Meeting Title: ICustomer Proposal Date: 2025-09-22 Meeting participants: Awaish Kumar, Uttam Kumaran, Justin Breshears
WEBVTT
1 00:02:13.810 ⇒ 00:02:15.060 Uttam Kumaran: Hello!
2 00:02:20.730 ⇒ 00:02:21.340 Awaish Kumar: I don’t…
3 00:02:22.200 ⇒ 00:02:23.380 Uttam Kumaran: How’s everything?
4 00:02:23.860 ⇒ 00:02:24.640 Awaish Kumar: I’m good.
5 00:02:41.030 ⇒ 00:02:41.960 Uttam Kumaran: Beauty.
6 00:03:50.420 ⇒ 00:03:53.629 Uttam Kumaran: One second, let me just finish something, and then I’ll start.
7 00:03:59.560 ⇒ 00:04:00.610 Justin Breshears: Hello, hello!
8 00:04:01.670 ⇒ 00:04:02.900 Uttam Kumaran: Hello!
9 00:04:05.230 ⇒ 00:04:07.610 Justin Breshears: Boysh, good to see you, and…
10 00:04:08.120 ⇒ 00:04:13.819 Justin Breshears: in all of your glory, I haven’t seen your face before, but it’s nice to get a chance to meet you, and…
11 00:04:14.030 ⇒ 00:04:17.869 Justin Breshears: hang out with you a little bit. I think I’ve only seen you, kind of, on group meetings.
12 00:04:19.079 ⇒ 00:04:19.659 Awaish Kumar: Nope.
13 00:04:21.860 ⇒ 00:04:22.530 Justin Breshears: Oh.
14 00:04:23.060 ⇒ 00:04:24.440 Justin Breshears: I don’t think I can hear you.
15 00:04:26.060 ⇒ 00:04:27.070 Awaish Kumar: You can’t hear me?
16 00:04:27.470 ⇒ 00:04:33.500 Justin Breshears: Oh, now I can. Okay, maybe just a glitch. My laptop is struggling with Zoom today, so… Apologies.
17 00:04:34.640 ⇒ 00:04:35.550 Uttam Kumaran: All good.
18 00:04:35.970 ⇒ 00:04:41.070 Uttam Kumaran: I’m just like… Monday is fun day for me, so I’m just, like, trying to…
19 00:04:41.430 ⇒ 00:04:47.719 Uttam Kumaran: get some things going that I don’t have to be part of, and then we can talk about iCustomer.
20 00:04:47.890 ⇒ 00:04:55.029 Awaish Kumar: I have shared an AI-generated Diagram in the sections.
21 00:04:55.930 ⇒ 00:04:56.570 Uttam Kumaran: Okay.
22 00:04:57.160 ⇒ 00:05:03.800 Uttam Kumaran: Let me… So, let me give you a sense of, like, where we are with Hi, customer,
23 00:05:05.260 ⇒ 00:05:19.099 Uttam Kumaran: So they asked us for… they sent us an RFP, I sent them back a version, I missed a couple things, kind of, like, figured it out, but mainly they didn’t tell me… they didn’t really give me accurately, like.
24 00:05:20.000 ⇒ 00:05:32.319 Uttam Kumaran: how long they want this to take, so they basically were like, we needed 30 days, blah blah blah. So, 30 days is quite a lot for this amount of work. I would think you’re probably thinking that too, Awash, but let’s see, I mean…
25 00:05:32.580 ⇒ 00:05:44.660 Uttam Kumaran: I’m gonna just increase the price if that’s… they want it fast. So, kind of like where we were in our proposal, and I’ll just share,
26 00:05:45.170 ⇒ 00:05:49.860 Uttam Kumaran: This Notion doc is what I shared with them last time.
27 00:05:50.620 ⇒ 00:05:56.130 Uttam Kumaran: So I’ll send this in the channel…
28 00:05:56.820 ⇒ 00:06:00.490 Uttam Kumaran: This is what I sent to them last time.
29 00:06:05.590 ⇒ 00:06:14.209 Uttam Kumaran: And so, what they asked for was, yeah, like, one or two diagrams, work plans, hours and rate, assumptions and risk.
30 00:06:14.320 ⇒ 00:06:17.870 Uttam Kumaran: Quality plan, relevant samples, start date, a little…
31 00:06:17.970 ⇒ 00:06:23.659 Uttam Kumaran: start date availability. So, I think for a way… for me and you to think about is…
32 00:06:24.500 ⇒ 00:06:26.289 Uttam Kumaran: 1 and 2?
33 00:06:27.160 ⇒ 00:06:29.930 Uttam Kumaran: And make sure that that’s… Accurate.
34 00:06:30.190 ⇒ 00:06:35.739 Uttam Kumaran: And then… or, like, 1, 2, and 5?
35 00:06:36.140 ⇒ 00:06:41.080 Uttam Kumaran: And then I think, Justin, where we kind of overlap is, like, 2, 3.
36 00:06:41.620 ⇒ 00:06:45.559 Uttam Kumaran: and… 4?
37 00:06:46.130 ⇒ 00:06:51.250 Uttam Kumaran: But I guess… Yeah, like, I think…
38 00:06:51.450 ⇒ 00:06:55.590 Uttam Kumaran: their… their RFP was completely AI-generated, so…
39 00:06:55.810 ⇒ 00:07:11.479 Uttam Kumaran: I’m kind of like… I think they just want to make sure we cross all the boxes. Not sure if they’re, like, they have some sort of structured process for RFPs, but I guess… is everybody kind of comfortable with, like, the scope of this right now before we sort of try to dive in?
40 00:07:13.760 ⇒ 00:07:23.600 Justin Breshears: So I’m trying to pull up the notion. I’m working off of my wife’s 5-year-old laptop, and I’m discovering that it may no longer be viable at this point.
41 00:07:23.600 ⇒ 00:07:25.150 Uttam Kumaran: Okay, okay, alright.
42 00:07:25.150 ⇒ 00:07:29.959 Justin Breshears: Because I have… I have my desktop that I’ve been working on, the last couple of weeks, but…
43 00:07:30.410 ⇒ 00:07:35.669 Justin Breshears: you know, if something happens, or I need to be mobile, or whatever, like, I was like, okay, let’s get a laptop set up.
44 00:07:35.670 ⇒ 00:07:41.489 Uttam Kumaran: For the most part, we’re all cloud-based, so there’s not, like, anything that’s very file-based, but yeah, like.
45 00:07:41.670 ⇒ 00:07:44.520 Uttam Kumaran: Notion can be memory intensive, and.
46 00:07:44.990 ⇒ 00:07:52.409 Justin Breshears: It’s Zoom that’s killing me. I’ve been fine, except for being on Zoom. I’m looking at my CPU usage, and it’s, like, 95% on Zoom. Yeah.
47 00:07:53.620 ⇒ 00:08:01.690 Uttam Kumaran: Yeah, I have a mirror that I usually do if I’m mobile. I have to, like, quit… I just quit everything, basically. Yeah. Yeah.
48 00:08:02.630 ⇒ 00:08:05.089 Justin Breshears: We’re gonna figure that out. Okay, finally pulled up.
49 00:08:06.610 ⇒ 00:08:09.439 Awaish Kumar: Okay, have you seen this diagram I shared?
50 00:08:09.940 ⇒ 00:08:12.510 Uttam Kumaran: Let me look at it. So this is for… okay.
51 00:08:15.300 ⇒ 00:08:21.870 Uttam Kumaran: So, I guess to give Justin some context, it’s a little bit of a technical prop… it’s more technical than…
52 00:08:22.430 ⇒ 00:08:25.540 Uttam Kumaran: Our other clients, for the most part.
53 00:08:25.900 ⇒ 00:08:28.660 Awaish Kumar: Basically, to give you, like, the gist.
54 00:08:28.700 ⇒ 00:08:31.179 Uttam Kumaran: They want to package a set of…
55 00:08:32.240 ⇒ 00:08:37.729 Uttam Kumaran: SQL models and Python code into something that they can
56 00:08:37.850 ⇒ 00:08:54.420 Uttam Kumaran: deliver to clients through the Snowflake marketplace. Compared to some of our other clients, we develop it for them in-house. They themselves are a product company, and this is their product. They’re having us help them develop that.
57 00:08:54.580 ⇒ 00:09:04.679 Uttam Kumaran: So, they have a couple things that are unique about this project, is one, the Snowflake marketplace piece, is, like, how do we build this solution within Snowflake?
58 00:09:04.680 ⇒ 00:09:16.440 Uttam Kumaran: and get… and they are using Snowflake as the method of distribution for these sets of files and… and processes. And so there are some… that is where I think the unique
59 00:09:16.810 ⇒ 00:09:23.719 Uttam Kumaran: complexity comes across in this. All of that is sort of documented in the existing scope.
60 00:09:23.920 ⇒ 00:09:31.010 Uttam Kumaran: And so if we just talk about, like, the architecture, so do you want to walk me through this diagram, Awash?
61 00:09:31.890 ⇒ 00:09:35.680 Awaish Kumar: Yeah, like, what I’m, thinking about…
62 00:09:35.860 ⇒ 00:09:38.450 Awaish Kumar: Is that, like, we will have some…
63 00:09:40.330 ⇒ 00:09:47.430 Awaish Kumar: Like, source data, which will… maybe we will generate it, or… downloaded from somewhere.
64 00:09:47.530 ⇒ 00:09:58.900 Awaish Kumar: as a sample, then it go to dbt models. So, there are two ways, to handle that. Number one is that we have a dbt project which will run through
65 00:09:59.010 ⇒ 00:10:02.950 Awaish Kumar: GitHub CICD and… and pre-build the models.
66 00:10:03.120 ⇒ 00:10:10.690 Awaish Kumar: And then, basically, the… the sec… the Snowflake app, which will read that tables only.
67 00:10:10.840 ⇒ 00:10:14.080 Awaish Kumar: And then maybe use some more…
68 00:10:14.210 ⇒ 00:10:22.090 Awaish Kumar: tables which are, there in the Snowflake database itself about stage, or signals.
69 00:10:22.220 ⇒ 00:10:26.740 Awaish Kumar: And then join these tables in the app, using some…
70 00:10:27.000 ⇒ 00:10:36.479 Awaish Kumar: like, tasks, and then, basically, it will create the Stimulate app also on the same Snowflake-based URL.
71 00:10:37.860 ⇒ 00:10:44.850 Awaish Kumar: Second thing is that, which route… which I get from their name, which is native Slovak Native Packaging App.
72 00:10:45.090 ⇒ 00:10:54.770 Awaish Kumar: That means they don’t want to maybe use dbt and CICD thing, so basically, we maybe use it for development, but at the end, all these queries will just
73 00:10:55.220 ⇒ 00:11:00.270 Awaish Kumar: Go into that app, and will be executed as a task.
74 00:11:00.850 ⇒ 00:11:05.859 Uttam Kumaran: I guess here’s where I would probably want us to try something, is, like, dbt… Snowflake shipped
75 00:11:05.960 ⇒ 00:11:11.540 Uttam Kumaran: dbt, directly within Snowflake, so you can now run.
76 00:11:12.150 ⇒ 00:11:12.830 Awaish Kumar: Yeah, but…
77 00:11:12.830 ⇒ 00:11:13.640 Uttam Kumaran: entirely.
78 00:11:14.110 ⇒ 00:11:18.909 Uttam Kumaran: So I… we can explore this, basically running dbt Core directly in SnowBank.
79 00:11:19.790 ⇒ 00:11:21.900 Awaish Kumar: Yes, but I’m… but I’m…
80 00:11:22.230 ⇒ 00:11:31.750 Awaish Kumar: Okay, I haven’t looked into it, like, how that app will be working, like, will that app will be able to communicate with the dbt?
81 00:11:32.830 ⇒ 00:11:35.389 Awaish Kumar: project in the Salt Lake, or…
82 00:11:36.390 ⇒ 00:11:37.209 Awaish Kumar: Let me go.
83 00:11:40.630 ⇒ 00:11:41.270 Uttam Kumaran: Okay.
84 00:11:42.680 ⇒ 00:11:46.430 Awaish Kumar: Yeah, this is where, like, I’m not sure, like, are people accessing…
85 00:11:48.650 ⇒ 00:11:52.380 Uttam Kumaran: This is where I think they’re… I don’t think they’re actually gonna ship this.
86 00:11:52.950 ⇒ 00:11:54.460 Awaish Kumar: Like, all of this…
87 00:11:54.510 ⇒ 00:11:56.010 Uttam Kumaran: is gonna happen.
88 00:11:57.150 ⇒ 00:11:58.749 Uttam Kumaran: Within their environment.
89 00:11:59.610 ⇒ 00:12:00.310 Awaish Kumar: Yes.
90 00:12:00.490 ⇒ 00:12:11.209 Uttam Kumaran: the source data is all that we’re going to be given, and that… I guess we should see, like, whether part of that… which parts are going to get executed on a client environment, and which parts are going to execute it
91 00:12:11.890 ⇒ 00:12:13.940 Uttam Kumaran: Server side, basically.
92 00:12:15.170 ⇒ 00:12:15.950 Uttam Kumaran: Right.
93 00:12:17.080 ⇒ 00:12:20.740 Awaish Kumar: But that all is going to be inside their environment, right?
94 00:12:23.430 ⇒ 00:12:27.899 Uttam Kumaran: I don’t know. This is where I’m saying, like, I don’t know how much of this is…
95 00:12:28.320 ⇒ 00:12:31.430 Uttam Kumaran: Within their environment versus, like, server-side.
96 00:12:31.980 ⇒ 00:12:35.620 Awaish Kumar: Yeah, but what we can propose is something which is…
97 00:12:35.740 ⇒ 00:12:50.440 Awaish Kumar: like, everything is in their environment, which is, like, creating an app, which is there, and Streamlit UI is also Snowflake hosted, and then there are some SQL queries, which
98 00:12:50.890 ⇒ 00:12:55.660 Awaish Kumar: Will be, like, some… like, we can just copy-paste the queries inside of…
99 00:12:56.140 ⇒ 00:13:00.319 Awaish Kumar: That, as, as… and execute them as tasks.
100 00:13:00.440 ⇒ 00:13:02.759 Awaish Kumar: So everything will just go in there.
101 00:13:04.170 ⇒ 00:13:04.920 Uttam Kumaran: Okay.
102 00:13:07.820 ⇒ 00:13:08.640 Uttam Kumaran: Okay.
103 00:13:13.090 ⇒ 00:13:17.629 Uttam Kumaran: Do we have any of those… do we have any of the… I don’t think we have any of those…
104 00:13:18.850 ⇒ 00:13:20.529 Awaish Kumar: As simple, smooth.
105 00:13:20.820 ⇒ 00:13:23.209 Uttam Kumaran: No, no, no, not examples, but, like.
106 00:13:23.840 ⇒ 00:13:26.479 Uttam Kumaran: I guess, what do they say in their email?
107 00:13:29.380 ⇒ 00:13:36.580 Uttam Kumaran: Yeah, so I think I want to kind of have a section for architecture,
108 00:13:38.000 ⇒ 00:13:42.020 Uttam Kumaran: So, roughly, we have these… these are sort of some… some risks.
109 00:13:42.730 ⇒ 00:13:49.290 Uttam Kumaran: I don’t know what they said, your dock is not working, it’s…
110 00:13:50.390 ⇒ 00:13:54.739 Uttam Kumaran: It’s a confirm timeline, so this one they said 30 days.
111 00:13:54.970 ⇒ 00:13:58.899 Uttam Kumaran: Let’s have a quick section here on architecture.
112 00:14:03.770 ⇒ 00:14:14.270 Uttam Kumaran: So, can you put your, your image in, and then… Yeah, basically…
113 00:14:14.640 ⇒ 00:14:21.779 Uttam Kumaran: I think we’re gonna talk through… so they kind of wanted to talk about, like, these concepts, right? Dataflow, UDFs.
114 00:14:22.020 ⇒ 00:14:24.830 Uttam Kumaran: streams… Tasks.
115 00:14:25.420 ⇒ 00:14:30.490 Uttam Kumaran: events, tables… Stream lit.
116 00:14:31.610 ⇒ 00:14:33.709 Uttam Kumaran: And native app, right?
117 00:14:34.570 ⇒ 00:14:35.360 Awaish Kumar: Okay.
118 00:14:36.970 ⇒ 00:14:38.740 Awaish Kumar: So… Freedom.
119 00:14:38.740 ⇒ 00:14:39.350 Uttam Kumaran: Yeah.
120 00:14:39.750 ⇒ 00:14:44.240 Awaish Kumar: That is app flow, right? The image we are showing is complete app flow.
121 00:14:45.640 ⇒ 00:14:54.130 Awaish Kumar: inside of it is, like, first part is the data flow, where we are saying dbt signals data, and will we be running UDFs, or…
122 00:14:54.300 ⇒ 00:14:57.679 Awaish Kumar: We will be running them using dbt, or things like that.
123 00:14:57.950 ⇒ 00:15:06.429 Uttam Kumaran: I guess just what I want to make sure is that as part of the proposal, we hit each of these pieces, both visually and via text.
124 00:15:06.550 ⇒ 00:15:13.579 Uttam Kumaran: So, for example, in Dataflow, we can put, like, basically, like, what is the goal?
125 00:15:13.700 ⇒ 00:15:22.399 Uttam Kumaran: And then, like, how, right? Like, so, goal of the data flow component is XYZ. How is, like, we’re using these specific tools within Snowflake.
126 00:15:23.870 ⇒ 00:15:30.890 Uttam Kumaran: For example, for, for streams and tasks, we could do…
127 00:15:31.000 ⇒ 00:15:36.750 Uttam Kumaran: utilize native dbt project, right? So this is the how.
128 00:15:36.940 ⇒ 00:15:42.410 Uttam Kumaran: And then the goal here is… orchestration…
129 00:15:43.260 ⇒ 00:15:49.190 Uttam Kumaran: Of data model and data ingestion Close.
130 00:15:50.250 ⇒ 00:15:52.179 Uttam Kumaran: Within no place, right?
131 00:15:53.610 ⇒ 00:15:56.859 Uttam Kumaran: So if you can do this, the goal how, for each…
132 00:15:56.990 ⇒ 00:15:58.370 Uttam Kumaran: I think we’re kind of good.
133 00:15:59.710 ⇒ 00:16:11.360 Uttam Kumaran: And then I think I would like to just continue to put, like, open questions wherever we have open questions, one of which is how much is client-side versus how much is server-side.
134 00:16:12.470 ⇒ 00:16:14.329 Uttam Kumaran: That’s something that I want to find out.
135 00:16:15.150 ⇒ 00:16:18.070 Awaish Kumar: And, like, what exactly do you mean by…
136 00:16:18.070 ⇒ 00:16:20.870 Uttam Kumaran: Like, how much is gonna get run on the…
137 00:16:21.470 ⇒ 00:16:33.270 Uttam Kumaran: for example, you can actually… when you install a native app in Snowflake, it will… it can actually install UDFs and things on the client side, right? But you can also have callbacks to…
138 00:16:33.920 ⇒ 00:16:46.489 Uttam Kumaran: iCustomer-related APIs. For example, there are a lot of native apps where all they say is, pick your table, and then all the processing happens in iCustomer, and then iCustomer
139 00:16:46.700 ⇒ 00:17:06.339 Uttam Kumaran: sends back a function that creates a table. The logic is all in iCustomer. You can also do it where the logic is within the client, right? Where when you install the native app, it comes with all these things, and so it’s completely running on the client infrastructure. So that’s what you want to decide what’s going to be running on the client.
140 00:17:06.390 ⇒ 00:17:09.470 Uttam Kumaran: Snowflake instance, and what’s gonna be running
141 00:17:09.550 ⇒ 00:17:11.420 Uttam Kumaran: Like, what’s gonna be an external call?
142 00:17:14.230 ⇒ 00:17:15.180 Awaish Kumar: Okay.
143 00:17:15.650 ⇒ 00:17:17.740 Uttam Kumaran: And, like, for example, shared back somehow.
144 00:17:20.630 ⇒ 00:17:32.660 Awaish Kumar: Okay, I understand. So, now we are thinking about, like, for example, if this app goes into Marketplace, and somebody installs it, where it runs, like, on the…
145 00:17:32.810 ⇒ 00:17:41.250 Awaish Kumar: Yes. Yeah, machines of the person who installed it, or on the instance, yep, okay.
146 00:17:46.010 ⇒ 00:17:46.840 Uttam Kumaran: Okay.
147 00:17:46.990 ⇒ 00:17:50.510 Uttam Kumaran: Great, so if you want to take care of this, I feel pretty good on…
148 00:17:51.460 ⇒ 00:17:54.540 Uttam Kumaran: I feel pretty good on the architecture,
149 00:17:55.780 ⇒ 00:17:58.439 Uttam Kumaran: And they can critique this and leave comments.
150 00:17:58.550 ⇒ 00:18:07.579 Uttam Kumaran: Let’s talk about, like… I mean, I feel like there are a lot of open questions here, like.
151 00:18:07.870 ⇒ 00:18:14.280 Uttam Kumaran: I just… this is where I’m kind of nervous, like, they have a very aggressive timeline, And…
152 00:18:14.850 ⇒ 00:18:22.739 Uttam Kumaran: they didn’t answer, like, any of my questions, so I’m like, I don’t want to do much more work until they, like, get back to us with, like, some answers.
153 00:18:22.740 ⇒ 00:18:23.320 Justin Breshears: This was…
154 00:18:23.500 ⇒ 00:18:34.110 Justin Breshears: That’s exactly where my biggest concerns are on this, because I’m like, 30 days for all of this, do we even have, like, the bandwidth to be able to tackle that as a team?
155 00:18:34.110 ⇒ 00:18:34.460 Uttam Kumaran: Yeah.
156 00:18:34.460 ⇒ 00:18:36.650 Justin Breshears: So the resourcing… So that’s kind of the…
157 00:18:37.300 ⇒ 00:18:38.320 Uttam Kumaran: Yeah.
158 00:18:38.810 ⇒ 00:18:41.840 Uttam Kumaran: That’s why, when he was like, I went over a few times.
159 00:18:42.300 ⇒ 00:18:46.380 Uttam Kumaran: are you adding additional info? And I said, yeah, but, like.
160 00:18:46.540 ⇒ 00:18:54.480 Uttam Kumaran: You said 30 days, is that how are you really… so, this is where, for me, like, this is a lot to get done in 30 days.
161 00:18:54.880 ⇒ 00:18:57.460 Uttam Kumaran: But, again, it’s like…
162 00:18:58.140 ⇒ 00:19:02.699 Uttam Kumaran: if they’re gonna pay a ton of money, then yeah, I can get his resource, for sure.
163 00:19:02.830 ⇒ 00:19:06.160 Uttam Kumaran: So, for us.
164 00:19:06.160 ⇒ 00:19:11.519 Justin Breshears: What are you thinking, like, structuring it, TNM, or on a fixed?
165 00:19:13.560 ⇒ 00:19:16.969 Uttam Kumaran: I would structure it like T&M.
166 00:19:17.610 ⇒ 00:19:19.220 Justin Breshears: I think that’d be better, yeah.
167 00:19:21.120 ⇒ 00:19:25.890 Uttam Kumaran: Yeah, I would probably structure a T&M. This seems like kind of a large thing.
168 00:19:26.710 ⇒ 00:19:32.219 Justin Breshears: And it protects us, too, if it becomes more complex, based on, like, stuff maybe they haven’t quite shared with us.
169 00:19:32.220 ⇒ 00:19:37.009 Uttam Kumaran: Yeah, so I would basically split it out, like, you have a data engineer, solution architect, PM,
170 00:19:37.370 ⇒ 00:19:39.089 Uttam Kumaran: And here are the… here are the rates.
171 00:19:40.850 ⇒ 00:19:43.129 Awaish Kumar: I just want to add one more thing, like…
172 00:19:43.270 ⇒ 00:19:49.910 Awaish Kumar: Yeah, we are putting some effort onto Answering these questions, or…
173 00:19:50.350 ⇒ 00:19:58.640 Awaish Kumar: building some visuals, but if they ask for more diagrams, I think we… we can make them as part of deliverables as well.
174 00:19:58.640 ⇒ 00:20:02.329 Uttam Kumaran: No, I’ll say, if you want us to do more, like, we have to do paid discovery, basically.
175 00:20:02.550 ⇒ 00:20:03.600 Awaish Kumar: Yeah, okay.
176 00:20:03.600 ⇒ 00:20:04.180 Uttam Kumaran: Yeah.
177 00:20:04.330 ⇒ 00:20:07.770 Uttam Kumaran: Because this is… this is quite a bit beyond…
178 00:20:08.080 ⇒ 00:20:13.379 Uttam Kumaran: I mean, like, again, I don’t have conf… like, I met them once, and we’re having this email back and forth.
179 00:20:13.720 ⇒ 00:20:16.450 Uttam Kumaran: But… I have been burned.
180 00:20:16.820 ⇒ 00:20:23.050 Uttam Kumaran: a lot… way too many times on stuff like this, so I… I really… I really want to see that they’re, like.
181 00:20:23.350 ⇒ 00:20:25.460 Uttam Kumaran: Okay, we’re down to move forward.
182 00:20:25.730 ⇒ 00:20:36.549 Uttam Kumaran: Ideally, if we’re gonna spend more hours on this proposal, I would love for them to pay for, like, a paid discovery, or we need some type of, like, commitment, basically, that, like, this is gonna go.
183 00:20:36.870 ⇒ 00:20:41.949 Uttam Kumaran: This is, like, this is the level of detail we do…
184 00:20:42.780 ⇒ 00:20:45.869 Uttam Kumaran: Like, after the thing is signed, usually, you know.
185 00:20:48.780 ⇒ 00:20:49.400 Awaish Kumar: Yep.
186 00:20:49.400 ⇒ 00:20:51.509 Uttam Kumaran: I don’t know, what do you think? What do you think, Justin?
187 00:20:52.160 ⇒ 00:20:56.850 Justin Breshears: Yeah, you… I think you draw the line here. I’m 100% in agreement with that.
188 00:20:57.480 ⇒ 00:21:04.980 Justin Breshears: You’ll get free work in a proposal. Yeah, yeah, I mean, yeah. I mean, they’re already asking for, like, architecture diagrams I’ve never…
189 00:21:05.010 ⇒ 00:21:07.469 Awaish Kumar: Never seen anybody provide that.
190 00:21:07.490 ⇒ 00:21:10.630 Justin Breshears: Prior to… contract signing.
191 00:21:10.820 ⇒ 00:21:11.440 Uttam Kumaran: Yeah.
192 00:21:12.850 ⇒ 00:21:17.490 Uttam Kumaran: But I guess this is what we’ll see, like… If they’re… yeah.
193 00:21:17.690 ⇒ 00:21:26.939 Uttam Kumaran: we’ll sort of see what the game is. So, I just want to get one more email to them with, like, we got you everything. I will make it very clear that, like.
194 00:21:27.140 ⇒ 00:21:30.280 Uttam Kumaran: This is quite a bit beyond what we do at this stage.
195 00:21:30.530 ⇒ 00:21:35.269 Uttam Kumaran: So, we need to have some type of commitment for us to spend more time on mapping this.
196 00:21:37.660 ⇒ 00:21:47.859 Justin Breshears: I think a paid discovery makes sense, because then you can fully draft up the contract, knowing all the details, because, like you said, there’s a ton of open questions still.
197 00:21:48.770 ⇒ 00:21:49.380 Uttam Kumaran: Yeah.
198 00:21:50.280 ⇒ 00:22:00.369 Uttam Kumaran: So in terms of, like, If we talk about, work plans and milestone.
199 00:22:01.110 ⇒ 00:22:07.319 Uttam Kumaran: basically… We talked about these kind of deliverables, they felt okay with it.
200 00:22:09.020 ⇒ 00:22:19.379 Uttam Kumaran: I guess my question was more of, like, I also asked, basically, we’re like, do we need to staff all this? Are they staffing all this? Which, again, they’ve left no comments here, so…
201 00:22:19.600 ⇒ 00:22:25.489 Uttam Kumaran: I’ll… I’ll list that in my feedback to them. Second is…
202 00:22:26.030 ⇒ 00:22:32.999 Uttam Kumaran: I… I mean, I… I think this is too much to get done within 30 days, Awash, so tell me what you think
203 00:22:33.340 ⇒ 00:22:41.340 Uttam Kumaran: even if I were to get you one or two more people, like, what do you… what do you think… I have an idea, but what do you think is possible within 30 days?
204 00:22:44.080 ⇒ 00:22:46.979 Uttam Kumaran: Like, I don’t… yeah, you go, and then I’ll go.
205 00:22:47.430 ⇒ 00:22:55.859 Awaish Kumar: Yeah, I will build the app, but I think, like, building… making a… like, what I’ve read is, sort of, like, provide some…
206 00:22:56.400 ⇒ 00:23:01.260 Awaish Kumar: like, the… Template app, which can… you can just…
207 00:23:01.380 ⇒ 00:23:10.360 Awaish Kumar: click and use it. So, app with some dbt projects, some modeling, obviously can be done very easily.
208 00:23:11.360 ⇒ 00:23:14.590 Awaish Kumar: like, I think all the modeling work and all the…
209 00:23:14.780 ⇒ 00:23:21.740 Awaish Kumar: like, all these signals and all that stuff could be done. I’m only concerned about the stream lift part of it.
210 00:23:23.880 ⇒ 00:23:29.650 Uttam Kumaran: Mmm, really? Okay, so I’m more… I’m more concerned about the snowflake stuff, because…
211 00:23:31.290 ⇒ 00:23:36.940 Uttam Kumaran: it’s kind of new, like, setting up the dbt and stuff like that is fine.
212 00:23:38.310 ⇒ 00:23:41.630 Uttam Kumaran: But
213 00:23:41.740 ⇒ 00:23:52.789 Uttam Kumaran: like, getting a Snowflake native app set up and in the marketplace can take quite a while, and getting… testing Streamlit on Snowflake is really painful.
214 00:23:52.790 ⇒ 00:23:55.619 Awaish Kumar: But they haven’t asked for, like, on the marketplace, right?
215 00:23:56.830 ⇒ 00:24:00.339 Uttam Kumaran: Oh, how… I don’t know how else they’re gonna distribute this, right? Like, oh, I guess…
216 00:24:00.340 ⇒ 00:24:02.380 Awaish Kumar: For the first, like, this time…
217 00:24:02.380 ⇒ 00:24:03.560 Uttam Kumaran: app.
218 00:24:04.000 ⇒ 00:24:11.649 Awaish Kumar: But for this 30 days implementation, they are more, like, build an app which can run, basically, like…
219 00:24:11.650 ⇒ 00:24:12.420 Uttam Kumaran: Oh, okay.
220 00:24:12.420 ⇒ 00:24:15.319 Awaish Kumar: Production, but just on one or two machines, or whatever.
221 00:24:15.930 ⇒ 00:24:17.489 Awaish Kumar: At least on the outside.
222 00:24:18.290 ⇒ 00:24:23.639 Uttam Kumaran: Okay, so let’s talk about, like, let’s say we set the… let’s say we set 30 days.
223 00:24:23.850 ⇒ 00:24:30.249 Uttam Kumaran: What do we think? So… I think, like, we have to basically either get
224 00:24:30.580 ⇒ 00:24:34.039 Uttam Kumaran: the data in Snowflake, or synthetic?
225 00:24:35.010 ⇒ 00:24:36.630 Uttam Kumaran: Generate, right?
226 00:24:36.890 ⇒ 00:24:37.380 Awaish Kumar: Yep.
227 00:24:37.380 ⇒ 00:24:41.910 Uttam Kumaran: Second is we, like, establish dbt project.
228 00:24:42.950 ⇒ 00:24:45.980 Uttam Kumaran: and build… link.
229 00:24:46.290 ⇒ 00:24:48.330 Awaish Kumar: a few MARTS models.
230 00:24:49.670 ⇒ 00:24:51.290 Uttam Kumaran: And then, what’s next?
231 00:24:53.210 ⇒ 00:24:57.140 Awaish Kumar: It’s a bit of a… Like, the snowflake app.
232 00:24:58.530 ⇒ 00:25:00.610 Awaish Kumar: Which can take up to 2 weeks.
233 00:25:02.100 ⇒ 00:25:02.670 Uttam Kumaran: Yeah.
234 00:25:02.670 ⇒ 00:25:03.350 Awaish Kumar: Sweet.
235 00:25:08.340 ⇒ 00:25:13.379 Justin Breshears: Is this a scenario where having more bodies thrown at it will speed up timeline, or no?
236 00:25:13.920 ⇒ 00:25:16.260 Uttam Kumaran: Yeah, like, we would… I mean, we’d definitely…
237 00:25:16.770 ⇒ 00:25:21.740 Uttam Kumaran: Like, we would have to have at least, two people on this that’s not me.
238 00:25:21.950 ⇒ 00:25:27.590 Uttam Kumaran: So, we’d have to go get… we’d get a data… we gotta bring on another data engineer for this project, for sure.
239 00:25:28.420 ⇒ 00:25:32.350 Uttam Kumaran: Which I could get, like, we can get and start someone pretty quickly.
240 00:25:32.600 ⇒ 00:25:37.640 Uttam Kumaran: But, like, I guess 30 days is actually just, like.
241 00:25:38.090 ⇒ 00:25:44.440 Uttam Kumaran: even if we were to have 10 people on this project, you wouldn’t be able to do it much faster. Like, 30 days is only, like, 30 days.
242 00:25:44.580 ⇒ 00:25:49.780 Uttam Kumaran: Like, a week early meeting in that time and stuff like that, right? So…
243 00:25:50.020 ⇒ 00:25:55.610 Uttam Kumaran: I think I agree, like, setting up a basic Snowflake app.
244 00:26:00.920 ⇒ 00:26:06.790 Justin Breshears: Are they cool with, like, a POC in 30 days, and then developing, like, a full… Solution?
245 00:26:07.330 ⇒ 00:26:11.989 Uttam Kumaran: Well, that’s what I’m hearing, Awasht, like, you’re kind of mentioning, is that this is just a…
246 00:26:12.390 ⇒ 00:26:15.030 Awaish Kumar: Yeah, demo app, like, proof of jobs.
247 00:26:15.030 ⇒ 00:26:15.620 Uttam Kumaran: Yeah.
248 00:26:18.720 ⇒ 00:26:26.109 Uttam Kumaran: Okay, so they said, ship a demo… so this is the goal. Ship a demo-ready, privately installable Snowflake native app. Okay, so no marketplace.
249 00:26:26.450 ⇒ 00:26:29.019 Uttam Kumaran: that runs inside the customer snowflake.
250 00:26:30.690 ⇒ 00:26:34.280 Uttam Kumaran: And provides value across these four things.
251 00:26:34.460 ⇒ 00:26:37.990 Uttam Kumaran: Zero external egress. Okay, so then they answered it, right?
252 00:26:38.200 ⇒ 00:26:43.999 Uttam Kumaran: So to answer my previous question, which is… It’s gonna be all client-side.
253 00:26:45.780 ⇒ 00:26:47.970 Uttam Kumaran: Right? Hind side.
254 00:26:48.380 ⇒ 00:26:49.370 Uttam Kumaran: private.
255 00:26:49.570 ⇒ 00:26:51.000 Uttam Kumaran: Snowflake app…
256 00:26:51.290 ⇒ 00:27:00.409 Awaish Kumar: But, so that means we can’t have a dbt project, but just the queries.
257 00:27:03.350 ⇒ 00:27:05.360 Uttam Kumaran: That, I don’t know.
258 00:27:05.530 ⇒ 00:27:09.720 Awaish Kumar: Because, like, if a client… a client installs an app.
259 00:27:09.980 ⇒ 00:27:16.469 Awaish Kumar: it can basically have queries and can run it, but I don’t think you can just go in and create a DVD project.
260 00:27:16.750 ⇒ 00:27:19.090 Awaish Kumar: Stuff like that. Automatically.
261 00:27:20.520 ⇒ 00:27:23.929 Uttam Kumaran: That, I don’t know, see, like, I don’t know, can you create the dbt project object?
262 00:27:27.490 ⇒ 00:27:28.740 Uttam Kumaran: Via the app.
263 00:27:29.540 ⇒ 00:27:32.000 Uttam Kumaran: Yeah, look, I mean, you can run… you can run that.
264 00:27:32.620 ⇒ 00:27:40.049 Awaish Kumar: Yeah, you can run it, but, like, the client has to create it. But I think they want it, like, plug-and-play kind of thing for clients.
265 00:27:40.050 ⇒ 00:27:43.600 Uttam Kumaran: Yeah, but like, why wouldn’t we just have the client create it?
266 00:27:49.020 ⇒ 00:27:53.060 Uttam Kumaran: You know what I mean? You can, like, you can just have… you can just have our app.
267 00:27:53.490 ⇒ 00:27:55.289 Uttam Kumaran: Run, create project.
268 00:27:57.630 ⇒ 00:27:59.150 Uttam Kumaran: And then create the models.
269 00:27:59.610 ⇒ 00:28:04.450 Uttam Kumaran: And then that way we don’t… because otherwise, like, yeah, I think that’s probably easiest.
270 00:28:04.880 ⇒ 00:28:07.059 Uttam Kumaran: Because then we just ship new projects.
271 00:28:09.490 ⇒ 00:28:11.310 Uttam Kumaran: We shook the entire project.
272 00:28:13.440 ⇒ 00:28:16.680 Uttam Kumaran: And then you run the… and then you… and you do execute… you just run execute.
273 00:28:21.300 ⇒ 00:28:24.950 Uttam Kumaran: I think it’s way easier than shipping a bunch of SQL and then having tasks.
274 00:28:25.330 ⇒ 00:28:26.489 Awaish Kumar: and UDFs.
275 00:28:26.590 ⇒ 00:28:29.360 Uttam Kumaran: Like, this is the wrapper around all the model logic.
276 00:28:31.560 ⇒ 00:28:32.160 Awaish Kumar: Okay.
277 00:28:33.190 ⇒ 00:28:34.620 Uttam Kumaran: I don’t know, that’s what I think.
278 00:28:34.620 ⇒ 00:28:38.769 Awaish Kumar: It has… has to go deeper into it, I don’t know, we don’t.
279 00:28:39.300 ⇒ 00:28:40.220 Uttam Kumaran: I mean, yeah, yeah.
280 00:28:47.080 ⇒ 00:28:48.570 Uttam Kumaran: Okay, so I’m just gonna s…
281 00:29:16.410 ⇒ 00:29:18.310 Uttam Kumaran: Okay.
282 00:29:19.350 ⇒ 00:29:22.250 Awaish Kumar: Like, but did you create this?
283 00:29:24.290 ⇒ 00:29:25.000 Uttam Kumaran: Huh?
284 00:29:25.450 ⇒ 00:29:28.149 Awaish Kumar: Like, the stable pre-kickoff requirements.
285 00:29:28.300 ⇒ 00:29:33.250 Awaish Kumar: It says, like, within 2 days of signing, you can… Did you eat enough food?
286 00:29:36.200 ⇒ 00:29:38.260 Uttam Kumaran: Hold on, let me check.
287 00:29:41.580 ⇒ 00:29:44.730 Uttam Kumaran: So these are, like, the core outcomes.
288 00:29:50.800 ⇒ 00:29:55.889 Uttam Kumaran: So… Basically, what we want to understand is, like, private native app.
289 00:29:58.110 ⇒ 00:30:00.430 Uttam Kumaran: this is just a dbt job, right?
290 00:30:02.430 ⇒ 00:30:07.249 Uttam Kumaran: This is a dbt job, this is a dbt job, this is a dbt job.
291 00:30:07.970 ⇒ 00:30:09.550 Uttam Kumaran: This is the stream lit.
292 00:30:11.030 ⇒ 00:30:16.760 Uttam Kumaran: This is… Streamlit, dbt job, Right?
293 00:30:17.370 ⇒ 00:30:23.170 Uttam Kumaran: So… this is gonna be the challenging thing. There’s not… I don’t think this happens in 30 days.
294 00:30:28.310 ⇒ 00:30:29.150 Uttam Kumaran: Right.
295 00:30:31.410 ⇒ 00:30:33.130 Awaish Kumar: So, I would probably say.
296 00:30:33.260 ⇒ 00:30:36.659 Uttam Kumaran: We could spin up, like, a demo of a private app.
297 00:30:37.500 ⇒ 00:30:43.200 Uttam Kumaran: But… Like, and show that it is installable.
298 00:30:43.430 ⇒ 00:30:48.650 Uttam Kumaran: but, like, to have it install, execute all the dbt, and… handle…
299 00:30:49.520 ⇒ 00:30:52.739 Uttam Kumaran: Orchestration and alerting? There’s no way.
300 00:30:55.190 ⇒ 00:30:55.950 Uttam Kumaran: Right.
301 00:30:57.770 ⇒ 00:30:58.660 Awaish Kumar: Okay, yep.
302 00:31:00.880 ⇒ 00:31:04.980 Uttam Kumaran: So, I would say this is, like, out of… this is probably, like, this is probably, like, 90 days.
303 00:31:05.790 ⇒ 00:31:12.130 Uttam Kumaran: And… In 30 days, we can… Create a demo.
304 00:31:12.730 ⇒ 00:31:14.760 Uttam Kumaran: Private app.
305 00:31:15.380 ⇒ 00:31:22.589 Uttam Kumaran: That can… I mean, this is where, like, That can execute dbt project?
306 00:31:24.090 ⇒ 00:31:26.549 Awaish Kumar: Yeah, I think so. We can do that.
307 00:31:28.660 ⇒ 00:31:36.219 Uttam Kumaran: So this would mean, like, you basically just have to have, like, a simple model that you can click run on Streamlit, and it runs the stuff.
308 00:31:39.460 ⇒ 00:31:40.250 Uttam Kumaran: Right?
309 00:31:41.370 ⇒ 00:31:43.390 Uttam Kumaran: Because then you can add more models.
310 00:31:44.090 ⇒ 00:31:47.639 Uttam Kumaran: and stuff like that, like, that’s probably what I would suggest.
311 00:31:47.900 ⇒ 00:31:54.790 Uttam Kumaran: And then, like, In 90 days, It could be distributable, And, like.
312 00:31:55.200 ⇒ 00:32:05.680 Uttam Kumaran: Yeah, so we want to create a private demo that execute Execute a dbt project, simple, streamlit UI that
313 00:32:06.080 ⇒ 00:32:08.649 Uttam Kumaran: Briggers, new BT job.
314 00:32:09.580 ⇒ 00:32:10.819 Awaish Kumar: And then…
315 00:32:11.540 ⇒ 00:32:16.970 Uttam Kumaran: run logs, Are logged in a table symbol, right?
316 00:32:18.980 ⇒ 00:32:23.339 Awaish Kumar: Can we divide it, like, 2 weeks, weekend, hour and a half?
317 00:32:24.080 ⇒ 00:32:29.709 Awaish Kumar: And next two weeks, we can have, like, DVT, with, some…
318 00:32:29.710 ⇒ 00:32:35.180 Uttam Kumaran: Well, I guess, yeah, you tell me, like, what is the, which, which, like, what’s the order of operations?
319 00:32:38.230 ⇒ 00:32:39.250 Awaish Kumar: Sorry?
320 00:32:39.730 ⇒ 00:32:41.400 Uttam Kumaran: What’s the order of operations?
321 00:32:43.160 ⇒ 00:32:44.889 Uttam Kumaran: Like, you would do the app first?
322 00:32:46.330 ⇒ 00:32:49.769 Awaish Kumar: Yeah, I can just, like, I would go for a spin-off.
323 00:32:50.010 ⇒ 00:32:52.290 Awaish Kumar: Go ahead and spin up the app first.
324 00:32:52.540 ⇒ 00:32:56.980 Awaish Kumar: Then I can create a DVD project, and…
325 00:32:58.200 ⇒ 00:33:01.689 Awaish Kumar: And then, handle the orchestration through the app.
326 00:33:02.810 ⇒ 00:33:11.990 Awaish Kumar: then I would go for… Like, adding extra things, like signals, data, and the… equity…
327 00:33:12.120 ⇒ 00:33:17.770 Awaish Kumar: Like, creating these three different audiences, so creating all the marks using all of this data.
328 00:33:18.710 ⇒ 00:33:21.660 Awaish Kumar: And then finally go for UI.
329 00:33:26.770 ⇒ 00:33:28.600 Uttam Kumaran: Yeah, like, I think this is 30 days.
330 00:33:31.290 ⇒ 00:33:35.530 Uttam Kumaran: The thing I don’t know, I don’t know about, like, this.
331 00:33:39.620 ⇒ 00:33:44.959 Uttam Kumaran: And certainly, like, we’re not gonna have the entire model written and tested and everything.
332 00:33:55.990 ⇒ 00:33:56.750 Uttam Kumaran: Okay.
333 00:33:58.890 ⇒ 00:34:04.750 Uttam Kumaran: I think it’s… I mean, I… I… yeah, I have a feeling it’s at least gotta be, like, probably 20 hours a week of work.
334 00:34:05.350 ⇒ 00:34:06.550 Uttam Kumaran: For 4 weeks.
335 00:34:06.790 ⇒ 00:34:07.550 Uttam Kumaran: Right?
336 00:34:08.060 ⇒ 00:34:08.780 Awaish Kumar: Yes.
337 00:34:10.710 ⇒ 00:34:11.679 Uttam Kumaran: Anything more?
338 00:34:15.620 ⇒ 00:34:17.040 Awaish Kumar: I, like…
339 00:34:20.420 ⇒ 00:34:25.009 Awaish Kumar: 20 hours, that means just, like, 4 hours per day.
340 00:34:25.770 ⇒ 00:34:27.350 Awaish Kumar: With Doug?
341 00:34:28.310 ⇒ 00:34:29.110 Awaish Kumar: Fair enough.
342 00:34:30.429 ⇒ 00:34:33.680 Awaish Kumar: Spin up an app, and get a DVD plugin.
343 00:34:35.909 ⇒ 00:34:38.449 Awaish Kumar: Yeah, it can go more than that.
344 00:34:41.840 ⇒ 00:34:44.060 Justin Breshears: I wouldn’t undersell this one.
345 00:34:44.060 ⇒ 00:34:44.730 Uttam Kumaran: Yeah, yeah.
346 00:34:44.739 ⇒ 00:34:45.959 Justin Breshears: What it smells like.
347 00:34:46.600 ⇒ 00:34:48.380 Uttam Kumaran: So, and then,
348 00:34:48.580 ⇒ 00:34:55.010 Uttam Kumaran: I guess this is what we can talk about. If we’re gonna do TNM, then I can just basically put, like, it’s gonna be at least 80 hours.
349 00:34:55.830 ⇒ 00:35:00.199 Uttam Kumaran: Could be at least 80 hours from, like, senior Data Engineer.
350 00:35:00.950 ⇒ 00:35:05.660 Uttam Kumaran: And then… we should loop in PM time.
351 00:35:07.090 ⇒ 00:35:08.240 Uttam Kumaran: And then…
352 00:35:11.270 ⇒ 00:35:15.260 Uttam Kumaran: I guess I should also budget for, like, Solution Architect, like, which could be me.
353 00:35:17.120 ⇒ 00:35:19.180 Uttam Kumaran: And then… Yeah.
354 00:35:19.750 ⇒ 00:35:21.539 Awaish Kumar: So what are you suggesting, like.
355 00:35:22.370 ⇒ 00:35:26.030 Awaish Kumar: Like, having this demo app in 4 weeks or 6 weeks?
356 00:35:26.640 ⇒ 00:35:27.460 Uttam Kumaran: 4…
357 00:35:28.220 ⇒ 00:35:29.410 Awaish Kumar: So, nice.
358 00:35:31.820 ⇒ 00:35:35.280 Uttam Kumaran: But again, all we’re gonna promise is that it runs on their instance.
359 00:35:35.910 ⇒ 00:35:36.600 Awaish Kumar: Yeah.
360 00:35:37.180 ⇒ 00:35:37.770 Uttam Kumaran: Yeah.
361 00:35:41.280 ⇒ 00:35:43.839 Justin Breshears: Is there an option for Solutions Architect that’s not you?
362 00:35:47.570 ⇒ 00:35:52.279 Uttam Kumaran: Depends on… if they sign it, and depends on if I have 5 days, yeah, I could go get somebody, but…
363 00:35:54.300 ⇒ 00:35:56.850 Uttam Kumaran: I need at least 2 days, I can get somebody.
364 00:35:57.480 ⇒ 00:35:59.470 Uttam Kumaran: 48 hours less is…
365 00:35:59.890 ⇒ 00:36:02.270 Awaish Kumar: Tough. I mean, we’ll pay for that.
366 00:36:02.340 ⇒ 00:36:06.240 Uttam Kumaran: But, if I have 5 days, yeah, I can get anybody within 5 days.
367 00:36:06.400 ⇒ 00:36:09.770 Uttam Kumaran: So, that’s… that’s like… you could leave that to me.
368 00:36:11.260 ⇒ 00:36:11.820 Justin Breshears: Right.
369 00:36:14.300 ⇒ 00:36:16.890 Justin Breshears: That’s how I’m trying to think about this, is let’s.
370 00:36:16.890 ⇒ 00:36:17.330 Uttam Kumaran: Okay.
371 00:36:17.330 ⇒ 00:36:23.530 Justin Breshears: Let’s see if we can run this, like, without all of your fingerprints in it, to see…
372 00:36:23.700 ⇒ 00:36:32.660 Justin Breshears: almost, like, how that… that works with, like, you’re gonna have me as a dedicated PM on it, like, setting up a solid plan, like, oh, which tech leading this thing, or…
373 00:36:32.880 ⇒ 00:36:38.369 Justin Breshears: And then, yeah, to this thing up, let’s see how we can do it. So the goal is to get you out of these pies.
374 00:36:41.110 ⇒ 00:36:50.949 Uttam Kumaran: So, I mean, I think, like, Awash, if you… you could… I mean, Awash would be… Awash could solution architect, and I would just get you and another engineer, Awash.
375 00:36:51.940 ⇒ 00:36:53.040 Awaish Kumar: Okay, yep.
376 00:36:53.460 ⇒ 00:36:58.050 Uttam Kumaran: like, we can either bring on some of the people we’ve interviewed, or I’ll go find someone that’s done a lot of Snowflake.
377 00:36:59.190 ⇒ 00:37:00.170 Uttam Kumaran: work.
378 00:37:00.380 ⇒ 00:37:02.289 Awaish Kumar: I would prefer someone who is…
379 00:37:02.800 ⇒ 00:37:06.559 Awaish Kumar: who has experience in Snowflake, maybe?
380 00:37:06.560 ⇒ 00:37:08.550 Uttam Kumaran: promise that there’s nobody that’s done this, like…
381 00:37:09.660 ⇒ 00:37:18.170 Uttam Kumaran: like, I’ve only done this because I was messing around, like, a few years ago. There’s not many people that have done this. So instead, I just gotta get you a great, like, full-stack person.
382 00:37:18.640 ⇒ 00:37:20.900 Awaish Kumar: Okay. Or at least a good, like, Python…
383 00:37:21.130 ⇒ 00:37:22.839 Uttam Kumaran: Back… back in person.
384 00:37:24.170 ⇒ 00:37:28.560 Uttam Kumaran: Like, getting someone to know the snowflake stuff is… you’re never gonna find that person.
385 00:37:29.020 ⇒ 00:37:31.850 Uttam Kumaran: Or they’re… or they’re just lying.
386 00:37:32.020 ⇒ 00:37:35.550 Uttam Kumaran: So let me,
387 00:37:36.830 ⇒ 00:37:46.080 Uttam Kumaran: Okay, so I feel good about this. I think they… otherwise, they asked about, assumptions and risk.
388 00:37:47.060 ⇒ 00:37:52.170 Uttam Kumaran: Have you done… I wish you do any work with Snowpark, Streamlit, or Native App before?
389 00:37:54.950 ⇒ 00:37:58.490 Awaish Kumar: Yeah, not with app, but…
390 00:37:59.140 ⇒ 00:38:03.790 Awaish Kumar: Only the snow park would, like, just for… Development.
391 00:38:05.100 ⇒ 00:38:07.290 Awaish Kumar: Look, like, my own.
392 00:38:07.760 ⇒ 00:38:08.520 Awaish Kumar: Goss.
393 00:38:09.710 ⇒ 00:38:11.959 Uttam Kumaran: Okay, I’ve done some work with Streamlit.
394 00:38:12.250 ⇒ 00:38:16.220 Uttam Kumaran: I built, like, a sample native app, but, like.
395 00:38:17.240 ⇒ 00:38:21.229 Awaish Kumar: I’ve worked with Streamlit, but not… I’ve had Streamlore, yeah.
396 00:38:21.650 ⇒ 00:38:22.350 Awaish Kumar: Losing Python.
397 00:38:22.350 ⇒ 00:38:24.480 Uttam Kumaran: But it’s just like, yeah, it’s easy.
398 00:38:26.600 ⇒ 00:38:32.849 Uttam Kumaran: But see, look, they literally, like, deliver, approvable, installable, that builds audiences in 30 days.
399 00:38:33.220 ⇒ 00:38:38.359 Uttam Kumaran: I mean, so at this point, like, I think it’s gonna be at least 40 hours from an engineer, it’s gonna be at least…
400 00:38:38.970 ⇒ 00:38:42.000 Uttam Kumaran: 20 hours from a solutions architect.
401 00:38:42.860 ⇒ 00:38:49.350 Uttam Kumaran: And at least… 20 hours of a… PM? Is that about right?
402 00:38:50.570 ⇒ 00:38:51.420 Awaish Kumar: Okay.
403 00:38:52.800 ⇒ 00:39:00.029 Awaish Kumar: Like… If we want to include it for… for increasing our rate, Then it’s okay, but…
404 00:39:00.430 ⇒ 00:39:02.150 Awaish Kumar: Do we really need a BM?
405 00:39:03.900 ⇒ 00:39:06.619 Uttam Kumaran: the… so, I guess my point there, and this is something that I’ve…
406 00:39:06.790 ⇒ 00:39:25.890 Uttam Kumaran: I’ve told everybody is, like, every client has to go through PM team. Whether the PM team decides to do light agile, or heavy agile, whatever, is up to that person, but we have, like, we made this mistake before where, like, it can’t be direct. So, no matter what, there will be some time from PM.
407 00:39:27.070 ⇒ 00:39:34.700 Uttam Kumaran: like, I would rather price… see, this is also the thing is now, if I’m talking sales, I’m gonna price it high. They’ll either tell me…
408 00:39:35.480 ⇒ 00:39:39.359 Uttam Kumaran: like, that’s way too much, like, there’s no way it’s gonna happen. Or they’ll say, like.
409 00:39:39.750 ⇒ 00:39:54.699 Uttam Kumaran: why do we need a PM? If they say, why do we need a PM? I’m actually happy, because then we could scale that back, we can think about, like, what that is, I can try to get the deal over. Like, but we’re not going to have any project we do that’s not going to go through project management, like, in some form, or a delivery team in some form.
410 00:39:54.810 ⇒ 00:39:58.720 Uttam Kumaran: Whether we bill for that or not, like, it’s up to us to decide.
411 00:39:58.840 ⇒ 00:40:02.049 Uttam Kumaran: And that’s, like, that could be… that’s a sales decision.
412 00:40:02.260 ⇒ 00:40:06.649 Uttam Kumaran: But, like, it… it has to… someone has to be there. It’s redundance, like, there has to be…
413 00:40:07.110 ⇒ 00:40:11.580 Uttam Kumaran: There has to be these, like, three stools, 3-leg stool, you know, on all these projects.
414 00:40:12.300 ⇒ 00:40:17.879 Justin Breshears: Yeah. We already have, like, trigger warnings in my head for tight timeline, like.
415 00:40:17.880 ⇒ 00:40:18.619 Uttam Kumaran: Yes, I know.
416 00:40:18.620 ⇒ 00:40:22.210 Justin Breshears: possible scope creep on, like, POC versus production.
417 00:40:22.210 ⇒ 00:40:30.630 Uttam Kumaran: No, so I’m gonna… I’m gonna pitch a huge scope, right? So even if this becomes psychotic, at least we win on the money side. Right.
418 00:40:30.800 ⇒ 00:40:33.029 Uttam Kumaran: If they see our price, and they’re like.
419 00:40:33.220 ⇒ 00:40:38.410 Uttam Kumaran: Get out of here? Okay, I don’t care, whatever. There’s no one that’s gonna do this type of work. It isn’t my type of work.
420 00:40:38.410 ⇒ 00:40:42.639 Justin Breshears: which… I’ll just say, like, your prices are not get-out-of-here prices, so…
421 00:40:42.640 ⇒ 00:40:50.490 Uttam Kumaran: Oh, and I know, I know that. So, like, even at a higher level, we’re not even that expensive. So, I mean, I’m gonna pitch… what I’m gonna pitch them is, like.
422 00:40:50.590 ⇒ 00:40:56.439 Uttam Kumaran: That 2020-40… I’m gonna put hours on each of those, and then…
423 00:40:56.820 ⇒ 00:41:06.129 Uttam Kumaran: we’ll kind of go from there. If they say yes, then we went too low. But then it’ll be a high enough number where we won’t pick ourselves that much, so…
424 00:41:07.220 ⇒ 00:41:08.400 Justin Breshears: I think it’s a good plan.
425 00:41:08.760 ⇒ 00:41:13.810 Uttam Kumaran: Okay, cool. So, I… I can wrap this up. Awash, do you… if you just want to do this piece.
426 00:41:14.330 ⇒ 00:41:22.950 Uttam Kumaran: And then… I guess, Justin, if you have any parts of this that you’re like, We should edit.
427 00:41:23.220 ⇒ 00:41:25.639 Uttam Kumaran: We can do that, but .
428 00:41:26.480 ⇒ 00:41:28.829 Justin Breshears: I’ll go back… I’ll go back through it and take a look.
429 00:41:28.830 ⇒ 00:41:34.879 Uttam Kumaran: Let me just do, like, a quick once-over. I’m going to make my email back to them, like, very, very…
430 00:41:35.180 ⇒ 00:41:46.419 Uttam Kumaran: clear, and then this, we can have all the detail, but I’m gonna make it very clear with, like, what we’re signing up for in 30 days, what we’re not signing up for. I’ll do, like, probably, like, a 30-day plan and a 90-day plan.
431 00:41:46.580 ⇒ 00:41:48.670 Uttam Kumaran: And to make it really clear for the CEO,
432 00:41:48.920 ⇒ 00:41:52.710 Uttam Kumaran: I can tell that both of those guys are different, like, that technical guy…
433 00:41:52.940 ⇒ 00:41:56.149 Awaish Kumar: It’s probably more in the weeds. CEOs probably just, like, want to hear the one-liner.
434 00:41:56.180 ⇒ 00:41:59.890 Uttam Kumaran: So we’ll answer both of those, and I’ll… Shift us today.
435 00:42:03.440 ⇒ 00:42:03.960 Awaish Kumar: Agreed.
436 00:42:04.860 ⇒ 00:42:06.510 Justin Breshears: 10 minutes late to the next meeting, so…
437 00:42:06.670 ⇒ 00:42:10.650 Uttam Kumaran: Yes. Okay, well, let’s go somewhere there. Thanks, S.