Meeting Title: Brainforge <> CTA: Follow Up on Data Initiatives Date: 2025-09-22 Meeting participants: Kyle Wandel, Katherine Bayless, Uttam Kumaran, kaybay
WEBVTT
1 00:00:31.360 ⇒ 00:00:35.359 Kyle Wandel: Three… three times when I look at the office, or right from the moment.
2 00:00:36.010 ⇒ 00:00:38.400 Kyle Wandel: Hello again, Catherine.
3 00:00:38.550 ⇒ 00:00:40.050 Katherine Bayless: Blue. Blue, Lou.
4 00:00:40.950 ⇒ 00:00:43.330 Katherine Bayless: Whew, man, how’s your Monday going?
5 00:00:43.790 ⇒ 00:00:49.370 Kyle Wandel: It’s good so far. We’re finally getting the tariff report out today, so, it’s a big…
6 00:00:49.640 ⇒ 00:00:54.199 Kyle Wandel: It’s off of my shoulders, so… Sorry, I didn’t get that.
7 00:00:54.200 ⇒ 00:00:55.110 Katherine Bayless: Yeah, yeah.
8 00:00:55.110 ⇒ 00:00:56.399 Kyle Wandel: forehand, leg.
9 00:00:57.530 ⇒ 00:00:57.970 Kyle Wandel: Excellent.
10 00:00:57.970 ⇒ 00:00:59.150 Katherine Bayless: All things.
11 00:00:59.150 ⇒ 00:01:18.930 Kyle Wandel: Yeah, I get that. I had a, so this also was my peak week for… one of my peak weeks for training. So, on Saturday, I did a 50-mile run… not a 50-mile run, wow, a 50-mile bike, and then, a Sunday I did a half marathon and a 2.2K swim the same day, so…
12 00:01:19.550 ⇒ 00:01:20.910 Katherine Bayless: That’s not nothing.
13 00:01:20.910 ⇒ 00:01:27.300 Kyle Wandel: Yeah, I have to do it… I’ll have to do it one more time this weekend, but then other than that, I start coasting and getting ready, basically.
14 00:01:27.470 ⇒ 00:01:29.279 Katherine Bayless: Okay. Yeah, my friend that’s training.
15 00:01:29.280 ⇒ 00:01:29.790 Kyle Wandel: Bye.
16 00:01:29.790 ⇒ 00:01:39.250 Katherine Bayless: She did 20 miles on Friday, and so I think she’s in a similar, like, getting towards the, yeah, just rinse and repeat until the day comes.
17 00:01:39.250 ⇒ 00:01:51.049 Katherine Bayless: She did say she’s like, I guess she’s bought running shoes usually that are a half size bigger. She’s like, I’m gonna have to get them a full size bigger. She’s like, my feet just turn into balloons by the end of, like, the 16th mile.
18 00:01:51.270 ⇒ 00:02:00.729 Kyle Wandel: Yep, that’s the… the biggest and one of the worst things about it is that you literally just need… people, like, wear Crocs afterwards, just because of how much that it gets big.
19 00:02:00.730 ⇒ 00:02:02.909 Katherine Bayless: Yeah, something I did not know.
20 00:02:05.200 ⇒ 00:02:06.960 Uttam Kumaran: Good morning. Nice to meet you, Kyle.
21 00:02:07.350 ⇒ 00:02:08.689 Kyle Wandel: Nice to meet you as well.
22 00:02:09.160 ⇒ 00:02:24.230 Katherine Bayless: Yeah, I tagged, Kyle in at the last minute, my fault for not forwarding the invites sooner. Kyle is, the new lead data engineer on the team, having started, technically last week, although, some lingering work has kind of… oh, my screen went…
23 00:02:27.200 ⇒ 00:02:28.890 Kyle Wandel: Yep, I think she just cut out.
24 00:02:28.890 ⇒ 00:02:32.050 Uttam Kumaran: Am I still… am I still here? Yeah, I’m still here. Yeah, yeah, I see you here.
25 00:02:32.050 ⇒ 00:02:32.959 Kyle Wandel: Okay, cool.
26 00:02:33.760 ⇒ 00:02:40.140 Kyle Wandel: I’ll message you quickly. Our… sometimes our, HQ… Internet is lovely.
27 00:02:40.140 ⇒ 00:02:46.329 Uttam Kumaran: Nice. Well, it’s great to meet you. I know I heard a little bit from Catherine about the overall setup, but
28 00:02:46.500 ⇒ 00:02:51.540 Uttam Kumaran: Yeah, I kind of… I don’t know if she ended up sharing,
29 00:02:51.960 ⇒ 00:02:56.950 Uttam Kumaran: this document with you, but we put together a little bit of a… of a document on,
30 00:02:57.790 ⇒ 00:03:08.850 Uttam Kumaran: on sort of, like, what… what we were, starting to think through work streams and… and sort of areas where you guys need help, and how we can be supportive. So, I’ll just… I’ll share.
31 00:03:08.850 ⇒ 00:03:09.340 Kyle Wandel: epic.
32 00:03:09.340 ⇒ 00:03:10.139 Uttam Kumaran: Nope, and then.
33 00:03:10.140 ⇒ 00:03:11.970 Kyle Wandel: Yeah, that’d be… that’d be great.
34 00:03:12.180 ⇒ 00:03:12.820 Uttam Kumaran: Yeah.
35 00:03:14.450 ⇒ 00:03:21.050 Uttam Kumaran: Do you mind sending me your email address, or you can just put it into this chat here? Yeah, one second.
36 00:03:23.020 ⇒ 00:03:24.990 Kyle Wandel: I love your dog in the background, by the way.
37 00:03:24.990 ⇒ 00:03:36.470 Uttam Kumaran: I appreciate it. He is, I think he helps me, sort of, like, reduce awkwardness on meetings and cut through the people… he’s usually the star attention, but…
38 00:03:36.810 ⇒ 00:03:37.270 Kyle Wandel: Good to see you.
39 00:03:37.270 ⇒ 00:03:40.380 Uttam Kumaran: done a whole lot of nothing today, and I don’t know why he’s still sleeping, like…
40 00:03:40.380 ⇒ 00:03:43.760 Kyle Wandel: Is he a… is he a lab mix, or…
41 00:03:43.760 ⇒ 00:03:47.079 Uttam Kumaran: He is a Yellow Loud Bray Pyrenees, German Shepherd mix.
42 00:03:47.080 ⇒ 00:03:56.180 Kyle Wandel: Nice, nice, yes, we have a… we have two dogs… I have two dogs at home, and one of them is a rescue… he’s a lab Chesapeake mix with a couple other things.
43 00:03:56.180 ⇒ 00:03:58.940 Uttam Kumaran: Oh, what is that? I’ve never heard about that.
44 00:03:58.940 ⇒ 00:04:06.690 Kyle Wandel: Yeah, a Chesapeake lab is a, like a furry lab, basically, is what I like to call it. Okay. They, they, they look…
45 00:04:07.070 ⇒ 00:04:09.950 Kyle Wandel: almost like a goat, if that makes sense? Like, their code is very.
46 00:04:09.950 ⇒ 00:04:10.930 Uttam Kumaran: Very implied.
47 00:04:10.930 ⇒ 00:04:24.079 Kyle Wandel: Curly? Yeah. Yeah, but they’re, like, they’re just like any other lab, basically. He’s a… he’s a mix of everything, so he’s… he’s a little smaller, he’s like 50 pounds, so… Okay. But then we also have a Shepherd. We have a driver shepherd who…
48 00:04:24.080 ⇒ 00:04:24.820 Uttam Kumaran: Oh, great.
49 00:04:24.820 ⇒ 00:04:26.740 Kyle Wandel: He’s, like, 90 pounds, so…
50 00:04:27.020 ⇒ 00:04:27.780 Uttam Kumaran: Right.
51 00:04:27.980 ⇒ 00:04:32.030 Kyle Wandel: Yeah, he’s 120 pounds. He looks… Oh, nice.
52 00:04:32.030 ⇒ 00:04:44.279 Uttam Kumaran: It may look a little small, but it’s a pretty big couch, and yeah, and yeah, he’s a big boy, and… but he’s very, very calm. I think the German Shepherd, he’s, like, very smart.
53 00:04:44.450 ⇒ 00:04:44.780 Kyle Wandel: Yeah.
54 00:04:44.780 ⇒ 00:04:55.699 Uttam Kumaran: Pyrenees, he’s just, like, likes to roam and protect, but then he’s also still very, very happy, because he’s a lab, and he’s just gotten very active, so I like the mixes, so…
55 00:04:55.850 ⇒ 00:05:03.560 Kyle Wandel: Yeah, no, that’s the best way to do it. Our shepherd is neurotic, as I’ll get out, so he has all the good and the bad, so…
56 00:05:03.560 ⇒ 00:05:04.150 Uttam Kumaran: I agree.
57 00:05:04.150 ⇒ 00:05:05.530 Kyle Wandel: I definitely get that.
58 00:05:05.530 ⇒ 00:05:06.280 Uttam Kumaran: Yeah.
59 00:05:06.740 ⇒ 00:05:09.829 Uttam Kumaran: Hey, Catherine, I assume the Wi-Fi just cut out.
60 00:05:10.150 ⇒ 00:05:12.169 kaybay: No, I don’t know.
61 00:05:12.170 ⇒ 00:05:13.620 Uttam Kumaran: Sorry.
62 00:05:13.620 ⇒ 00:05:15.429 kaybay: Lenovo ThinkPads.
63 00:05:15.770 ⇒ 00:05:25.309 kaybay: me question life. It just, for some reason, it just, like, goes, like, totally, like, black screen. Like, the light for, like, you’re on camera still is on, and sometimes people can still hear me.
64 00:05:25.590 ⇒ 00:05:27.829 kaybay: I was like, let’s just hop on the phone.
65 00:05:27.830 ⇒ 00:05:34.430 Uttam Kumaran: That’s okay, that’s okay. Great. So, I think, I just shared with Kyle that
66 00:05:34.700 ⇒ 00:05:48.299 Uttam Kumaran: we put together a little bit of, like, what we described last time, which is, like, okay, let’s talk about some specific work streams. I’m happy to do another brief intro if you think that’s helpful, Catherine, or… okay. Yeah, so, Kyle, I…
67 00:05:48.300 ⇒ 00:05:59.649 Uttam Kumaran: got first put in touch with the team at CTA, a few months ago, sort of before Catherine came on, and was just, sort of exploring, you know, where you guys need data help, and I think
68 00:05:59.650 ⇒ 00:06:11.240 Uttam Kumaran: Me and Catherine talked last week more specifically about the state of the world, there, and my background, I’m a data engineer. I worked as a data engineer, in New York, actually, for, a number of
69 00:06:11.240 ⇒ 00:06:14.220 Uttam Kumaran: of companies, I led data teams, and I started this business.
70 00:06:14.220 ⇒ 00:06:17.140 Kyle Wandel: I started this business here…
71 00:06:17.140 ⇒ 00:06:37.850 Uttam Kumaran: In Boston, like, roughly 2 years ago, and we’re a data analytics and AI consultancy, so my team is filled with former internal data folks that have built, sort of, reporting, so everything from ETL to warehousing to, data modeling and dbt, and then finally, like, BI and reporting.
72 00:06:37.850 ⇒ 00:06:51.919 Uttam Kumaran: So that’s sort of our bread and butter. More recently in the company, in the last year, we’ve also started to, develop, like, AI services and solutions. So these can be things like AI on top of your data to ask basic questions, like lookups and things like that.
73 00:06:51.920 ⇒ 00:07:05.490 Uttam Kumaran: But also building agentic workflows and things, so really excited, because I think a lot of the data world is getting pulled forward by that word. But it also, I think it allows more people to get access to data, versus having to
74 00:07:05.740 ⇒ 00:07:18.559 Uttam Kumaran: self-service or look through a dashboard that may not be built for them. So, I think Catherine and I kind of left off at that point. We didn’t go into a ton of specifics around all of the key issues, although I do understand that there are some legacy systems.
75 00:07:18.560 ⇒ 00:07:27.740 Uttam Kumaran: There are tons of different CRM sources. There doesn’t seem to be a cohesive source of truth for some information.
76 00:07:27.890 ⇒ 00:07:45.130 Uttam Kumaran: Yeah, there’s a thing, a lot of things around, like, identity resolution and enrichment. And then, lastly, it’s also, you know, my background has been working, in building data teams and procuring data infrastructure, and so very opinionated about the tools that we use.
77 00:07:45.130 ⇒ 00:07:47.960 Uttam Kumaran: And ideally trying to help you guys set up
78 00:07:48.080 ⇒ 00:08:02.719 Uttam Kumaran: infrastructure that not only supports your clients, your users, but also that helps your workflows, and that you guys don’t, get angry about all the, you know, painful tooling that exists. And so that’s a lot of, like, how we think about things.
79 00:08:02.780 ⇒ 00:08:07.179 Uttam Kumaran: So I, I roughly put together, and I’ll just share…
80 00:08:07.190 ⇒ 00:08:10.720 Uttam Kumaran: this on my screen, roughly put together, like, these…
81 00:08:10.730 ⇒ 00:08:27.519 Uttam Kumaran: for, like, work streams. They’re kind of… they’re a little bit generic now, so hopefully in this meeting we can talk a little bit about, ideally the quadrant for us to focus on is, like, what is the most important? That could be based on ROI or, like, timeline, and then also, like,
82 00:08:27.520 ⇒ 00:08:42.620 Uttam Kumaran: what is, like, the easiest and hardest to actually accomplish. So we kind of have, like, data foundations and pipelines. So I think, Kyle, this may be a lot of where you’re coming in and stepping in, which is, like, getting all of the data from all our resources in one place.
83 00:08:42.630 ⇒ 00:08:58.930 Uttam Kumaran: Making sure we have those API keys that we could structure pipelines. The second is, like, BI reporting. So I know that, like, there is Power BI. Do we want to talk about another, you know, way for folks to get access to data?
84 00:08:58.930 ⇒ 00:09:14.100 Uttam Kumaran: The third piece is sort of identity resolution. Again, it’s… a lot of the business, I know, is around customers, so how can we get a great view of those customers from all their touchpoints into one place that makes it easy for business users to look them up?
85 00:09:14.170 ⇒ 00:09:32.110 Uttam Kumaran: We can also talk about enrichment, and then finally, like, how does that end up back into a CRM, or back into some type of Klaviyo? Typically, these are, like, reverse ETL use cases. And the last thing is sort of just being, like, kind of a… I kind of describe it as, like, executive foil, which is just, like.
86 00:09:32.460 ⇒ 00:09:45.239 Uttam Kumaran: thinking about the strategy of, like, either the team organization, or people, process, or procurement, and, you know, sort of, like, just being a helpful assist there.
87 00:09:45.240 ⇒ 00:09:55.010 Uttam Kumaran: So maybe I’ll pause there and just talk about, like, do these four roughly seem, like, right? And maybe we could talk about, like, priorities or what’s been done so far.
88 00:09:55.060 ⇒ 00:09:56.739 Uttam Kumaran: Yeah, I can go ahead.
89 00:09:57.670 ⇒ 00:10:10.929 kaybay: Yeah, I think this is, a perfectly awesome start, definitely in line with what we talked about last time. So I think, some additional thoughts that might help us as we refine the list,
90 00:10:10.930 ⇒ 00:10:19.389 kaybay: So we just had the call with the AWS Professional Services team, because we were interested in working with them to kind of build out the control tower landing zone kind of thing.
91 00:10:19.390 ⇒ 00:10:26.129 kaybay: We do already have an AWS account and that kind of stuff. I’m just kind of like, I’d rather start fresh with something clean that we know is secure.
92 00:10:26.130 ⇒ 00:10:45.930 kaybay: So I think we’ll be able to lean on them for a lot of that, like, foundational, foundational AWS work, but I’m sure you know, like, they don’t do… or what would be out of scope would be, like, migration of existing resources, or helping us set up Snowflake, and stuff like that, so I think leaving the foundational piece in here is still… makes sense to me.
93 00:10:45.940 ⇒ 00:11:06.779 kaybay: And then additionally, I think I’d mentioned we were talking to a consulting group that a friend of mine had recommended, SDG. They are definitely a little bit more on the, like, sort of buttoned up and traditional, and I think they’re probably used to coming into environments that are clean and just need a little improvement versus the, like, hey, come join our chaos.
94 00:11:06.780 ⇒ 00:11:13.670 kaybay: So, I think for them, I asked them to scope out, like, a Power BI, sort of.
95 00:11:13.830 ⇒ 00:11:20.890 kaybay: review of the existing environment, put out, like, the fires that need to be put out this year with the current cycle CES.
96 00:11:20.890 ⇒ 00:11:45.879 kaybay: You know, and then kind of, like, that was the limit for that scope. The reason I like leaving, some of this stuff in here, though, is because, looking at the next year, even though it would be, you know, next year’s work, next year’s budget, I do think moving off of Power BI, or at least having a very serious conversation around it, is in our future. And so I think getting a sense while we’re doing the work this year of, like, whether or not we do want to stick with Power BI,
97 00:11:45.880 ⇒ 00:11:57.109 kaybay: versus going a different direction makes a lot of sense. Plus, also, there’s enough reporting to spread over a million consultants at this point. But I will say, too, I looked into that Omni,
98 00:11:57.110 ⇒ 00:11:57.430 Uttam Kumaran: Yeah.
99 00:11:57.430 ⇒ 00:11:59.499 kaybay: It looks really cool, so…
100 00:11:59.500 ⇒ 00:12:00.190 Uttam Kumaran: Really good.
101 00:12:00.190 ⇒ 00:12:00.950 kaybay: Yeah, huh?
102 00:12:00.950 ⇒ 00:12:17.060 Uttam Kumaran: I’m happy to give you access even to our demo instance if you want to play around. Like, we have a bunch of specific data on there, and I am not very easily impressed by data tools, I don’t know, and I really like it too. I think,
103 00:12:17.300 ⇒ 00:12:23.530 Uttam Kumaran: they’ve… they’ve sort of done, like, not just, like, a Looker Plus, but the development
104 00:12:23.740 ⇒ 00:12:31.710 Uttam Kumaran: life cycle of objects in there is really nice, and, like, the UI and being able to do spreadsheets natively, like.
105 00:12:32.060 ⇒ 00:12:34.109 Uttam Kumaran: It’s really, really nice, so…
106 00:12:34.110 ⇒ 00:12:34.580 kaybay: Yeah.
107 00:12:34.580 ⇒ 00:12:36.640 Uttam Kumaran: Something that we’re exploring a lot now.
108 00:12:37.290 ⇒ 00:12:49.199 kaybay: Yeah, I feel like I have a high, like, sensitivity to vaporware, I was kind of, like, I was wondering, you know, whether their website pass my laptop.
109 00:12:53.150 ⇒ 00:12:54.520 Uttam Kumaran: I think we’re back on the laptop.
110 00:12:54.520 ⇒ 00:13:01.130 Kyle Wandel: Yeah, exactly. Trying to be, at least. Yeah, as Catherine alluded to, our laptops are quite fun right now.
111 00:13:06.140 ⇒ 00:13:10.970 Uttam Kumaran: I guess, Kyle, where are you, where are you joining, CTA from before?
112 00:13:10.990 ⇒ 00:13:16.309 Kyle Wandel: Well, I’ve actually… so, I’m… I’ve been with CTA, so we have a… we have a market research department.
113 00:13:16.310 ⇒ 00:13:17.560 Uttam Kumaran: Oh, that’s right, that’s right.
114 00:13:17.710 ⇒ 00:13:31.019 Kyle Wandel: Yeah, my background is more data analytics and more data analysis in general, but like you mentioned at the very beginning of the call, AI and Genetic AI is gonna, like, starting to hit that realm pretty hard, so I went back to school and actually focusing more on data engineering stuff, and… Nice.
115 00:13:31.020 ⇒ 00:13:49.679 Kyle Wandel: I’ve done… I mean, I’ve done data… I’ve been around data for, like, 10 plus years, so I understand pretty much every part of the data stream, from the very beginning, foundational work, all the way to the analytics and reporting, but this is just really getting my feet wet more in, like, the data engineering standpoint, actually having an official title and term to it, so…
116 00:13:49.680 ⇒ 00:13:53.370 Uttam Kumaran: Cool, cool, great, awesome. Yeah, I don’t think the analyst stuff is actually, like.
117 00:13:54.040 ⇒ 00:14:01.360 Uttam Kumaran: going away, I just think it’s… a lot of the basic stuff, I would… you hope people can fish for themselves, right?
118 00:14:01.360 ⇒ 00:14:22.030 Kyle Wandel: Yeah, it’s… like you said, it’s more so just, like, being able to use a dashboard, like, I feel like that’s slowly going away, but you still need to have that analyst, that scientist that is able to, like, verify all the data, and to make sure the data is clean, and correct, and getting to the right places at the right time, and really educating the… where we are, educating our workforce, and how to use, and what questions to ask, basically.
119 00:14:22.050 ⇒ 00:14:23.419 Kyle Wandel: be consistent.
120 00:14:24.390 ⇒ 00:14:29.099 Uttam Kumaran: Yeah, I definitely think the ask better questions thing is a big one. Yes, yeah.
121 00:14:29.130 ⇒ 00:14:30.890 kaybay: Sorry about the…
122 00:14:30.890 ⇒ 00:14:31.990 Uttam Kumaran: No, no, no, it’s like…
123 00:14:32.000 ⇒ 00:14:38.809 kaybay: I had a… I was afraid it would start doing the, like, echo thing back and forth. I’m gonna just stay on my phone rather than trust the laptop.
124 00:14:38.810 ⇒ 00:14:39.590 Uttam Kumaran: Okay.
125 00:14:39.590 ⇒ 00:14:50.040 kaybay: Okay, my last piece of my thought around this stuff was, the identity resolution, I think is probably the…
126 00:14:50.830 ⇒ 00:15:02.900 kaybay: it is definitely one of the most urgent things, I think, to solve for. And I also think this is probably something that’s really well-suited to, at least as far as you’ve described, like, your company and the people you have, because
127 00:15:03.000 ⇒ 00:15:11.889 kaybay: I really don’t. I know there are answers, I know identity resolution is an entire, you know, field of science, right? Like, there’s totally different tactics out there. I’ve dabbled in some.
128 00:15:11.890 ⇒ 00:15:27.710 kaybay: but bringing some additional brainpower to this to just kind of, like, crunch and solve it, to the extent possible, rather than kind of leaving it for me and Kyle to, like, chip away at. I will admit, since we spoke last, I’ve, I’ve been…
129 00:15:28.520 ⇒ 00:15:50.710 kaybay: just, like, tearing my hair out, trying to get the list of web domains that should authenticate somebody as a member for registration together, and it sounds so easy, and yet here I am four days later, and I’m like, I’ll try and get it to you today. So I think that might be an area that would make a lot of sense for us to work together on, so that we can kind of buy some speed and get through a little bit of that.
130 00:15:50.710 ⇒ 00:15:54.370 Uttam Kumaran: a good North Star for a lot of the other work streams, like…
131 00:15:54.370 ⇒ 00:15:55.100 kaybay: Yeah, exactly.
132 00:15:55.100 ⇒ 00:16:02.470 Uttam Kumaran: This… it is what gets powered by all the pipelines, so if the AWS work from those guys.
133 00:16:02.570 ⇒ 00:16:10.480 Uttam Kumaran: like, the… their, like, their standard is set by that work stream. Like, if their… all their work needs to be able to support this.
134 00:16:10.480 ⇒ 00:16:27.109 Uttam Kumaran: Secondarily, it all needs to be easily reportable, so they have to be marts that you can put Power BI on top of that, of course, refresh and have cadences and can be validated. So, I do think that, like, I agree, a lot of where we excel, and we’ve done
135 00:16:27.110 ⇒ 00:16:39.559 Uttam Kumaran: in my career, I’ve done dbt data modeling, and Snowflake my whole career, and that’s what we’re really, really great at. But everything goes to power that, and so whatever data we need to build pipelines for.
136 00:16:39.560 ⇒ 00:16:54.469 Uttam Kumaran: it’s for that reason, and then similarly, the marts and the structure at which we set up dbt, which is, like, raw, staging, intermediate marts, like, dimension and fact tables, all go to enable the reporting layer, so that,
137 00:16:54.470 ⇒ 00:17:08.429 Uttam Kumaran: Kyle and whoever it is in your past life, or other people that are reporting, they have an easy way of getting it. They’re not just, like, 5 tables that have 100 columns, and, like, they take hours to refresh. There is, like, an established data mart with
138 00:17:08.430 ⇒ 00:17:16.680 Uttam Kumaran: testing with some level of observability, run ads, things like that. So that’s a perfect way to utilize us. We have a… we have a lot of clients that
139 00:17:16.680 ⇒ 00:17:32.139 Uttam Kumaran: That’s what we do, and then our job is to enable those analysts, so we… we’re making sure that they know which tables to leverage, that if we… that we have verifiable columns or verifiable tables versus something that may have gone created once and sort of
140 00:17:32.240 ⇒ 00:17:40.529 Uttam Kumaran: left. And it’s actually nice that you guys are starting some stuff fresh, because sometimes we will have to walk into an environment, and it’s a lot of, like.
141 00:17:40.700 ⇒ 00:17:45.030 Uttam Kumaran: We need to fix existing stuff and build it, and then, like, kind of, like, hot swap.
142 00:17:45.180 ⇒ 00:17:47.640 Uttam Kumaran: And it’s, like, it’s painful, so…
143 00:17:47.650 ⇒ 00:17:50.299 kaybay: Yeah, it’s very painful.
144 00:17:50.750 ⇒ 00:18:00.760 kaybay: I mean, like, so actually, I mean, like, today, since now that registration is open, there’s, like, an inbox for, like, you know, registration issues and customer support kind of thing.
145 00:18:00.880 ⇒ 00:18:11.989 kaybay: and the… one of my counterparts on the marketing team was like, a lot of these emails are people saying, like, hey, you know, so-and-so’s left the company, I should be the new contact now. And I’m like, that’s great.
146 00:18:12.020 ⇒ 00:18:21.299 kaybay: I can totally make that change in my data, but in terms of getting that pushed out to the 8 other places that is probably in this organization, I’m like, right?
147 00:18:21.300 ⇒ 00:18:21.820 Uttam Kumaran: Yeah.
148 00:18:21.820 ⇒ 00:18:26.439 kaybay: Yeah, so many, so many little questions that’ll come out of it. Yeah.
149 00:18:26.440 ⇒ 00:18:36.430 Uttam Kumaran: But that’s also something to consider for the pipeline work that’s unique here, is I think there’s also pipelines in and then pipelines out, right? Like, once you do the solution and you have
150 00:18:36.580 ⇒ 00:18:53.780 Uttam Kumaran: you know, a concrete understanding of that customer, what systems does that data need to go back into? Whether it is for, like, marketing, like emails, whether it is for, for mail, for whatever the outcomes are, how does that data get back into there? So.
151 00:18:53.780 ⇒ 00:18:58.479 Uttam Kumaran: I think, Kyle, on your part, it is a really interesting project, because you have both sides.
152 00:18:58.500 ⇒ 00:19:05.749 Uttam Kumaran: Of that, and that data has to be up-to-date pretty quickly. So this is… this identity resolution piece
153 00:19:05.820 ⇒ 00:19:21.790 Uttam Kumaran: And again, in identity resolution, there are both a lot of vendors that do this, but I do… I am hearing that this is something I think, that we want CTA to own, and I think given the complexity, it will be a pretty penny to go buy something, and they will promise you that they can do this, and…
154 00:19:22.030 ⇒ 00:19:27.729 Uttam Kumaran: like, it’s not… it’s… it will take quite a while, you know? Sure, Catherine, as you know, you know.
155 00:19:27.980 ⇒ 00:19:42.089 kaybay: Right, I mean, that’s… that’s exactly kind of my thought, is like, you know, even if we went out and bought some sort of, you know, seven-figure, you know, unicorn of a system that proposed to do it, I think we just still haven’t even done enough of the discovery struggle to figure out
156 00:19:42.090 ⇒ 00:19:48.950 kaybay: you know, maybe it’s not that hard, maybe we just haven’t had a chance to do it yet, or maybe we really do need a very specific feature set, or etc, etc, so…
157 00:19:49.370 ⇒ 00:19:50.979 kaybay: Oh, sweet, yeah.
158 00:19:50.980 ⇒ 00:19:56.510 Kyle Wandel: I mean, that’s kind of where my institutional knowledge comes in, in terms of I could just describe, like, 10 to 15
159 00:19:56.510 ⇒ 00:20:19.309 Kyle Wandel: 20 different use cases, basically, of how we’ve had to change something over time and how it didn’t track over time, and quite frankly, I think the most important thing, like Kevin said, the basics of it is, like, understanding when an individual inputs something on their registration form, or another particular form as a free-form text, is that the data is able to pick up on that and put them in the right bucket, basically.
160 00:20:19.310 ⇒ 00:20:26.009 Kyle Wandel: that’s something that, quite frankly, is what I think is most important from a membership standpoint and a categorization standpoint.
161 00:20:26.070 ⇒ 00:20:29.129 Kyle Wandel: And so that would be perfect if you guys could help us out with that.
162 00:20:29.740 ⇒ 00:20:30.440 Uttam Kumaran: Right.
163 00:20:30.440 ⇒ 00:20:31.130 kaybay: Yeah.
164 00:20:31.240 ⇒ 00:20:35.660 kaybay: So I think painting the picture altogether, I think
165 00:20:35.780 ⇒ 00:20:44.010 kaybay: perhaps, kind of blending from your buckets. Definitely the strategic advisory thing all day, bring it on. I think probably looking at
166 00:20:44.010 ⇒ 00:21:08.949 kaybay: like, foundational Snowflake and dbt in support of the identity resolution work, so setting those up so that we can focus on that identity resolution, and then for the near term, probably reporting out to Power BI, but as we go through these journeys together, figuring out, like, what business intelligence platform makes more sense to onboard with next year’s funding. That’s kind of where my head is at, but I’m totally open to thoughts and refinement.
167 00:21:09.180 ⇒ 00:21:15.650 Uttam Kumaran: And then if you can tell me, like, what is… what are the specifics of the AWS Professional Service? Like, what do you have in scope to take on?
168 00:21:15.880 ⇒ 00:21:26.650 kaybay: Yeah, so it’s… it’s a pretty narrow scope, to be honest. It’s basically, they will deploy… and I mean, I know you’ll probably laugh, because it’s like, this is a solution in the marketplace that you can, like, click to launch.
169 00:21:26.740 ⇒ 00:21:43.120 kaybay: But there’s enough of a networking component that I just want to have experts in-house to help with. So basically, what they’ll do is they’ll deploy the foundational pieces for control tower, landing zone, they’ll connect it to our Active Directory auth provider, which we use Okta.
170 00:21:43.120 ⇒ 00:21:50.820 kaybay: And then they’ll put out together all of the, like, runbooks and IAC templates so that we can deploy accounts as we need them. I mean…
171 00:21:50.820 ⇒ 00:22:10.099 kaybay: The reality is, I probably foresee us just having one stream of, like, playground dev staging prod, but the AWS kind of recommended best practice is the ability to deploy accounts in that set for each project or organizational unit, so we’ll have that ability, even if the near-term reality is that it’s just kind of a four-account sort of setup.
172 00:22:10.110 ⇒ 00:22:18.729 kaybay: And then they’ll also help us with putting in place all of the AWS config rules and the logging and monitoring and that kind of stuff, so…
173 00:22:18.730 ⇒ 00:22:20.039 Uttam Kumaran: I think it’s worth it, like.
174 00:22:20.040 ⇒ 00:22:20.520 kaybay: Yeah, huh.
175 00:22:20.520 ⇒ 00:22:24.949 Uttam Kumaran: It’s also work that, like, nobody wants to do, and it’s sort of like…
176 00:22:25.140 ⇒ 00:22:33.769 Uttam Kumaran: And honestly, it’s so confusing once you open up Console. Like, I open up Console all the time, and still I’m like, where am I?
177 00:22:33.770 ⇒ 00:22:34.920 kaybay: Right?
178 00:22:34.920 ⇒ 00:22:41.670 Uttam Kumaran: So, I don’t know, I feel like it’s worth them doing it. They’ve made their tool extremely complicated, they can set it up, so…
179 00:22:41.670 ⇒ 00:22:57.809 kaybay: Right? Yeah, exactly. And then I think it’ll just kind of give everybody a little bit of peace of mind, like, you know, okay, the foundation is solid. We had it built by the people that build the thing, now we can put stuff on top of it without needing to worry about, like, oh, what is that VPC config again?
180 00:22:57.810 ⇒ 00:23:07.750 Uttam Kumaran: Yeah, exactly. Yeah. And then for Snowflake, like, is that… do you guys get that separately, then, from AWS? And then, yeah, I guess maybe that’s the first question.
181 00:23:07.750 ⇒ 00:23:16.859 kaybay: So, I’m trying to be a little cheeky here. You can purchase the Snowflake instance through the marketplace, which will circumvent our contracting process.
182 00:23:16.860 ⇒ 00:23:17.540 Uttam Kumaran: Yes.
183 00:23:18.440 ⇒ 00:23:20.790 kaybay: This is… this is recorded, isn’t it? Damn it.
184 00:23:21.290 ⇒ 00:23:32.459 kaybay: I have asked the IT guy, he was like, that’s fine. So I’ll be able to do the Snowflake instance through the marketplace. I will probably… I need to do it before October, when we get access to that data share.
185 00:23:32.460 ⇒ 00:23:42.070 kaybay: So I’ll probably wind up spinning up that instance in my current playground account, and then as soon as we have the landing zone, kind of hot-swap it over there. But there is…
186 00:23:42.070 ⇒ 00:24:00.949 Uttam Kumaran: It depends, yeah, I don’t know, it depends, like, if… yeah, I guess if you guys got a great deal from AWS, then you could put it in there. You could also just do both and, like, see what comes out right. The main question I have is this, like, if they have extra scope, they can make sure that, like, storage integrations and stuff are set up.
187 00:24:01.360 ⇒ 00:24:26.079 kaybay: Yeah, I think leaning on them for some of that does make sense. I will admit, like, I have worked with AWS’s, like, database, sort of, solution architects before, and they’re very, very, very smart, lovely people, but they also really struggle to, like, adapt the best practices to the, like, realities of a 150-person nonprofit team. And so I’m kind of like, I like them for the infrastructure side.
188 00:24:26.080 ⇒ 00:24:28.379 kaybay: So, for the actual, like, data management.
189 00:24:28.380 ⇒ 00:24:32.370 Uttam Kumaran: I mean, honestly, even easier is if I just have one place where I can see, like.
190 00:24:33.070 ⇒ 00:24:35.579 Uttam Kumaran: EPC names and things, that would be…
191 00:24:35.580 ⇒ 00:24:36.150 kaybay: Yeah.
192 00:24:36.150 ⇒ 00:24:37.580 Uttam Kumaran: That’s ideal, so…
193 00:24:37.580 ⇒ 00:24:39.830 kaybay: Yeah, yeah, exactly.
194 00:24:40.240 ⇒ 00:24:42.340 kaybay: Oh, are we being defended?
195 00:24:42.520 ⇒ 00:24:46.589 Uttam Kumaran: No, but there’s just… Come here, relax, relax, relax.
196 00:24:46.840 ⇒ 00:24:49.060 Kyle Wandel: He’s alert, he’s alerting, he’s alerting you.
197 00:24:49.060 ⇒ 00:24:50.009 kaybay: That one.
198 00:24:50.010 ⇒ 00:24:50.730 Kyle Wandel: He’s alerting you.
199 00:24:50.730 ⇒ 00:24:52.209 Uttam Kumaran: We can just go…
200 00:24:55.010 ⇒ 00:25:01.400 Kyle Wandel: I know what that’s like. Our dogs go absolutely bonkers, especially the Shepherd. I think it happened last…
201 00:25:02.870 ⇒ 00:25:08.380 kaybay: Yeah. I used to have a little dog who would bark in the window, and then you could see his, like, shadow on the wall behind me on calls.
202 00:25:10.120 ⇒ 00:25:11.580 kaybay: Very cartoonish.
203 00:25:25.930 ⇒ 00:25:27.550 kaybay: My phone balancing.
204 00:25:28.610 ⇒ 00:25:29.559 kaybay: Let’s try this.
205 00:25:29.560 ⇒ 00:25:32.610 Kyle Wandel: Is it blue screen, or it just… it just turned off, went black?
206 00:25:33.310 ⇒ 00:25:34.610 kaybay: Oh, me, or…
207 00:25:34.610 ⇒ 00:25:36.919 Kyle Wandel: The, no, the, your laptop that it blew.
208 00:25:36.920 ⇒ 00:25:37.640 kaybay: I don’t know.
209 00:25:37.640 ⇒ 00:25:38.280 Kyle Wandel: blood.
210 00:25:38.580 ⇒ 00:25:50.840 kaybay: Yeah, it does this thing where, yeah, it just, like, turns black, and, like, you can’t, like, see anything, or access anything, or, like, you know, none of the keys do anything. It’s like it’s frozen, I guess, kind of, but…
211 00:25:50.840 ⇒ 00:25:55.609 Uttam Kumaran: It happens on my laptop, too, and then some services stay live, and then some…
212 00:25:56.360 ⇒ 00:25:58.000 kaybay: Yeah. I don’t know.
213 00:25:58.140 ⇒ 00:26:04.660 kaybay: Honestly, I was really surprised when it, like, just suddenly started going again, because I even closed the clamshell, and I’m like, what? Like, okay.
214 00:26:07.390 ⇒ 00:26:10.960 kaybay: When I see Lenovo at CES, I’ll have a word with them about their ThinkPads.
215 00:26:14.290 ⇒ 00:26:24.960 Uttam Kumaran: Okay, cool, so I feel… I feel kind of pretty good about where we land. I think the biggest thing is, like, I think the IR work is a great downstream, like,
216 00:26:25.080 ⇒ 00:26:30.249 Uttam Kumaran: basically a user of all the AWS work, so I just want to establish that. And then I think I’ll…
217 00:26:30.310 ⇒ 00:26:49.480 Uttam Kumaran: probably working with Kyle, since you have a lot of the institutional knowledge, understand, like, where this is landed Power BI, and, like, the data model requirements, I think, can probably come a lot from you. And then I think we… I think a lot of the scope initially is, like, setting up dbt best practices, setting up version control, wherever that lands, orchestration for that.
218 00:26:49.600 ⇒ 00:27:06.089 Uttam Kumaran: Also would love to, you know, also suggest a decision on, like, some type of observability tool. Like, we’ve had a lot of success with Metaplane. Something where you can just understand, like, column values, and you can set up alerting and tests right out the gate.
219 00:27:06.400 ⇒ 00:27:20.119 Uttam Kumaran: That way, when we tend to ship data models, we try to also ship tests and, you know, understanding of, like, values and, you know, run at times, updated at times, things like that.
220 00:27:22.000 ⇒ 00:27:29.170 Uttam Kumaran: And then I think, yeah, long-term, I think we can consider options for evaluating other BI tools. The lovely thing about
221 00:27:29.170 ⇒ 00:27:45.170 Uttam Kumaran: we just have… I try every single thing in market. Because we’re a consultant, they like us, because we… we’re typically… I either have the ability to say, yeah, you should consider them, or, like, no. So we can get demo instances for everything, and, like, also, I can… you don’t have to talk directly to them.
222 00:27:45.170 ⇒ 00:27:57.249 Uttam Kumaran: Which is also a real joy, so, like, if they… I know most of the people at all these places, so if we want to get something set up, they’ll give us stuff, and we can just play with it for a while until we’re comfortable.
223 00:27:57.550 ⇒ 00:28:00.949 Uttam Kumaran: Cool, so… Yeah, go ahead.
224 00:28:01.150 ⇒ 00:28:07.849 kaybay: Oh, no, I should say, the observability piece, yeah, is a good one. I don’t know if we had necessarily talked about it before, per se, but, like.
225 00:28:08.070 ⇒ 00:28:11.529 kaybay: It’s always something I’ve done, kind of, like, with duct tape and popsicle sticks.
226 00:28:11.530 ⇒ 00:28:12.040 Uttam Kumaran: Yes.
227 00:28:12.040 ⇒ 00:28:16.030 kaybay: Like, yeah, like, kind of very, like, manual hands-on, but moving to a more.
228 00:28:16.030 ⇒ 00:28:26.449 Uttam Kumaran: Yeah, like, I’ve done it too, where I’ve run SQL queries, where it’s like, look at the values within this. One thing that I realized is it just has to be some… like, one, you have to have a system where
229 00:28:26.520 ⇒ 00:28:38.000 Uttam Kumaran: just like P0 and P1 alerts are able to be monitored, and then that can get sent into, like, either PagerDuty or Slack or whatever, wherever it gets… these get picked up, and that you have, like, runbooks.
230 00:28:38.650 ⇒ 00:28:45.270 Uttam Kumaran: So there’s a lot of process stuff here. The other thing about Metaplane that I like is they just do, like.
231 00:28:45.440 ⇒ 00:28:48.160 Uttam Kumaran: Data table monitoring very well.
232 00:28:48.320 ⇒ 00:29:00.619 Uttam Kumaran: tools like Monte Carlo, DataFold, they’re trying… because they raise a lot of money, they’re trying to, like, become the star of your stack, when in fact, I think they’re just supposed to be just, like, a plumbing
233 00:29:00.620 ⇒ 00:29:08.900 Uttam Kumaran: In the wall, and they should… yes, it’s, like, almost like an insurance policy, where it should help you identify issues, and then ideally.
234 00:29:08.920 ⇒ 00:29:21.039 Uttam Kumaran: triage faster, and then prove that there are no issues, right? Like, both of those, I think, all those three matter. A lot of these tools, you’ll find that they’re more expensive than, like.
235 00:29:21.260 ⇒ 00:29:38.390 Uttam Kumaran: the parts in your stack where you’re storing or running data, it doesn’t make any sense. But we’ve… I found that Metaplan has been pretty good. You can do standard deviations for common metrics. They have historical, so you can see, like, the job, like, which jobs are running for a long time, what to mitigate.
236 00:29:38.580 ⇒ 00:29:44.660 Uttam Kumaran: It’s, like, 500 bucks a month, typically, and I don’t know if you guys use Datadog, but they got purchased by Datadog, so…
237 00:29:44.900 ⇒ 00:29:47.019 kaybay: Oh, okay, okay. Yeah, I used…
238 00:29:47.020 ⇒ 00:29:47.860 Uttam Kumaran: Yeah.
239 00:29:47.860 ⇒ 00:30:01.010 kaybay: Yeah, I used Datadog at my last place. I really liked it. I mean, I was blown away by how powerful that tool is. There’s a lot of config, but it was super powerful, and I’ve been encouraging Jay to kind of consider it for next year. I actually asked the ProServe folks
240 00:30:01.010 ⇒ 00:30:21.179 kaybay: I was like, are you guys gonna make us use, or not make us, but, like, set it up with, like, CloudWatch CloudTrail, or push us towards Datadog? And they’re like, oh, we would use our services, and I’m like, oh, interesting, okay. So, I mean, I’m fine getting started there, but yeah, I think in the next year or so, I would like to see if Jay’s got appetite to use Datadog across the board, because it’s a really nice platform.
241 00:30:21.180 ⇒ 00:30:28.319 Uttam Kumaran: Yeah, I still don’t think there’s… I think they’re… they’re expensive, but I don’t think there’s a better winner. Like, LogRocket is the only other company I’ve heard of.
242 00:30:28.320 ⇒ 00:30:29.460 kaybay: Yeah.
243 00:30:29.460 ⇒ 00:30:44.390 Uttam Kumaran: But, I mean, for our use case, really, what I want to know is, like, for core metrics that we’re standing… we’re within values, I want to know which tables haven’t run in a while. And then I was just talking to another client, it’s just like, we want to find out there’s a problem faster than the business.
244 00:30:44.470 ⇒ 00:30:53.580 Uttam Kumaran: You know, and that’s… until that point, I think that’s our… like, that’s our North Star there, you know? So, it’s not the fact that data issues won’t happen.
245 00:30:53.830 ⇒ 00:30:54.300 kaybay: Right.
246 00:30:54.300 ⇒ 00:30:56.330 Uttam Kumaran: We have to be able to find out and triage it.
247 00:30:56.800 ⇒ 00:30:57.200 kaybay: Dude.
248 00:30:57.200 ⇒ 00:30:59.229 Uttam Kumaran: Yeah, it’d be the first action, so…
249 00:30:59.690 ⇒ 00:31:07.500 kaybay: Yeah, it always looks better if we’re the team that finds the problem, and then also fixes it, versus, is told about the problem and eventually fixes it.
250 00:31:07.500 ⇒ 00:31:08.150 Uttam Kumaran: Yeah.
251 00:31:08.420 ⇒ 00:31:27.910 Uttam Kumaran: So I guess, like, in terms of, you know, I think I’ve… I think we’re pretty settled on, sort of, focusing on identity resolution. I guess, like, I guess, talk to me about, sort of, like, what you’re thinking about for, like, a team and, like, sort of timelines and stuff. I mean, I think the big questions for me is, it seems like it would most likely be
252 00:31:28.050 ⇒ 00:31:48.020 Uttam Kumaran: I think what would seem fair is, like, me and one person with one engineer on our side that can come on. For all of our, clients, we do also have… everything runs through our delivery team, so we have a project manager, we do linear tickets, and we do, sort of, like, both weekly and, like, monthly updates, and, like.
253 00:31:48.020 ⇒ 00:31:57.059 Uttam Kumaran: we can run as beefy of a sprint as you want. As I mentioned last time, we’re also very opinion about this, like, running a data team, like, the day-to-day processes, so…
254 00:31:57.060 ⇒ 00:32:00.650 Uttam Kumaran: Stand-ups and things… and, like, grooming and things like that are all things that we…
255 00:32:00.650 ⇒ 00:32:09.529 Uttam Kumaran: we could do. Of course, for some teams, they have typical project managers. Some teams, it’s… like, for some of our… some of our clients, they’re so scrappy that it’s just, like.
256 00:32:09.550 ⇒ 00:32:22.910 Uttam Kumaran: everything’s async, so I think for us, it’s important to establish, like, the rituals of a great data team as well. But that… that would sort of be, like, probably the format of the… of the team for ours, from our side.
257 00:32:23.470 ⇒ 00:32:39.109 kaybay: Yeah, I think that makes perfect sense, to be honest. And I think, especially given that, you know, there’ll be, you know, probably, like, November, December timeline, like, a lot of overlap between the different engagements, and so having everybody singing off that same sheet of music is gonna be critical.
258 00:32:39.110 ⇒ 00:32:50.129 kaybay: the ProServe folks, they do always put a PM on their thing. I think SDG… off the top of my head, I think they actually didn’t have a PM resource that was baked into the scope by default.
259 00:32:50.130 ⇒ 00:32:57.329 kaybay: So I think between the AWS person, your person, and then the other humans involved, yeah, I think we’re probably…
260 00:32:57.330 ⇒ 00:33:11.860 Uttam Kumaran: I’m not a big, like, PM gatekeeps everything, or everything’s super, like… again, like, most… we’re, like, an engineering company first, so all of our engineers are required to be, like, customer-facing, or at least, like, can be, right? So…
261 00:33:11.860 ⇒ 00:33:19.990 Uttam Kumaran: it’s not a… it’s not, like, a camera off and, like, everything goes through PM, like, sort of org. Everybody’s really, really…
262 00:33:20.000 ⇒ 00:33:21.260 Uttam Kumaran: collaborative.
263 00:33:21.370 ⇒ 00:33:24.300 Uttam Kumaran: It’s just nice that they keep us really organized.
264 00:33:24.580 ⇒ 00:33:25.350 kaybay: Yeah.
265 00:33:25.350 ⇒ 00:33:26.390 Uttam Kumaran: Yeah.
266 00:33:27.000 ⇒ 00:33:32.749 Uttam Kumaran: That’s the right flavor. Yeah, yeah, I agree. I used to be a product manager, so I kind of, like.
267 00:33:33.080 ⇒ 00:33:41.500 Uttam Kumaran: I was an engineer first, I became a product manager, and then I, like, did my own flavor of, like, what I felt like there, so…
268 00:33:41.640 ⇒ 00:33:44.440 Uttam Kumaran: Cool, and then I assume that… yeah, go ahead, go ahead.
269 00:33:44.600 ⇒ 00:34:04.320 kaybay: I would just say, in terms of, like, next steps, that kind of thing, so I have a meeting on Wednesday with my boss, who’s our COO, and then, my counterpart on the IT side, Jay, who’s the VP of IT. And my plan is, I’ve got kind of all my homework done, so I’m going to present to them, here’s the scope from AWS, from SDG, from Brainforge.
270 00:34:04.320 ⇒ 00:34:25.010 kaybay: you know, this is kind of the total dollar amount we’re looking at. I know it’s a large price tag for an end-of-year project, but, you know, pitching them that story, like we talked about, right? Like, helping them understand that, yes, it’s a lot of money to spend all at once, but it’s also building a foundation that we can then build many, many, many things on top of. So, I’m hoping that after that meeting, I’ll have, sort of.
271 00:34:25.110 ⇒ 00:34:31.129 kaybay: Enough of the executive buy-in to then sort of try to push through the contracting process as quickly as possible.
272 00:34:31.130 ⇒ 00:34:50.170 kaybay: And then kind of targeting, like, a mid-October, ideally mid to late October sort of starting time, just in terms of, like, resource management on your end. But I think, yeah, the team that you’ve described makes sense. I think the focus on the identity resolution stuff makes sense. Just really excited to start working with you guys.
273 00:34:50.489 ⇒ 00:35:06.119 Uttam Kumaran: Yeah, great. No, me too. I guess, what would be helpful before that meeting? I mean, again, like, we can go as far as you want on, like, getting beefier project plans, but probably the next piece is just, like, you guys have a sense of, like, the modeling and stuff like that. This is… this is most of, like.
274 00:35:06.119 ⇒ 00:35:16.969 Uttam Kumaran: well, when we come in and we do discovery, that’s… all of our questions would be there, but do you need anything before Wednesday to… to help with your pitch, or how can I, you know, enable that?
275 00:35:17.470 ⇒ 00:35:31.860 kaybay: I think all I would need to know is just, I’m assuming time and materials kind of pricing, and then just what the hourly would be for you, for the engineer, and for the PM, if it’s not the same rate. I was actually surprised, AWS is now charging the same across the board. I thought they.
276 00:35:31.860 ⇒ 00:35:45.309 Uttam Kumaran: Yeah, I mean, this is a… this is, again, this is a big thing in consulting where, like, I… because I’m not a consultant by trade, I tend to say the quiet thing in the room, which is, like, hourly pricing is, like, kind of dying, because…
277 00:35:45.320 ⇒ 00:35:55.800 Uttam Kumaran: like, we use AI to help speed things up all the time, and so on my side, I’m like, why would we charge hourly when it’s getting faster? So for us, for a lot of our clients.
278 00:35:55.900 ⇒ 00:36:10.500 Uttam Kumaran: I’ve challenged that, and, like, I try to both do outcome-based and hourly. Some people are… of course, some orgs are used to procuring everything TMM, and so for them, it’s… it’s going to be that way until it isn’t. But, I don’t know, I feel like…
279 00:36:10.660 ⇒ 00:36:22.340 Uttam Kumaran: as a… as the client, I think it’s helpful for you guys to really focus on milestone-based outcomes. I mean, of course, I believe in our team and that we’re gonna hit it, and… but I also, like.
280 00:36:22.760 ⇒ 00:36:30.000 Uttam Kumaran: I’m new to this world of running a consulting company, and I don’t know, it’s always been weird, because our optimization typically is, like.
281 00:36:30.050 ⇒ 00:36:41.190 Uttam Kumaran: just increase rates and lower cost, but we’re using AI for a lot of stuff, and the client should benefit from that. And so, for us, we’re moving towards as much of, like.
282 00:36:41.190 ⇒ 00:36:54.979 Uttam Kumaran: if we can hit a milestone at an appropriate time, something more outcome-based. But again, we’re new, so I can do those things. Like, EY, Deloitte, most of your big professional services, they’re having a big crisis moment right now, because
283 00:36:55.220 ⇒ 00:37:00.340 Uttam Kumaran: And those companies are almost, like, pseudo-government. They have all those people that are based on pension.
284 00:37:00.400 ⇒ 00:37:02.879 kaybay: Yeah. And they have all the junior people that are, like.
285 00:37:02.880 ⇒ 00:37:05.390 Uttam Kumaran: Billing for them, and
286 00:37:05.820 ⇒ 00:37:11.859 Uttam Kumaran: they are… they cannot, like… they don’t have much room for creativity in this world, you know? So…
287 00:37:12.600 ⇒ 00:37:15.439 kaybay: Yeah, I love all of what you’ve just said.
288 00:37:15.440 ⇒ 00:37:15.980 Uttam Kumaran: That’s strange.
289 00:37:15.980 ⇒ 00:37:25.120 kaybay: I had a conversation with somebody else, last week, or the one before, about the outcome-based thing, and, like, it’s… I mean, it… yeah, it does make… it’s the only thing that really makes sense.
290 00:37:25.120 ⇒ 00:37:40.519 Uttam Kumaran: It’s pure, it’s very secure. I think with data, though, you know, the challenge is some of the stuff enables a lot, right? Like, foundational snowflake reg… like, how can you put a price on that, in that it just, like, powers a lot of stuff? So more, it’s more of, like.
291 00:37:40.600 ⇒ 00:37:51.310 Uttam Kumaran: the more… the easier the scope we can agree on, or maybe it’s even a mix of both, right? Like, there is something that’s more hourly, and there’s something that kicks on an outcome.
292 00:37:51.350 ⇒ 00:37:59.749 Uttam Kumaran: For a lot of our clients sometimes, too, we’re just like, hey, if we’re handling, like, 2 or 3 work streams, and you’re roughly okay with the pace, then it’s just gonna be a monthly fee.
293 00:37:59.750 ⇒ 00:38:12.589 Uttam Kumaran: And that way, my team can ladder up and down, but then it’s also very… there’s no surprises, versus every time we have to get a change order, when it’s, like, just 10 more hours or something, I push my team to get it done.
294 00:38:12.690 ⇒ 00:38:28.909 Uttam Kumaran: And that’s how we work, is like, even if we don’t make the most money, I want to make sure we get it done, because we’re here to build longer-term relationships, versus some places, oh, you need a change order for this, or stuff like that, but it’s all kind of getting figured out in consulting. It’s pretty interesting right now.
295 00:38:28.910 ⇒ 00:38:31.670 Kyle Wandel: My first ever job was management consulting, so I.
296 00:38:31.670 ⇒ 00:38:46.329 Uttam Kumaran: Oh, great, okay. I hear the expert, I don’t… like, that’s where I’m just learning about… it’s… I get… I get it, like, I’m now on that boat, I get why it happened, but the AI, I think, is really changing my mind to how stuff should work.
297 00:38:46.530 ⇒ 00:38:59.260 Kyle Wandel: Well, personally, I think it benefits both parties. I mean, because there were many, many times during my contract when we were told not to do something anymore because we were reaching that hourly cap limit, and so our contractor got screwed over because of it, so…
298 00:38:59.260 ⇒ 00:38:59.830 Uttam Kumaran: Yeah.
299 00:39:00.890 ⇒ 00:39:04.639 kaybay: Yeah, I think… so maybe, interestingly, to that end.
300 00:39:04.640 ⇒ 00:39:05.520 Kyle Wandel: Might be worth…
301 00:39:05.520 ⇒ 00:39:13.370 kaybay: the real trick, I think, for… just, you know, to be honest, right, is, like, I… I’m not confident we could really define
302 00:39:13.720 ⇒ 00:39:28.579 kaybay: outcomes super reasonably just yet, but maybe we could take a hybrid approach? So I wonder if we had, like, an MSA, and maybe, like, an initial itty-bitty scope around the discovery work to define the outcomes?
303 00:39:28.580 ⇒ 00:39:43.569 kaybay: And then putting in something that would say, like, okay, now we have the outcomes, these are the things we want to work towards, and trying to move that through really quickly, because we’d have the MSA already approved, we would just need to get the change order signed off on. I know we just talked about the pain of change orders, but, like, one change.
304 00:39:43.570 ⇒ 00:39:50.720 Uttam Kumaran: No, but it’s also helpful, like, look, I think even just to get a sense of our pace and working with us, and then move towards, like, something that’s more fixed.
305 00:39:50.720 ⇒ 00:39:51.090 kaybay: Yeah.
306 00:39:51.090 ⇒ 00:40:06.260 Uttam Kumaran: It’ll be better. But it’s all going to… I think that just, for me, I… I don’t like getting put in that position, Kyle, that you mentioned, which is, like, we’re hitting hourly caps, and instead, I’m like, I know it’ll balance out, and if we’re egregiously, like, signing up for too much.
307 00:40:06.260 ⇒ 00:40:19.140 Uttam Kumaran: We didn’t have a conversation about it, but, like, I don’t know, I think there’s something more fair there. And you’re right, for me, the AI piece is actually, we deliver way faster, and I want our client to benefit, and that’s what makes us competitive.
308 00:40:19.180 ⇒ 00:40:38.029 Uttam Kumaran: Versus, like, your super slow, big consulting company. And I, like, yes, you could be very greedy and be like, well, we want to make all of that, but I actually think there’s, like, that is our competitive advantage, and that’s why clients work with us, is that we are very quick, and that we use AI to advantage, and clients benefit from some of that, we benefit from some of that.
309 00:40:38.090 ⇒ 00:40:54.519 Uttam Kumaran: So yeah, I don’t know, maybe we can talk about it. Worst case, it ends up the same way, but, I’m, like, I’m enjoying working with clients that are more flexible, especially because, like, that’s the direction we want to head. It’s something that is… we’re hitting outcomes. But yeah, some of this stuff is hard to…
310 00:40:54.650 ⇒ 00:41:02.409 Uttam Kumaran: we don’t… we kind of, like, don’t know until you open the chest up. And you guys are seeing… I’m sure you guys are seeing that too, so…
311 00:41:02.700 ⇒ 00:41:04.220 Uttam Kumaran: Right.
312 00:41:04.220 ⇒ 00:41:10.059 kaybay: Right, like, I feel like I could sit here and say, like, outcome is, you know, identity resolution, XYZ goal.
313 00:41:10.390 ⇒ 00:41:15.440 kaybay: Once we get in there, it’s gonna be like, you know, oh no, this is no longer possible, right?
314 00:41:15.440 ⇒ 00:41:16.929 Uttam Kumaran: Yeah, yeah, so…
315 00:41:16.930 ⇒ 00:41:22.140 Kyle Wandel: To that testament, I will say that our CEO is much more willing to
316 00:41:22.140 ⇒ 00:41:37.409 Kyle Wandel: enjoy the investigative process, like, okay that, other than being able to oversell yourself. So, I definitely think I’m in agreement with Captain, like, maybe an initial scope of just discovery to see where we’re at, and then maybe in 2026 or something, push forward to actual working stuff, but…
317 00:41:37.410 ⇒ 00:41:39.280 Kyle Wandel: I would agree.
318 00:41:39.750 ⇒ 00:41:42.299 Uttam Kumaran: Cool. Okay, I would love to explore that, yeah.
319 00:41:43.710 ⇒ 00:41:48.100 kaybay: How long… I mean, I know the answer is obviously also, it depends, but, like…
320 00:41:48.430 ⇒ 00:41:53.760 kaybay: what sort of timeline do you think sounds reasonable for the discovery piece? Like.
321 00:41:53.760 ⇒ 00:41:55.059 Uttam Kumaran: Yeah, so…
322 00:41:55.060 ⇒ 00:41:56.070 kaybay: Couple.
323 00:41:56.070 ⇒ 00:42:09.119 Uttam Kumaran: to give you a sense of kind of how long things take for, like, for miles, for things that are pretty repeatable. So, one, we do a lot of, like, basic dbt infrastructure setup, right? So that is, if you guys have GitHub or Bitbucket or whatever, setting up
324 00:42:09.920 ⇒ 00:42:28.569 Uttam Kumaran: control setting up, like, the original project. We have a lot of, like, reusable templates that we use to set up, like, core Martin projects. That’s making sure we can run our first model, and then we also have, like, a staging process where you can… there’s a CI-CD process, basically. That’s usually, like, two to four weeks.
325 00:42:28.570 ⇒ 00:42:34.029 Uttam Kumaran: Okay. Like, that doesn’t require… the only thing… the only blockers there typically are, like, access.
326 00:42:34.150 ⇒ 00:42:42.389 Uttam Kumaran: And then, like, do what we usually do there. I would say from the point after that is, one, is just, like.
327 00:42:42.590 ⇒ 00:42:52.529 Uttam Kumaran: the predictability of the source data that we’re getting in, and then it’s how fast can we iterate on the requirements for the core modeling. Some situations.
328 00:42:52.530 ⇒ 00:43:07.490 Uttam Kumaran: it’s like, we… they’re like, you decide what the data model’s gonna be like, and then we get there, and then it’s like, oh, we’re missing this column, we’re missing this column. Sometimes it’s like, we already… I already have it, perfect ERD for you, just build towards that, right? And that’s… that’s a… that’s a difference there.
329 00:43:07.650 ⇒ 00:43:13.410 Uttam Kumaran: And then the other thing is, like, there… it seems like if Snowflake is gonna be new, there’s gonna be some Snowflake…
330 00:43:13.430 ⇒ 00:43:23.650 Uttam Kumaran: like, role-based access control, like, how we’re gonna do naming conventions and things like that. Again, we have templates, and we have a typical way we do that.
331 00:43:23.650 ⇒ 00:43:35.719 Uttam Kumaran: But that’s usually, again, like, one to two weeks of work of making sure, cool, we have users, they have roles, we can then power… we can test Power BI on top of it, we can test dbt on top of it, and test ETL into it.
332 00:43:35.770 ⇒ 00:43:39.660 Uttam Kumaran: You know, so that’s usually… and then for discovery.
333 00:43:39.830 ⇒ 00:43:45.470 Uttam Kumaran: that’s where I think it’s some… it’s a little bit more like it depends, like, it depends on the amount of people involved, like…
334 00:43:45.600 ⇒ 00:43:57.670 Uttam Kumaran: if you’re, like, okay, there’s, like, there’s, like, one or two key people that are the gatekeepers, everything, okay, then it’s just, like, meetings with them. If you’re, like, it’s them, but maybe there’s other people, and, like, IT’s typically slow to get access to stuff.
335 00:43:57.910 ⇒ 00:44:09.259 Uttam Kumaran: it could totally sort of go higher. It’s also… it’s… it’s also, like, I don’t like to… I’m not someone who’s, like, get everything perfectly until we execute, like…
336 00:44:09.290 ⇒ 00:44:27.779 Uttam Kumaran: Right. Some of it we will figure out, because also my job is that that person has data problems, so I want to solve whatever they need while we’re learning about them, and builds a lot of buy-in and confidence. Yep. So, part of it is, like, we’ll just kind of get a sense from y’all on, like, how difficult it is. Like,
337 00:44:27.860 ⇒ 00:44:33.739 Uttam Kumaran: Are there systems, all of them have APIs, that we can get data from them within a week time and start to, like.
338 00:44:33.990 ⇒ 00:44:40.130 Uttam Kumaran: spot check stuff? Or is there… is there quite a lot more work to do?
339 00:44:40.480 ⇒ 00:44:48.980 Uttam Kumaran: You know, but again, we want to end up with a plan, right? So, typically, it’s like a… we end up with a… with a roadmap, a pretty sophisticated roadmap with
340 00:44:48.980 ⇒ 00:45:00.879 Uttam Kumaran: list of the open questions that we maybe didn’t get to, or that we would have to get to, and so at least, at that point, you could say, like, this plan is robust enough for us to go at it, or it’s like, I think we should do a couple more weeks of discovery.
341 00:45:00.920 ⇒ 00:45:20.750 Uttam Kumaran: You know, and so we’re constantly writing that plan, pulling together platform documentation as we do the discovery. So it’s not a, okay, we’ll talk to you in, like, 6 weeks thing. And we also try to build stuff, like, during that moment, right? While we have someone on the hook, they need something quick, it’d be nice to build them something in our new environment, and so that’s, like, our typical process.
342 00:45:21.370 ⇒ 00:45:30.300 kaybay: Yeah, I mean, it’s also very much what I’ve been doing, is, like, I definitely don’t have all the things yet, but I have enough to give you better data, and so, like, the one sort of small…
343 00:45:30.440 ⇒ 00:45:43.700 kaybay: ground I have gained is with the marketing team. Instead of this horrendous, like, mishmash of lists and groups and data extensions in Marketing Cloud, now I have the, like, appropriate sort of sync between my faux data warehouse.
344 00:45:43.700 ⇒ 00:45:56.380 kaybay: all of the logic that needs to go into it, and then all the, like, downstream stuff. And then this morning, I was so excited, because my counterpart on the marketing team was like, but it occurred to me that I think I can self-serve out of this, and I was like, yes, yes, yes, yes!
345 00:45:57.620 ⇒ 00:45:58.360 kaybay: So I’m like.
346 00:45:58.360 ⇒ 00:46:07.520 Uttam Kumaran: Also, like, if we can ship even a crappy version of Identity Resolution, and people ask questions, we’re like, it’s missing this. I’m like, yes. Yeah, yeah. So that’s…
347 00:46:07.690 ⇒ 00:46:13.629 Uttam Kumaran: I don’t need to… we don’t need to get it right, we just need to get people interested in it, and then they’ll unblock.
348 00:46:13.770 ⇒ 00:46:18.090 Uttam Kumaran: Oh, because we didn’t get access, oh, like, let me call so-and-so, right? So that’s…
349 00:46:18.090 ⇒ 00:46:24.769 kaybay: Yeah. That’s how it works, so… Yeah, no, 100%. I love my red pen, folks. Yeah.
350 00:46:25.390 ⇒ 00:46:28.110 kaybay: Okay, I think…
351 00:46:30.500 ⇒ 00:46:32.870 kaybay: Oh, sorry, I actually don’t know what time it is and how close we are to.
352 00:46:32.870 ⇒ 00:46:34.980 Uttam Kumaran: Dude, 145, yeah.
353 00:46:34.980 ⇒ 00:46:37.199 kaybay: Okay, we go to 3 or 2?
354 00:46:37.200 ⇒ 00:46:38.859 Uttam Kumaran: Yes, we have 15 more minutes.
355 00:46:38.860 ⇒ 00:46:46.880 kaybay: Yeah, yeah, okay. Okay, so I think maybe we could, to the hybrid extent, If we take
356 00:46:47.020 ⇒ 00:47:05.979 kaybay: the outcome sort of approach, and we say the snowflake thing, right, get it baseline set up, ready to go, dbt, same thing, those two pieces, I think, could be outcome-based, because, yeah, I think saying, like, have it be able to do things is an outcome that I can define. And then the identity resolution piece.
357 00:47:07.450 ⇒ 00:47:11.459 kaybay: Is a little bit more slippery, but maybe if we scope…
358 00:47:11.460 ⇒ 00:47:31.619 kaybay: discovery with the outcome of the plan for identity resolution sounds a little silly, but, like, I think that might be a way to do it. And I do think, exactly to your point, I think we can move very quickly on delivering a far better state of the data with some identity resolution, like, you know, quick and dirty stuff.
359 00:47:31.620 ⇒ 00:47:36.860 kaybay: Even if the, you know, ultimate golden dataset is more of a, you know, longer project.
360 00:47:36.860 ⇒ 00:47:37.780 Uttam Kumaran: Yes.
361 00:47:37.820 ⇒ 00:47:52.059 kaybay: I think right now, you know, to now and the end of the year, like, the big focus is on all the CES stuff, and so I think there’s a logical and excellent use case there to get started with. It’ll be sort of ironic that we’ll probably finish the work, or that phase of the work.
362 00:47:52.060 ⇒ 00:48:07.049 kaybay: in time for CES to be done, we won’t really need it until next year, but I think the things that we would build to support the CES process currently would translate easily to some of the other data streams that we’re needing to bring in, because it’s already sort of bringing in a bunch of the data.
363 00:48:07.050 ⇒ 00:48:12.279 Uttam Kumaran: That we want to work with. To your point about the API, or your question about the APIs and stuff.
364 00:48:12.870 ⇒ 00:48:29.180 kaybay: I think we’re looking at flat files and FTP, for the most part, for now. The vendors that we work with do have APIs, but as far as I understand, their, like, rate limits are pretty draconian because their systems are not terribly resilient for an event of our size.
365 00:48:29.180 ⇒ 00:48:40.800 kaybay: And so there’s a lot of hesitancy around, like, granting additional API access, even for something like me, because it’s just disrupting, you know, the delicate house of cards that they’ve already got kind of built.
366 00:48:40.800 ⇒ 00:48:45.139 Uttam Kumaran: If they can land everything in S3, you can do a storage integration into S3, and…
367 00:48:45.160 ⇒ 00:48:46.020 kaybay: Yeah.
368 00:48:46.020 ⇒ 00:48:53.120 Uttam Kumaran: build snowpipe on it and bring it in, like, it’s… that’s fine. Again, the biggest thing is just, like, that schema doesn’t change, and that…
369 00:48:53.300 ⇒ 00:48:53.930 kaybay: Right.
370 00:48:54.230 ⇒ 00:48:56.849 Uttam Kumaran: Come on a regular cadence, you know, so…
371 00:48:57.020 ⇒ 00:49:01.569 Uttam Kumaran: We’ve seen it all on my side, so it’s just like, as long as we can get something from them.
372 00:49:01.780 ⇒ 00:49:02.910 Uttam Kumaran: That’s great.
373 00:49:03.010 ⇒ 00:49:03.569 Uttam Kumaran: To stop there.
374 00:49:03.570 ⇒ 00:49:11.189 kaybay: Exactly, exactly, exactly. We can start nimbly. So yeah, I think that all sounds good to me. Okay.
375 00:49:11.790 ⇒ 00:49:12.470 kaybay: Okay?
376 00:49:12.470 ⇒ 00:49:28.460 Uttam Kumaran: Yeah, I feel… I feel that’s… I feel pretty fair. I think the Snowflake piece and the dbt initialization piece are both, fine to kind of scope as is. I agree in that, like, I… I think if the timelines roughly, like, end with a plan towards
377 00:49:28.560 ⇒ 00:49:33.769 Uttam Kumaran: identity and enrichment in 2026 is, like, a great outcome for the initial scope, like, that’s…
378 00:49:34.050 ⇒ 00:49:47.490 Uttam Kumaran: that’s that. I think in that process, like, we’re not saying we won’t, like, build small things at the same time. Yeah. So that’s… I don’t… I wouldn’t… I would… I’ll kind of bake that in, in that, like, if there are quick wins or things for us to do. The other thing is, like.
379 00:49:47.530 ⇒ 00:49:54.959 Uttam Kumaran: you know, there is part of data which is, like, as more people start to use stuff, more questions come up. And it is, like, a positive thing, so…
380 00:49:54.960 ⇒ 00:50:09.320 Uttam Kumaran: one of the things that we’ve found in companies that we’ve gone into is that, like, as we right-size things, just more users pop up. And so I kind of want to make sure that we are flexible, and that, like, yes, there is some… some outcome, but there’s also, like, you have time from us, like.
381 00:50:09.630 ⇒ 00:50:24.149 Uttam Kumaran: okay, we need just extra hours to continue to build models for a new team that wants to show us some love, right? And so, that’s something I think is helpful. And all the clients, as soon as we start to get things going, everybody comes out of the woodwork to ask for data.
382 00:50:24.150 ⇒ 00:50:32.650 Uttam Kumaran: You know, and ideally, hopefully, that is partly us building, but then it’s enabling people on Power BI to go grab it themselves, or whatever the BI tool is, you know.
383 00:50:32.950 ⇒ 00:50:38.310 kaybay: Yeah, yeah, exactly. Definitely already, seeing that, and it’s a good, it’s a good problem to have.
384 00:50:38.310 ⇒ 00:50:38.660 Uttam Kumaran: Great.
385 00:50:39.030 ⇒ 00:50:39.580 kaybay: I do love.
386 00:50:39.580 ⇒ 00:50:40.350 Uttam Kumaran: What the fuck’s second.
387 00:50:40.350 ⇒ 00:50:41.319 kaybay: That would work.
388 00:50:41.320 ⇒ 00:50:47.330 Uttam Kumaran: you’re like, oh, now I’m just, like, handling, like, day-to-day, like, fixes or whatever, so, yeah.
389 00:50:47.620 ⇒ 00:50:49.790 Kyle Wandel: Yeah, yeah, oh my god.
390 00:50:49.790 ⇒ 00:50:59.900 kaybay: There is that interesting, like, conflict of desire where I’m like, I’d really like to solve tech debt today, but actually I’m probably just gonna upload 4 spreadsheets to an FTP server and then go home and cry a little bit.
391 00:50:59.900 ⇒ 00:51:07.950 Uttam Kumaran: Yeah, or there’s a day… there’s, like, weeks where you’re like, all I did was tech that I didn’t get anything new done, and you’re like, you’re like, but I fixed, like, 10 jobs, and, like, I mean.
392 00:51:07.950 ⇒ 00:51:08.529 kaybay: Thank you.
393 00:51:08.530 ⇒ 00:51:14.880 Uttam Kumaran: It’s… but, yeah, this is the challenge. Like, we try to… even in some of our teams, it’s, like, 40% ad hoc work.
394 00:51:14.880 ⇒ 00:51:16.530 kaybay: Yeah. I just have to tell our…
395 00:51:16.530 ⇒ 00:51:19.809 Uttam Kumaran: PM team, like, that’s okay in beta. Like, that is not, like, a…
396 00:51:19.980 ⇒ 00:51:27.040 Uttam Kumaran: bad thing. That doesn’t mean we actually even play in poorly. Data teams are a service organization, in my mind, a service organization.
397 00:51:27.040 ⇒ 00:51:31.489 kaybay: And so there’s gonna be… usually we budget at least 20% of our time to, like.
398 00:51:31.660 ⇒ 00:51:38.889 Uttam Kumaran: my report failed, or, like, I need this new data source, or whatever, and we’re, like, we were okay with… with that, you know?
399 00:51:39.210 ⇒ 00:51:42.330 kaybay: Yeah, yeah, totally agree. Totally agree.
400 00:51:42.330 ⇒ 00:52:01.129 Kyle Wandel: And I will say that the appetite’s here. I mean, like Captain said, not only are the individual department leads kind of, like, starting to take initiative and kind of understanding that we have access to the data, but also executives are really, kind of, their appetite’s pretty high as well. It’s a little bit of changing of the guard, so it’s a really good opportunity for really everybody right now.
401 00:52:01.130 ⇒ 00:52:01.750 Uttam Kumaran: Yeah.
402 00:52:02.890 ⇒ 00:52:07.620 Uttam Kumaran: Okay, so let me try to get you something, Catherine, just, like, tomorrow, I’ll just put something together.
403 00:52:07.620 ⇒ 00:52:07.990 kaybay: Yeah.
404 00:52:07.990 ⇒ 00:52:13.659 Uttam Kumaran: If that works, and then I’ll look forward to hearing how that meeting goes, and then we can go from there.
405 00:52:13.950 ⇒ 00:52:32.920 kaybay: Okay, that sounds perfect. And yeah, I would say anything you can get me tomorrow is great. Don’t worry about, like, it being super polished, because I think as long as I have some numbers and stuff I can speak to in the meeting, it’s Wednesday afternoon, actually, so even Wednesday morning is fine. As long as I have stuff that I can speak to to give them a sense of the number and a sense of the story, I’ll be in good shape.
406 00:52:32.920 ⇒ 00:52:34.570 Uttam Kumaran: Yeah. Okay, perfect.
407 00:52:34.960 ⇒ 00:52:37.730 kaybay: Okay, yeah, I’m excited.
408 00:52:37.730 ⇒ 00:52:41.339 Uttam Kumaran: Alright, great. I appreciate it. Thank you both, and then, yeah, we’ll be in touch.
409 00:52:42.020 ⇒ 00:52:43.020 kaybay: Thank you so much.
410 00:52:43.020 ⇒ 00:52:43.900 Uttam Kumaran: Alright, well, thanks.
411 00:52:43.900 ⇒ 00:52:45.530 Kyle Wandel: End of life, I guess.