Meeting Title: Uttam <> Robert <> Nick - Stella-Analytics-Opportunity Date: 2024-01-22 Meeting participants: Nick Baker, Robert Tseng, Uttam Kumaran


WEBVTT

1 00:00:30.290 00:00:35.360 Uttam Kumaran: Hey? Hey? How’s it going, hey? Good. Thanks. Sorry about that.

2 00:00:35.630 00:00:36.740 Robert Tseng: No worries.

3 00:00:38.050 00:00:44.210 Uttam Kumaran: Okay. Me and Nick were just talking, and then we hopped off. But let’s see, he’ll be back.

4 00:00:44.470 00:00:45.330 Robert Tseng: Okay.

5 00:00:48.860 00:00:49.610 Uttam Kumaran: cool.

6 00:00:51.170 00:01:03.169 Uttam Kumaran: How’s the day going? Monday? Monday is always kind of crazy for me. So yeah, it’s been actually pretty good. I’ve been I’ve been trying to like really

7 00:01:04.239 00:01:06.600 Uttam Kumaran: group my meetings on like a few days.

8 00:01:06.620 00:01:19.810 Uttam Kumaran: And so I’ve I’ve kept like today, pretty free, but then it lets me like take on stuff like this and so trying to do better, II just feel like last month I was like every day. I like meetings in actual working. Now

9 00:01:19.890 00:01:28.220 Uttam Kumaran: I’m trying to have, like some, some like focus time for work and then putting meetings and stuff on other days. But I think it’s Eb and flow.

10 00:01:28.930 00:01:40.529 Uttam Kumaran: Yeah, yeah, I mean for me, it’s Monday is kinda when I have to like answer to anything that happened over the weekend or yeah. And then on Tuesday is when I get to really focus a bit more. So

11 00:01:40.760 00:01:51.100 Uttam Kumaran: yeah, I try to have like end of week stuff on Friday, and then I just try to plan everything as I can. And then I come on Monday a little. But then I, yeah, II really don’t not like

12 00:01:51.660 00:01:55.390 Uttam Kumaran: not too dependent on the weekend. So yeah, I feel that

13 00:01:56.230 00:01:57.260 Robert Tseng: for sure.

14 00:01:59.250 00:02:03.370 Uttam Kumaran: okay, message, Nick.

15 00:02:20.430 00:02:21.110 okay.

16 00:02:24.210 00:02:25.030 Uttam Kumaran: And I,

17 00:02:27.320 00:02:42.299 Nick Baker: what’s up? Oh, I look so. You are sick days ago. So I’m I’m II mean, I’m normally this pale, but I’m like extra nasty version of myself right now, so

18 00:02:42.580 00:02:44.780 Nick Baker: I normally have a little more life in my eyes.

19 00:02:44.830 00:02:52.750 Uttam Kumaran: You look great man, I appreciate it. I don’t feel great, but I’m still here.

20 00:02:53.490 00:02:54.300 Robert Tseng: Thanks

21 00:02:55.190 00:03:00.690 Uttam Kumaran: cool. So I’m glad we could all hop on. Maybe we could just hop right in, I think, the big thing.

22 00:03:00.720 00:03:13.479 Uttam Kumaran: Rob, we just want to get a sense more about the company and more about the quoting flow. Yeah, I think pretty much. We’re we’re good on, like, generally like, we need this intermari intermediary layer between

23 00:03:13.620 00:03:20.470 Uttam Kumaran: segment and amplitude. The big thing is like, can you give us a little bit more sense of like? What

24 00:03:20.860 00:03:26.809 Uttam Kumaran: what business logic on the analytic side. They want to try to replicate. And I guess my question was.

25 00:03:27.030 00:03:34.649 Uttam Kumaran: are we able to get that directly from the events? Is that something we have to replicate and like, just a couple of like examples, maybe we can talk through.

26 00:03:35.470 00:03:48.610 Robert Tseng: Yeah, I mean, I think right now, everything can just go from the events I mean, that’s like, why, I’m setting up all the event tracking for them. So yeah, what I’ve gone in and done is like they have their application running. But then, like.

27 00:03:49.220 00:03:52.110 Robert Tseng: obviously, there’s a lot of back end

28 00:03:52.250 00:03:55.640 Robert Tseng: stuff that doesn’t need to be tracked. And so the

29 00:03:55.780 00:04:18.150 Robert Tseng: what I’m doing in segment is just like defining like click and click events and pulling like data data off of off of screen site like the this flow that we’re focused on is like a like a quote building Sas tool. For like metal and fabrication, specifically. So, there’s like a lot of data that they want to scrape off of pages as they’re going through as users are going through this flow. And

30 00:04:18.734 00:04:26.629 Robert Tseng: I mean, that’s that’s why I’ve been setting up. So then the the analytics downstream will just be doing some of the math to like, I guess.

31 00:04:26.740 00:04:27.909 Robert Tseng: Yeah, it’s just

32 00:04:28.360 00:04:40.890 Robert Tseng: to to calculate. Do some of the calculations in the quotes. And then, yeah, I guess just different basic reports on on usage and certain milestones that they want to be looking at.

33 00:04:41.540 00:04:58.000 Uttam Kumaran: Yeah. So I don’t think it doesn’t seem like there should be a huge concern that we’re replicating a ton of the stuff that engineering is doing. In fact, I think most of the calculations will be counts and sums on top of like events and like segmentation, things that are purely needed for reporting. If there are like.

34 00:04:58.220 00:05:02.100 Uttam Kumaran: If there are like costs or other like calculations, we need to do.

35 00:05:02.110 00:05:11.009 Uttam Kumaran: I mean again, as long as they don’t, those don’t change. Those can be pretty easily defined in sequel. So so I think that kind of tackles some of the probably the concern.

36 00:05:11.120 00:05:13.970 Uttam Kumaran: and we could probably include that a little bit about just like

37 00:05:14.770 00:05:18.399 Uttam Kumaran: like what kind of calculations we will be doing in sequel?

38 00:05:18.820 00:05:33.359 Robert Tseng: sure. Yeah, I mean, I guess my my take on that is you know, if they’re defining, because even what they’re defining as like a active user is like changing, I think. I mean there.

39 00:05:33.660 00:05:52.399 Robert Tseng: And and so the way that yeah, you know when it, when it, when that has to change on the app, it it just doesn’t change on the application side, being able to own that on the and then, like an analytics stack will allow you to like, change these definitions faster without actually impacting the the application. So I think that’s like one of the

40 00:05:52.680 00:06:00.780 Robert Tseng: like, the the cells that I’ve been using for like, why, why we should be having this engineering replica or at analytics replica as well.

41 00:06:01.230 00:06:19.339 Uttam Kumaran: Yeah, I think that’s pretty common. Like. Again, you have, you have an application, and you have like things that are necessary for the application to work. And then you have reporting, which is just like as your goals change, and as we’re able to get more insight, or we’re able to consider more information. You can build a definition for active user, whatever. So I would say.

42 00:06:19.390 00:06:21.500 Uttam Kumaran: that sort of split is pretty common.

43 00:06:21.630 00:06:28.549 Uttam Kumaran: And and where and where, in in opportunities where, like, I think, we can benefit from just using whatever definition engineering has

44 00:06:28.720 00:06:34.359 Uttam Kumaran: just, we would use that. But again, there’s gonna be times where either they change the definition for reporting purposes.

45 00:06:34.430 00:06:40.810 Uttam Kumaran: or it’s just not a definition. So this is a pretty common thing that I think would be easy to describe.

46 00:06:41.090 00:06:56.680 Robert Tseng: Yeah. And they just gave me. Then they just gave me access to the application database. Today. I haven’t actually looked through the schema yet. But I was thinking that once I get a handle on that that could be a good starting place as well as your defining like tables and stuff. Yeah.

47 00:06:59.570 00:07:25.279 Robert Tseng: okay, stakeholders is Thursday. We have, like everybody on the team there. I’m kind of like we pushing this next phase of the data strategy. And that’s kind of when I want to really like, get them to like, be ready to to decide. I’m gonna whatever document we put out. I love to share them for the meeting, and, you know, try to get try to get by and like in that in that meeting. So I think they have. Thursday is what I’m aiming for them.

48 00:07:25.700 00:07:27.529 Uttam Kumaran: Nick, like anything else

49 00:07:27.920 00:07:32.279 Uttam Kumaran: you think we need to include. I think me and you could probably just work on outing like

50 00:07:32.350 00:07:36.949 Uttam Kumaran: Snowflake, dbt. and, like Github, are like, I don’t know

51 00:07:37.040 00:07:39.619 Nick Baker: what else we what do they have?

52 00:07:39.730 00:08:07.010 Nick Baker: Who do they like? Who do they have internally that you know. So like, let’s say, we set something up. We get like the Mvp. And then we do some flushing out, maybe building some testing, like, you know, do some sexy, cool stuff? Who do they have internally? That would then adopted it long term like, like, are there particular stakeholders? We need to cater this towards that are literate in one tool versus another. That this would this like might help us or help us get that influence that we need.

53 00:08:07.120 00:08:20.220 Robert Tseng: Yeah. So there are no technical people that have used this tech stack. I mean, this is like A, you know. I think it’s what I’ve noticed in industries like this is, they’re not. They have. You know, there’s a lag in terms of adopt adopting modern data tools. So

54 00:08:20.260 00:08:36.059 Robert Tseng: yeah, I mean, there’s they have a data analyst. They have a data analyst and a data scientist. I guess that. I’ve been trying to also train with the things that I’ve been bringing into the company. But yeah, I think that’s that’s all they have right now.

55 00:08:36.299 00:08:56.049 Nick Baker: Okay, yeah. And those those those people don’t feel like they have a ton of influence. It would, I would. Yeah, okay, alright. So there’s no really no reason to cater that except for to say that like, maybe inclusive of some sort of additional like documentation guidance training that we can offer as like a piece of the hand off. Ultimately.

56 00:08:56.140 00:09:13.269 Robert Tseng: yeah, I’m also okay with not handing it off like they’re comfortable paying me a retainer fee. Just so like, yeah, fair point. Yeah. In that case, then no, no reason to do that. Yeah.

57 00:09:13.430 00:09:18.409 Nick Baker: Okay. Yeah. I mean, I think we could look through it a little bit more from what I saw. I mean it feels

58 00:09:19.430 00:09:32.560 Nick Baker: it feels like for the most part, is pretty straightforward. I guess, without having any more context of what their their current setup really looks like beyond what what we said here, which it seems like. It’s again also pretty straightforward. It seems like we kind of got what we need in there.

59 00:09:32.860 00:09:38.920 Robert Tseng: Yeah, I mean, it’s it’s a, it’s a react front end. And

60 00:09:38.990 00:09:42.180 Robert Tseng: yeah, they just have. I mean, they have like a a few.

61 00:09:42.270 00:09:59.159 Robert Tseng: I guess I haven’t looked into too much detail on their application database. But it’s just a few tables that you know capture like basic customer information. And then, like they just yeah. And and they, I don’t know. And then it gets a bit complex with their logs there. So

62 00:09:59.300 00:10:01.709 Robert Tseng: but yeah, I mean, I have like a

63 00:10:02.100 00:10:16.470 Robert Tseng: I’ve been. I’ve been building and maintaining this like stream of like event data defined for them. And that’s kind of their new like application data that’s gonna be into whatever you build, I guess

64 00:10:16.560 00:10:24.599 Nick Baker: does the and then it’s sort of it’s using. It’s a quoting flow effectively. Is this like their their own app for the quoting flow? Are they using it?

65 00:10:24.670 00:10:44.150 Nick Baker: This is their own native app for a specific kind of quoting in this industry, and that’s like one of their tools. They have a few other tools that I haven’t touched yet, but this is the first one we’ve been focusing on. Do we have any sense of like the reliability of the event? Data?

66 00:10:44.660 00:10:51.450 Robert Tseng: Well, since I’ve been setting it up now, I think it’s more reliable. Yeah. But

67 00:10:52.310 00:11:01.790 Robert Tseng: yeah, I mean, I think I think their their application is pre, relative is like, it’s pretty pretty solid in terms of like. I haven’t seen too many like data gaps after I’ve set this up so.

68 00:11:02.130 00:11:08.959 Uttam Kumaran: And Robert the main, the main outputs gonna be just piping through to your amplitude dashboards.

69 00:11:09.130 00:11:28.390 Uttam Kumaran: So dashboards. So that’s kind of what I’m just gonna call that. So that’s the big thing is like we’ll try to. I’ll just try to litter in here. A little bit more information about like these are gonna go directly to power, not only like the speed and like the reliability of those inputs dashboards. But your ability to go build a ton of those

70 00:11:28.600 00:11:29.960 Uttam Kumaran: like way quicker

71 00:11:30.080 00:11:42.320 Uttam Kumaran: and without having to go straight from raw events every time. And again, that that’s that’s like, actually, I think, the benefit of it. And then again, as you include more complex business logic.

72 00:11:42.370 00:12:02.680 Uttam Kumaran: you’re not just writing that on on the fly sequel models like, those are all version control. And and Github, and yeah, so yeah, I think the version control data management piece is a big big sell for them, because, yeah, right now, in their quoting flow, whenever like something changes like in the middle, then goes back to change it. Their database. Their their their data doesn’t

73 00:12:02.790 00:12:17.810 Robert Tseng: the application database just like gets rid of the previous info. Right? There’s no his. There’s no history of so, yeah, just being able to do that would be like a huge, another level of functionality for them.

74 00:12:17.910 00:12:23.459 Uttam Kumaran: I think the only thing maybe you can think of is like, let’s say we were to work like what that initial

75 00:12:23.610 00:12:33.619 Uttam Kumaran: like output is that we can like team up on whether it’s like a dashboard that you have coming up to deliver, or some extra insight that you’re like.

76 00:12:33.990 00:12:38.689 Uttam Kumaran: In 2 weeks we can drive towards that as like an initial thing, or in 4 weeks we can drive towards that or whatever.

77 00:12:38.830 00:12:54.810 Uttam Kumaran: Yeah, I I’ve been trying to push them to like, pick one. Yeah. But I think again. That’ll just mean, like we can really set stuff up in a way just to drive towards that again. It’ll be a little bit of time. I’m just getting in for a setup. But

78 00:12:55.030 00:13:15.710 Robert Tseng: yeah, form details a bit more. I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m I’m

79 00:13:15.740 00:13:18.720 Robert Tseng: just like, a one pager. They like to read that kind of stuff.

80 00:13:18.740 00:13:31.110 Robert Tseng: and then, because their head of engineering is going to be there, I think some sort of like diagram like they don’t have an Erd or or dag, or anything like that right now. So, being able to just like show them a piece of that like

81 00:13:31.220 00:13:37.419 Robert Tseng: would be would be great. Just to kind of speak more directly to him and get his sign off. Yeah.

82 00:13:37.840 00:13:40.749 Uttam Kumaran: okay, yeah. I mean, I think we could just put literally like it.

83 00:13:41.260 00:13:51.520 Uttam Kumaran: a basic technical diagram of snowflake. Dbt, yeah, like, I have, I have a couple of those lying around. Yeah, nothing. Fancy. I think something just like that would be good.

84 00:13:51.650 00:13:52.960 Uttam Kumaran: Okay, yes.

85 00:13:55.660 00:14:05.299 Uttam Kumaran: Okay. So maybe, Nick, we’ll just chat slack. And I think it would. This is like, just like an hour or 2 away from like rounding this out. And then. yeah, in terms of Doc.

86 00:14:06.470 00:14:07.300 Nick Baker: deal

87 00:14:07.910 00:14:08.780 Robert Tseng: cool.

88 00:14:08.940 00:14:20.320 Uttam Kumaran: Okay, cool. Alright. So then talk with slack. And then, yeah, let me let us know what Thursday goes, or if anything else in between. yeah. And then do you wanna stay on it? Yeah, yeah, yeah.

89 00:14:20.470 00:14:30.770 Uttam Kumaran: yeah. If you wanna let me know. What do you think? Nick, we’re just talking about another opportunity that came up. I was actually thinking about looping in Brian on this one. But you mentioned.

90 00:14:30.800 00:14:37.460 Uttam Kumaran: Rob, that this is just the It’s like something around like Hippodada. How did this? How does lead come in?

91 00:14:37.980 00:14:45.739 Robert Tseng: Yeah. So I mean, II I like to create some content on Linkedin or whatever. And

92 00:14:45.840 00:14:48.309 Uttam Kumaran: I was gonna ask you, like how it’s going.

93 00:14:48.350 00:14:58.769 Robert Tseng: Yeah, you know, you get an inbound lead here and there. So yeah, I posted today that I liked this new snowflake feature. And I got like 300 views.

94 00:14:59.060 00:15:03.500 Uttam Kumaran: I don’t know. I feel like I’m you know. I’m learning from you a little bit. But I

95 00:15:03.610 00:15:07.999 Uttam Kumaran: it’s just still so hard for me, but

96 00:15:08.150 00:15:18.110 Robert Tseng: just posting twice a week is enough for me like II don’t want to spend too much time doing this stuff, but if it gets me like 10 leads a week. I think I’m I’m good with that. So

97 00:15:18.610 00:15:22.190 Robert Tseng: but yeah, so they’re I guess they’re.

98 00:15:22.280 00:15:41.189 Uttam Kumaran: I don’t know if you if you forwarded the details to Nick. But yeah, I guess, like, basically II could just like kind of summarize it. So it looks like again. I don’t know the company, but it just looks like they have a ton of tools like Wordpress, Hubspad salesforce. They have some sort of data management tool.

99 00:15:41.260 00:15:44.570 Uttam Kumaran: For like processing medical records.

100 00:15:44.610 00:15:48.310 Uttam Kumaran: And they want to bring that data into.

101 00:15:48.650 00:16:03.890 Uttam Kumaran: They want to bring their hub salesforce data into Snowflake. They have some Rds application data that they probably want to bring in as well. They need to develop some like customer like de-du dupes. They need to have some sort of break

102 00:16:04.460 00:16:12.980 Uttam Kumaran: role based access control for that data. And they probably want Dbt. Github. All that stuff, literally, the responsibilities are like.

103 00:16:13.350 00:16:17.990 Uttam Kumaran: pretty much exactly like. well, we there’s like nothing on there except

104 00:16:18.140 00:16:24.429 Uttam Kumaran: the refining of the Rds. Yeah, it’s like which I again, I could do.

105 00:16:24.530 00:16:32.539 Uttam Kumaran: But I think like I’m surprised that they want somebody to to come in and do that, because that’s more like a back end thing. But everything else is like pretty upper.

106 00:16:33.080 00:16:42.260 Uttam Kumaran: It’s like pretty Upper Alley again. It’s more interested in like, would they be good to use 5 train? And if they already have snowflake like? Well, how are they? What are they already using it for? But

107 00:16:42.880 00:16:50.209 Uttam Kumaran: I mean, I don’t know also, do do you know whether they want like dashboarding and stuff? Do they already have analysts like, that’s another question out to

108 00:16:50.390 00:17:18.319 Robert Tseng: yeah. So II mean, I guess when I see something like this, I mean, this is all they give me, and it’s like it. It. I’ve I’ve done a few of these where they’re like, hey, these are what we’re looking for, submitted Rfp, so you try to you. You create like an outline. You try to hit on as many things like, if you don’t, there’s we wanna ignore the Amazon Rds far we can like we. And then we can, we we kinda it’s a kind of up to us to like. Imagine it out of it for them. Would that possibly look like? So

109 00:17:18.940 00:17:35.290 Robert Tseng: yeah, I mean, II don’t have hipaa experience. I don’t really work with healthier companies, but the tech stack look very like up your alley. So that’s why I pass it on along to you to see if you’d wanna go after it like the amplitude in salesforce hubspot piece like I, you know, I’m I’m good with with that. But

110 00:17:36.020 00:17:43.049 Uttam Kumaran: yeah, I mean, I think I think we should cause Nick. I think Brian has, like hipaa experience from

111 00:17:43.150 00:17:48.030 Uttam Kumaran: an earlier client, or where he was, he was full time somewhere.

112 00:17:48.070 00:17:57.409 Uttam Kumaran: But I mean again, I think that’s probably just like I think that’s honestly just like would push us forward. I don’t. I think there’s nothing really specific about that, apart from

113 00:17:57.580 00:18:20.700 Nick Baker: just needing to know which I feel like they’re always like extra Co, obviously, it’s like hippets client, you know, confidential data. So they’re gonna be way more cautious about it and working with somebody that already has experience actually different in practice. What you’re doing, which it probably isn’t. Just like more stringent snowflake provisioning. But yeah, II feel like Brian would be able to kill that with

114 00:18:20.720 00:18:23.260 Nick Baker: with, probably it’s like 30 Madison.

115 00:18:23.280 00:18:30.640 Uttam Kumaran: Yeah, yeah, it’s definitely like a shitload of work, though. Yeah, so like, like.

116 00:18:30.950 00:18:33.510 Uttam Kumaran: it’s like a lot of tools. And that’s why

117 00:18:33.740 00:18:40.420 Uttam Kumaran: I think we could easily put together if this was like nothing set up. Here’s all the things we would do. It’s easily like

118 00:18:40.770 00:18:46.819 Uttam Kumaran: it’s easily like 6 months of work to manage and build data models and bring all that for

119 00:18:47.030 00:18:50.650 Uttam Kumaran: for all those tools? So the questions would be.

120 00:18:50.820 00:18:52.519 Uttam Kumaran: What’s a timeline?

121 00:18:52.630 00:18:57.259 Uttam Kumaran: Can we bring more people on? And then also is there. Do you have an existing team?

122 00:18:57.320 00:19:08.680 Uttam Kumaran: And then II think we could easily put together a one pager with like this is what we would do. Here’s a rough timeline and like key milestones. Here’s the experience our team has on working with all these tools.

123 00:19:08.760 00:19:14.070 Uttam Kumaran: and then, of course, the medical experience. And then I mean pretty much line by line

124 00:19:14.460 00:19:22.260 Uttam Kumaran: like this is, that’s exactly what we do. So I mean, I think we should just throw something at it like again. Maybe I just I could just put together a doc.

125 00:19:22.280 00:19:24.560 Uttam Kumaran: I’ll loop in Brian, and we can see

126 00:19:25.190 00:19:29.839 Uttam Kumaran: and I’ll just send that to you if you or if you want to send that, or wh what up? What do you think?

127 00:19:30.600 00:19:37.769 Robert Tseng: Yeah, I mean, I’m I’m cool with that. I mean, I’ll ask those questions. But if you want to put in like, yeah, that that snippet of

128 00:19:38.110 00:19:48.389 Robert Tseng: the experience to have the document and brag with him, and I think that’d be good to just send that over with him. I mean, again, I like going like above me on on these, because

129 00:19:48.700 00:19:56.739 Uttam Kumaran: again, a lot of those guys, I think they get a lot of contractors applying or individuals. And so if you’re set up, if you’re able to be like yo, here’s an actual like.

130 00:19:57.160 00:20:05.790 Uttam Kumaran: pretty much a project plan. and we’ve done every single item here. let’s get on a call. You know that we could try that. So. But yeah.

131 00:20:07.840 00:20:13.469 Uttam Kumaran: cool. And then you have. And then you have that, doc. So in case it comes up again. Okay, so let’s do that, and then I will.

132 00:20:13.690 00:20:16.409 Uttam Kumaran: I’ll try to get that, too, sometime tomorrow.

133 00:20:16.810 00:20:18.549 Robert Tseng: Okay, that sounds good.

134 00:20:20.120 00:20:24.149 Uttam Kumaran: Okay, cool. And do that. Now, I gotta start posting on Linkedin. If you’re in 10 leads a week.

135 00:20:25.300 00:20:26.480 Uttam Kumaran: Yeah.

136 00:20:27.220 00:20:35.449 Robert Tseng: I mean, you know, maybe one out actually worth something. But yeah, that’s awesome.

137 00:20:35.550 00:20:51.109 Uttam Kumaran: Yeah. I mean you. You start posting. I sorry, like I lost your invite to the ether. But now I’m every time you post I’m gonna I’m going to be interact engaging with it. That’s how it works. Yeah.

138 00:20:51.130 00:20:52.480 Robert Tseng: alright, thanks, guys.

139 00:20:52.580 00:20:53.990 Nick Baker: back to soon.