different
351 00:42:51.550 ⇒ 00:42:59.440 Awaish Kumar: different dimensions like different things. And then we have that data.
352 00:42:59.660 ⇒ 00:43:02.540 Awaish Kumar: So 1st of all, that data is going to come
353 00:43:03.173 ⇒ 00:43:22.580 Awaish Kumar: to the warehouse so like that you you might like, we don’t care about the ingestion tools, for now we we will be using something to get that data to warehouse. But when it lands to warehouse, we want to architect that right, how it is, how it is going to be when it’s ingested, and then how we are going to
354 00:43:22.690 ⇒ 00:43:29.460 Awaish Kumar: architected for our modeling work. And then how we are going to, basically what.
355 00:43:29.700 ⇒ 00:43:33.940 Awaish Kumar: how we are going to solve that to end, to end users.
356 00:43:34.100 ⇒ 00:43:41.139 Awaish Kumar: So what that architecture looks like? Or do you know anything about database modeling, or things like that?
357 00:43:41.480 ⇒ 00:43:56.029 Zoran Selinger: No. So I mean, I’m just. I’m a user of of bigquery in that work. But I mostly I mean, Bi, Bi does that. They are. They are modeling and and doing.
358 00:43:56.030 ⇒ 00:43:59.350 Awaish Kumar: Are you familiar with databases for your have you worked with postgres?
359 00:43:59.955 ⇒ 00:44:02.850 Zoran Selinger: Yes, of course, of course.
360 00:44:02.850 ⇒ 00:44:03.349 Awaish Kumar: How would you.
361 00:44:03.350 ⇒ 00:44:12.619 Zoran Selinger: Those are like transact, like document databases transactional databases. I’m familiar with basic concepts. I I can.
362 00:44:12.890 ⇒ 00:44:13.530 Zoran Selinger: I’ve never.
363 00:44:13.530 ⇒ 00:44:14.409 Awaish Kumar: I have one question.
364 00:44:16.590 ⇒ 00:44:25.019 Awaish Kumar: yeah, I have a question for postgres. For example, how would you optimize like, for example, I have a table that’s
365 00:44:25.480 ⇒ 00:44:34.729 Awaish Kumar: very big. And when I run a query or to find something on from that table it is basically taking a lot of time.
366 00:44:34.730 ⇒ 00:44:35.540 Awaish Kumar: Yeah, yeah, it’s
367 00:44:35.540 ⇒ 00:44:45.600 Awaish Kumar: that time that query. So how would you go and optimize that table to improve the query, Execution time.
368 00:44:47.210 ⇒ 00:44:52.179 Zoran Selinger: I mean, I I think we we need to use indices. But I I can’t tell you
369 00:44:53.590 ⇒ 00:44:57.959 Zoran Selinger: detail about that. I’ve never done that kind of work, never.
370 00:45:00.790 ⇒ 00:45:03.300 Awaish Kumar: Okay, okay. And
371 00:45:05.840 ⇒ 00:45:12.990 Zoran Selinger: So yeah, just mostly interacted with document stores, because it’s
372 00:45:13.980 ⇒ 00:45:28.169 Zoran Selinger: it’s just appropriate for for the use cases that that I had in in marketing. They they can be, they can be appropriate. But, like, for example, that company they do. They do a lot of they do a lot.
373 00:45:28.170 ⇒ 00:45:30.649 Awaish Kumar: What do you use for document storage, for example.
374 00:45:31.040 ⇒ 00:45:32.270 Zoran Selinger: Elastic.
375 00:45:32.540 ⇒ 00:45:35.870 Zoran Selinger: For example, that was, I kind of inherited that system.
376 00:45:35.870 ⇒ 00:45:38.766 Awaish Kumar: That’s more about like for for logging
377 00:45:40.910 ⇒ 00:45:42.560 Zoran Selinger: Yeah. So when I, when I.
378 00:45:42.560 ⇒ 00:45:43.900 Awaish Kumar: Yeah, or something else here.
379 00:45:43.900 ⇒ 00:45:50.049 Zoran Selinger: When I arrived, it was basically, all the event data
380 00:45:50.580 ⇒ 00:46:04.180 Zoran Selinger: was was inelastic. We we used it for a while then then. Now everything is in in bigquery, and it wasn’t a decision that was made by me, because I don’t. I don’t deal with that part at all.
381 00:46:08.420 ⇒ 00:46:16.590 Awaish Kumar: Okay, so what I’m getting is that you mostly have. what? Basically as a
382 00:46:19.230 ⇒ 00:46:23.243 Awaish Kumar: data heavy back in yeah, back in engineer. But
383 00:46:23.840 ⇒ 00:46:26.540 Zoran Selinger: Doing some data related work.
384 00:46:27.020 ⇒ 00:46:27.960 Zoran Selinger: I think.
385 00:46:28.920 ⇒ 00:46:30.260 Awaish Kumar: Like in a data, heavy company.
386 00:46:30.260 ⇒ 00:46:32.810 Zoran Selinger: I mean, yeah, right now, I understand
387 00:46:32.970 ⇒ 00:46:38.240 Zoran Selinger: most. Mostly, I think that’s accurate. That is, that is accurate, does it?
388 00:46:38.240 ⇒ 00:46:46.810 Awaish Kumar: Yeah. As a back, like as a back end engineer, you mostly use the data which is coming from. Maybe someone is there who’s putting data to bigquery. Right?
389 00:46:47.060 ⇒ 00:46:48.350 Zoran Selinger: He might discontinue.
390 00:46:48.350 ⇒ 00:46:54.190 Awaish Kumar: Engineer team. Or maybe they have a data engineer in the team for some someone there.
391 00:46:54.300 ⇒ 00:47:19.820 Awaish Kumar: So and then you, you are basically using that data. But I understand there are some companies who who basically have such use cases to show a lot of data to user or things like that. And and as a back end engineer, you have to deal with that like you have to run some group by some applications and things like that to serve the customers needs on the front end side.
392 00:47:21.560 ⇒ 00:47:42.809 Zoran Selinger: Yes, I mean, we have a Us. For example, in in that we front and team is completely separate, right? And like we mostly ingest and and deal with data. That is, that is, like event by event basis. We inject business business logic.
393 00:47:43.110 ⇒ 00:47:46.220 Zoran Selinger: that is event by event, event basis
394 00:47:46.320 ⇒ 00:48:03.590 Zoran Selinger: and the front end team. If they have to show data, do visualizations, they do that with a completely different stack. Right? I don’t at all touch front end. I don’t serve so front, end and all I serve. I I serve the platforms and the tracking team
395 00:48:04.330 ⇒ 00:48:05.520 Zoran Selinger: with with that work.
396 00:48:05.520 ⇒ 00:48:11.360 Awaish Kumar: Okay. So so so you have been using Javascript to do your back end work.
397 00:48:12.000 ⇒ 00:48:20.820 Zoran Selinger: Yeah, it’s yeah. I just inherited the the guy that that’s that set up it initially. It was just it was in old.
398 00:48:21.940 ⇒ 00:48:22.960 Zoran Selinger: okay.
399 00:48:28.390 ⇒ 00:48:29.837 Awaish Kumar: Yeah, okay,
400 00:48:30.960 ⇒ 00:48:44.259 Awaish Kumar: I don’t have any questions right now, like, it was really nice talking to you. And now we have. Only I think 10 min left or 12, if you have you, if you have any questions.
401 00:48:44.400 ⇒ 00:48:45.310 Awaish Kumar: Gosh!
402 00:48:45.510 ⇒ 00:48:47.199 Zoran Selinger: Yeah, I mean, I was.
403 00:48:48.760 ⇒ 00:48:55.229 Zoran Selinger: I’m interested in in kinda like the tracking work.
404 00:48:55.680 ⇒ 00:49:01.840 Zoran Selinger: And what what do AI engineers do in Benforge.
405 00:49:02.820 ⇒ 00:49:11.599 Awaish Kumar: So like tracking work we have is like we, we do get some tracking like we as I, as I give an example of product analytics project.
406 00:49:12.069 ⇒ 00:49:12.540 Zoran Selinger: Yeah.
407 00:49:12.540 ⇒ 00:49:13.310 Awaish Kumar: So.
408 00:49:13.520 ⇒ 00:49:22.844 Awaish Kumar: And similarly, we we might get some projects for for Gtm work. Right? So we have. My client have might have a
409 00:49:24.160 ⇒ 00:49:36.760 Awaish Kumar: a platform or website, and then they might have set up some. Gtm, but that’s not efficient, or that’s something needs to be improved. Some missing events. Things like that number. One second thing is
410 00:49:37.540 ⇒ 00:49:44.220 Awaish Kumar: we have. We might have to. We might have to set up some other tools, for example, post hook or something else
411 00:49:44.751 ⇒ 00:50:07.900 Awaish Kumar: capture those events. So we might get some general requirements like they didn’t even touch anything right. The client comes in and they say, like we? We are running of operations. But we don’t know anything about analytics work. So we we come in and then figure out what are the requirements and what tools can we use, and what is the easy to
412 00:50:08.360 ⇒ 00:50:16.150 Awaish Kumar: setup and plus less like a scalable solution.
413 00:50:16.960 ⇒ 00:50:28.653 Awaish Kumar: and then finally make some decisions and and start from there. So that’s like the kind of tracking work that’s required.
414 00:50:30.391 ⇒ 00:50:33.438 Awaish Kumar: And then you mentioned about AI, the Eid
415 00:50:34.270 ⇒ 00:50:41.669 Awaish Kumar: is basically doing a lot of different like there. They are basically buildings of agents, right? Some they are using.
416 00:50:42.385 ⇒ 00:50:49.734 Awaish Kumar: Open a open AI and and different. They’re like different analytics
417 00:50:50.470 ⇒ 00:50:55.289 Awaish Kumar: versions like for the O 3 or 4, or or, for example.
418 00:50:55.975 ⇒ 00:51:03.499 Awaish Kumar: they are also utilizing different other AI tools like perplexity, or things like that don’t do it. Just
419 00:51:03.982 ⇒ 00:51:23.950 Awaish Kumar: get a sense of like how these tools work, and then how better soft clients. So what they are doing in the back end is just like as an AI engineer. What you do is is build model. But they are basically building complete solutions. So, for example, a client comes in and they want a checkboard which basically can book A
420 00:51:24.210 ⇒ 00:51:28.159 Awaish Kumar: can make a booking on during the chatting right? So.
421 00:51:28.160 ⇒ 00:51:29.059 Zoran Selinger: Yeah, yeah.
422 00:51:29.060 ⇒ 00:51:34.080 Awaish Kumar: So plan comes in and you tell with it, like, okay, we have column.
423 00:51:34.350 ⇒ 00:51:59.879 Awaish Kumar: And then what is replying that we have okay, nice room or you know, for you all these days. Okay, would you like to book it? Okay, show. If you want to show big, see pictures and things like that. And if if they say yes like, they confirm for the booking. So basically, the chat board can basically figure out the information dates, pricing everything and then send a request
424 00:52:00.220 ⇒ 00:52:03.719 Awaish Kumar: to the backend system that okay, make a booking. So
425 00:52:04.210 ⇒ 00:52:08.379 Awaish Kumar: and then basically, we have have given an interface for the chat.
426 00:52:08.800 ⇒ 00:52:12.450 Awaish Kumar: and they can interact like their like.
427 00:52:12.610 ⇒ 00:52:19.860 Awaish Kumar: The client’s client, like customers, can basically interact with that checkboard and and make a booking and basically
428 00:52:20.290 ⇒ 00:52:24.951 Awaish Kumar: building that chat board and all the back end communication happening with
429 00:52:25.820 ⇒ 00:52:34.810 Awaish Kumar: with any agent or whatever. AI agent. So that’s that’s handled by our AI engineers and all that infrastructure as well. So
430 00:52:35.588 ⇒ 00:52:38.021 Awaish Kumar: how how to train.
431 00:52:38.930 ⇒ 00:52:48.399 Awaish Kumar: for example, like how to get context for for the kind of request we are getting, and to to give it to the, for example, the
432 00:52:50.100 ⇒ 00:53:02.739 Awaish Kumar: our agent, and give the context and the the platform Urls and everything. How to make a booking, basically. And then you communicate with the the customer, and
433 00:53:03.070 ⇒ 00:53:07.316 Awaish Kumar: and finally, it makes a more comfortable. So in the back end that’s happening
434 00:53:10.225 ⇒ 00:53:17.809 Awaish Kumar: that communication like, so the front end up on the model, and also
435 00:53:19.120 ⇒ 00:53:22.360 Awaish Kumar: the backend, and how the data is stored.
436 00:53:22.560 ⇒ 00:53:27.269 Awaish Kumar: how the communication is going to be stored. And if we need to do some.
437 00:53:27.500 ⇒ 00:53:36.250 Awaish Kumar: So if you need to do some rag or things like that like that, then then 1st sector level decisions are are made by AI engineers.
438 00:53:36.520 ⇒ 00:53:37.590 Zoran Selinger: Yeah. Yeah.
439 00:53:37.890 ⇒ 00:53:41.173 Zoran Selinger: So let’s say, let’s say,
440 00:53:42.660 ⇒ 00:53:47.279 Zoran Selinger: I come on board just for for the tracking tasks.
441 00:53:47.570 ⇒ 00:53:56.059 Zoran Selinger: And you say there’s an opportunity to kind of learn whatever we want, right?
442 00:53:57.350 ⇒ 00:54:08.670 Zoran Selinger: So if I if I do, if I mostly do that like missing tracking tasks, or however you want to call it but I also want to touch a little bit of AI
443 00:54:09.470 ⇒ 00:54:14.659 Zoran Selinger: is, would there be an opportunity about? Would there be opportunity? There.
444 00:54:16.800 ⇒ 00:54:22.449 Awaish Kumar: Yeah, like. So, as I mentioned it, it is going to be some give and take kind of situation.
445 00:54:23.029 ⇒ 00:54:42.510 Awaish Kumar: So, for example, like, as I mentioned like, for example, if I if in that in our team we are, we are going to make a decision hiring you. We like we have to figure out what clients or what projects can work on. If you are, if you’re a good fit for that, you can work on that.
446 00:54:42.920 ⇒ 00:55:02.019 Awaish Kumar: Basically. Then, obviously, if you are now, when you are part of the team, right, you you get to know everything. Will you have access to all the team members or the engineers and everyone you can interact with them. You can ask for like more work on the I side, or
447 00:55:02.427 ⇒ 00:55:17.170 Awaish Kumar: and maybe like you can. You can pair with them right? So that will be that all will be available. So. But it’s it’s like. But but as an like, if you haven’t worked with the AI, you are not going to be hired as an AI engineer.
448 00:55:17.650 ⇒ 00:55:20.230 Awaish Kumar: And and if I mean 1st item.
449 00:55:20.230 ⇒ 00:55:42.830 Awaish Kumar: so basically, you come something else. And then we can learn like you can get some tickets which basically, you can try to work with the AI team to to finish them pair with them. They might. They will be willing to like help. You have some working session with you. That. That’s okay, like that’s we have we? Everyone has very collaborative.
450 00:55:43.080 ⇒ 00:55:43.469 Zoran Selinger: You can.
451 00:55:43.823 ⇒ 00:55:56.209 Awaish Kumar: Get that help from them tickets from them. And once once you are like capable enough, you can maybe might might just switch being a engineer as well. But that’s going to come on.
452 00:55:56.580 ⇒ 00:56:00.900 Awaish Kumar: But initially, it’s it’s going to be like some talents.
453 00:56:02.680 ⇒ 00:56:09.960 Zoran Selinger: Of course, of course I understand that. I just wanna see if I if I do have like an ability to raise my hand to say.
454 00:56:10.530 ⇒ 00:56:13.960 Zoran Selinger: could I do this? Could I try and do do that? One.
455 00:56:13.960 ⇒ 00:56:14.650 Awaish Kumar: Yeah.
456 00:56:14.650 ⇒ 00:56:15.360 Zoran Selinger: But.
457 00:56:15.360 ⇒ 00:56:20.749 Awaish Kumar: You can always do that. You can pick pick some tasks. You can ask for tasks. That’s okay.
458 00:56:21.230 ⇒ 00:56:29.160 Awaish Kumar: Yeah, until. And unless that’s something urgent client task, right? Which needs to be done anyway. But yeah, otherwise you can. We get.
459 00:56:30.470 ⇒ 00:56:36.719 Zoran Selinger: Yeah, that sound that that sounds great. I would like to.
460 00:56:40.510 ⇒ 00:56:49.634 Zoran Selinger: you know. Try and have a little bit of a sample work for you guys if that’s possible. For whatever you think is.
461 00:56:50.000 ⇒ 00:56:50.470 Awaish Kumar: Yeah, I know.
462 00:56:50.833 ⇒ 00:56:51.559 Zoran Selinger: For you.
463 00:56:51.560 ⇒ 00:56:55.378 Awaish Kumar: Like like Britain’s, have very
464 00:56:56.350 ⇒ 00:56:59.610 Awaish Kumar: great policy of hiring. And we
465 00:57:00.322 ⇒ 00:57:04.530 Awaish Kumar: we can like get some some of your availability, for example.
466 00:57:05.030 ⇒ 00:57:11.369 Awaish Kumar: few hours per week 10 h or 20 h. Whatever fits for.
467 00:57:11.370 ⇒ 00:57:13.219 Zoran Selinger: We’ll do so as well.
468 00:57:13.220 ⇒ 00:57:14.789 Awaish Kumar: Then we can work like
469 00:57:15.219 ⇒ 00:57:21.209 Awaish Kumar: work. For, for example, for 2 weeks or 4 weeks, and and we can get to know each other right?
470 00:57:21.900 ⇒ 00:57:25.408 Awaish Kumar: Once we are doing real work, like we know, like
471 00:57:26.230 ⇒ 00:57:29.559 Awaish Kumar: how how would you add or like, how good we are for you?
472 00:57:30.390 ⇒ 00:57:38.699 Zoran Selinger: Yeah, yeah, yeah, I’m yeah. That sounds good I’d like, I’d like to try you.
473 00:57:39.080 ⇒ 00:57:43.020 Zoran Selinger: I like what you’re doing. Seems like
474 00:57:43.130 ⇒ 00:57:46.819 Zoran Selinger: a group of really smart people, and that’s
475 00:57:47.330 ⇒ 00:57:53.939 Zoran Selinger: that’s obviously where everyone should be happy to be with. So
476 00:57:54.700 ⇒ 00:58:01.869 Zoran Selinger: I’d like to get a sample of of work, and maybe maybe put in 1020 h one of the weeks.
477 00:58:02.760 ⇒ 00:58:05.330 Zoran Selinger: and see if I can contribute.
478 00:58:05.620 ⇒ 00:58:09.500 Zoran Selinger: You know, when there’s something appropriate, right?
479 00:58:10.370 ⇒ 00:58:13.294 Awaish Kumar: Okay, yeah, sure. No problem. I think
480 00:58:14.000 ⇒ 00:58:17.066 Awaish Kumar: it, it was a nice conversation. I would.
481 00:58:17.949 ⇒ 00:58:30.849 Awaish Kumar: like, we’ll be the team and then we are hoping that we can give you feedback by end of week, like whatever the decision is. And our team Rico from our operations team is going to
482 00:58:31.000 ⇒ 00:58:31.590 Awaish Kumar: cool.
483 00:58:32.590 ⇒ 00:58:33.650 Awaish Kumar: Get in touch with you.
484 00:58:33.890 ⇒ 00:58:34.690 Zoran Selinger: Sure, sure.
485 00:58:35.050 ⇒ 00:58:36.179 Awaish Kumar: Thank you.
486 00:58:36.180 ⇒ 00:58:38.020 Zoran Selinger: I appreciate it. Thank you for your time.
487 00:58:38.590 ⇒ 00:58:39.360 Zoran Selinger: Bye-bye.