Meeting Title: Brainforge x CTA: Weekly! Date: 2025-12-19 Meeting participants: Katherine Bayless, Kyle Wandel, Uttam Kumaran, Ashwini Sharma
WEBVTT
1 00:01:06.920 ⇒ 00:01:08.629 Uttam Kumaran: Hey everyone, good morning.
2 00:01:08.630 ⇒ 00:01:10.519 Katherine Bayless: Hey, good morning, how’s it going?
3 00:01:10.520 ⇒ 00:01:12.219 Uttam Kumaran: Good! How’s everything?
4 00:01:13.280 ⇒ 00:01:15.819 Katherine Bayless: Going? It’s going. Going.
5 00:01:16.370 ⇒ 00:01:17.700 Katherine Bayless: Tom, are you in Texas?
6 00:01:18.060 ⇒ 00:01:18.730 Uttam Kumaran: Yes.
7 00:01:18.730 ⇒ 00:01:21.039 Katherine Bayless: I have beef with your state.
8 00:01:21.040 ⇒ 00:01:26.050 Uttam Kumaran: Okay, please, yeah, I’m the representative, I can make anything happen, so…
9 00:01:26.050 ⇒ 00:01:27.059 Katherine Bayless: The mayor of Texas.
10 00:01:28.440 ⇒ 00:01:44.590 Katherine Bayless: Yeah, no, actually, I mean, you guys maybe are aware too, but, I was totally thrown a curveball yesterday afternoon. We were on the integrations call for all the CES stuff, and the mobile app vendor was like, anybody who wants to stick around and talk about the new Texas law, and I’m like, I’m sorry, what?
11 00:01:44.590 ⇒ 00:01:51.760 Katherine Bayless: And I guess the age verification stuff is going into effect January 1st, Android.
12 00:01:51.760 ⇒ 00:01:54.110 Uttam Kumaran: For… age verification for what?
13 00:01:54.340 ⇒ 00:02:13.500 Katherine Bayless: for… sorry, yeah, yeah, sorry, I skipped ahead. App Store downloads. So, like, our mobile app for CES will only work for people who are either 18 or older or have parental consent in the state of Texas, which sounds fine, except that Android is supporting it fully, and Apple is only supporting it on the latest version of their iOS.
14 00:02:13.650 ⇒ 00:02:15.120 Uttam Kumaran: Oh.
15 00:02:15.120 ⇒ 00:02:18.659 Katherine Bayless: So we have to decide if we want to comply.
16 00:02:18.950 ⇒ 00:02:28.679 Katherine Bayless: Or tell all of our attendees to download the latest version of iOS, or else they can’t use our mobile app. And so I have a call with a whole bunch of lawyers later today.
17 00:02:28.680 ⇒ 00:02:30.579 Uttam Kumaran: No. Oh, God.
18 00:02:30.580 ⇒ 00:02:31.190 Katherine Bayless: Great.
19 00:02:31.490 ⇒ 00:02:32.810 Uttam Kumaran: I was like…
20 00:02:32.810 ⇒ 00:02:35.620 Katherine Bayless: Oh, didn’t see that coming. Didn’t have that on my bingo card for this.
21 00:02:35.620 ⇒ 00:02:36.910 Uttam Kumaran: It’s like, literally.
22 00:02:36.910 ⇒ 00:02:37.520 Kyle Wandel: random.
23 00:02:37.520 ⇒ 00:02:38.240 Uttam Kumaran: bang.
24 00:02:38.240 ⇒ 00:02:42.959 Kyle Wandel: Completely random thought, Catherine. Is there any way we can, entice Apple to,
25 00:02:43.560 ⇒ 00:02:45.899 Kyle Wandel: Make it go compat- backwards compatible.
26 00:02:45.900 ⇒ 00:02:56.040 Katherine Bayless: I mean, right? Like, it is funny that, like, we are the organization that, like, that’s not the most ridiculous question. Like, Christina was like, just have Gary write an op-ed and get it killed.
27 00:02:56.040 ⇒ 00:03:04.199 Kyle Wandel: I was about to say, can we have Gary reach out to, Tim Cook, or offer them free exhibiting space, or free, front stage access? So, I don’t know.
28 00:03:04.200 ⇒ 00:03:07.670 Katherine Bayless: Right? Yeah, exactly, exactly. I’m like, we know some people.
29 00:03:08.180 ⇒ 00:03:10.050 Uttam Kumaran: Sorry, guys, that’s brutal.
30 00:03:10.650 ⇒ 00:03:11.540 Katherine Bayless: Yeah, so…
31 00:03:12.330 ⇒ 00:03:21.619 Katherine Bayless: I was like, honestly, as it sort of, like, started to sink in, I was like, well, actually, this is a problem that’s so far above my pay grade that all I need to do is listen to the action items.
32 00:03:22.310 ⇒ 00:03:23.860 Katherine Bayless: Yeah.
33 00:03:24.000 ⇒ 00:03:24.740 Katherine Bayless: Oop, yeah.
34 00:03:25.130 ⇒ 00:03:26.949 Uttam Kumaran: That’s so funny. Wow.
35 00:03:26.950 ⇒ 00:03:28.839 Katherine Bayless: so yeah.
36 00:03:28.840 ⇒ 00:03:31.640 Uttam Kumaran: How’s the… how’s the week been otherwise?
37 00:03:32.260 ⇒ 00:03:36.580 Katherine Bayless: Pretty good. What would you say, Kyle? I think it’s been a decent week, actually.
38 00:03:36.850 ⇒ 00:03:53.779 Kyle Wandel: Yeah, mostly. I’m mostly wrapped up with all of my, other work that I’ve had to do, so now I can focus more on data ops stuff, and so, honestly, like, I think the… having this on Friday will be… or having also the Snowflake training on Friday will be really nice, because it’ll give us thoughts and more excitement going forward. Yeah.
39 00:03:54.380 ⇒ 00:03:55.040 Katherine Bayless: Yeah.
40 00:03:56.090 ⇒ 00:04:04.919 Uttam Kumaran: Cool. I think so. Today, we have a bit of a deck prepared, so just kind of highlight, you know, progress so far. I think part of the…
41 00:04:05.200 ⇒ 00:04:10.109 Uttam Kumaran: we’re not, like, when we started, you know, the company, I’m not really big on…
42 00:04:10.220 ⇒ 00:04:21.759 Uttam Kumaran: I actually was, like, really bash… bashing decks, because I think a lot of consultants sort of just, like, oh, we’re gonna put a big deck together and not do other work. So we’ve… we’ve sort of come at it where
43 00:04:21.760 ⇒ 00:04:36.239 Uttam Kumaran: later, now we’re starting to do… it’s just basically a good way to summarize, but also for… for the CTA team, it’s a good way to just screenshot and be like, here’s what we got done. And so it’s a nice codification of, like.
44 00:04:36.390 ⇒ 00:04:44.209 Uttam Kumaran: what the things we’re working on, so that’s what, I can go through, and we’re super open to, you know, any,
45 00:04:44.650 ⇒ 00:04:54.629 Uttam Kumaran: any feedback, so let me get this going, and then I’ll probably… Ashwini, I can drive and let you, you know, comment as we go into things.
46 00:04:59.800 ⇒ 00:05:00.780 Uttam Kumaran: Great.
47 00:05:16.000 ⇒ 00:05:16.830 Uttam Kumaran: Cold.
48 00:05:17.630 ⇒ 00:05:36.639 Uttam Kumaran: So this is a slide that we’ve started to just put at the beginning of all of our decks, which is just, like, what are we here for, and constantly hammering, like, what our, like, North Stars are. I think, you know, especially given the amount of opportunities that exist within CTA, I think it’s helpful
49 00:05:36.730 ⇒ 00:05:48.629 Uttam Kumaran: to see this every time, and remind us as, like, data team, like, what we’re here to do. So, I mean, I don’t think any of this will be surprising to y’all, but for our team, I think
50 00:05:49.160 ⇒ 00:05:52.729 Uttam Kumaran: You know, it’s sort of good to see
51 00:05:52.740 ⇒ 00:06:08.019 Uttam Kumaran: This, and when we’re making decisions, you know, within data models, or training, or new opportunities, like, our goal is not… our goal is to, you know, help the team, like, enable self-serve analytics, build a great foundation.
52 00:06:08.020 ⇒ 00:06:13.999 Uttam Kumaran: And then in particular, I like 3 being super, super clear, which I think, Catherine, you’ve…
53 00:06:14.000 ⇒ 00:06:26.410 Uttam Kumaran: you recognize that, and I think you made sure that, you know, we’re aware, which is, like, how do we drive the member experience, and then how do we sort of take on or reduce the manual overhead, so…
54 00:06:28.190 ⇒ 00:06:44.569 Uttam Kumaran: this, week, you know, I think we… we did, you know, got a lot of progress done on the technical side. So dbt set up, our Snowflake, you know, environment is… is in a great place. The active members report, like.
55 00:06:44.630 ⇒ 00:06:51.529 Uttam Kumaran: you know, V0 is there, ready to review. You know, we’re in… and I think this…
56 00:06:51.810 ⇒ 00:06:58.260 Uttam Kumaran: this week, we’ll… today, we’re going to kind of talk through what each of these items kind of looks like in Snowflake, and then…
57 00:06:58.920 ⇒ 00:07:17.289 Uttam Kumaran: run, you know, the training with Kyle and start to get more people into the environment. Kind of next is just the ingestion of sources, data modeling for the customers, Martin, then continuing to build up our, like, documentation. I think one thing that we are going to be working on, especially, like.
58 00:07:17.500 ⇒ 00:07:36.989 Uttam Kumaran: kind of… until… kind of into the next two weeks and, like, early Jan is just, like, starting to get track of all of the metrics that we’ve heard from, you know, all the different… from the folks and the conversations that we’ve had. And so when we go meet other stakeholders in the company, we’ll come prepared with, like, okay, we are aware of these metrics.
59 00:07:37.130 ⇒ 00:07:41.819 Uttam Kumaran: And then making sure that our marts sort of, like, can support all of those.
60 00:07:42.080 ⇒ 00:07:46.470 Uttam Kumaran: Yeah, so I guess… Yeah, go ahead, please.
61 00:07:46.470 ⇒ 00:07:55.629 Katherine Bayless: Just one note on the polyatomic piece. So I had a really good conversation, with him yesterday, and I was… we sort of kind of realized, like, you know, actually the…
62 00:07:55.740 ⇒ 00:08:04.330 Katherine Bayless: prioritization of getting data into Snowflake is a little different for, like, the things we want in Snowflake so we can start building against them, versus, like.
63 00:08:04.810 ⇒ 00:08:18.140 Katherine Bayless: the things that we can consume right away using Polytomic versus need to be built out, and knowing that, like, some of the P0 data sources we want to be able to work with are show data, but the show’s going to be over in 3 weeks.
64 00:08:18.140 ⇒ 00:08:34.360 Katherine Bayless: I was like, it doesn’t really seem to make any sense for you guys to focus on building a connector to a platform we won’t use for another 9 months. So, I think it was probably… it was a good conversation, so we’ll lean on, like, the ones that they have ready to go out of the box, and then, yeah, prioritizing the Marketing Cloud build.
65 00:08:34.360 ⇒ 00:08:44.259 Katherine Bayless: So yeah, it was good. I’m really excited. I think he said they were gonna have the listing up in the marketplace as, like, a private offer, that’s the route they’re gonna take, so they should have it maybe next week.
66 00:08:44.520 ⇒ 00:08:49.770 Uttam Kumaran: Okay, great. Yeah, I mean, I think that’s exactly, like, what we kind of needed here. I mean, it just sort of…
67 00:08:50.130 ⇒ 00:09:03.070 Uttam Kumaran: depends on their… what they… what they can do, and also if it’s going to be valuable for us to have that for Showtime or right after. So, I’m glad that we sort of arrived, you know, but again, I’m not…
68 00:09:03.380 ⇒ 00:09:12.770 Uttam Kumaran: I’m not worried about them in short order hitting all of those, you know, going into January, February, so I… I think they’ll… we’ll just have them kind of work down the list as needed.
69 00:09:13.330 ⇒ 00:09:14.599 Katherine Bayless: Yeah, yeah, yeah.
70 00:09:18.500 ⇒ 00:09:26.729 Uttam Kumaran: Cool, I guess maybe, Ashmini, do we want to just talk about… go through Snowflake, and then maybe… or do you want to talk about the Active Remembers Report?
71 00:09:26.860 ⇒ 00:09:29.399 Uttam Kumaran: Or do you want to… what order do you want to take things?
72 00:09:30.310 ⇒ 00:09:32.610 Ashwini Sharma: Yeah, anything is fine, right?
73 00:09:33.200 ⇒ 00:09:43.809 Ashwini Sharma: Maybe I’ll start with the active members report. So this is created on a dev environment for now. We’ll move it to the prod as the CI pipeline is complete.
74 00:09:44.000 ⇒ 00:10:01.760 Ashwini Sharma: What it does, essentially, is whatever spreadsheet that you had shared, Catherine, earlier, which is created by looking into the remembers data and generating a report, and then verifying it against the system, right? It just scans through the data that is there in Snowflake.
75 00:10:01.920 ⇒ 00:10:08.159 Ashwini Sharma: You know, there’s a couple of joints between different data sets, organization, individuals, the relationship between
76 00:10:08.350 ⇒ 00:10:11.100 Ashwini Sharma: Organizations, and
77 00:10:11.360 ⇒ 00:10:26.750 Ashwini Sharma: getting the primary representative of organization, and then marrying that data with the individual data, and then producing that report, right? So if you query that table that I had shared on Slack, it should give you, on any day, what are the list of active members.
78 00:10:26.860 ⇒ 00:10:33.580 Ashwini Sharma: Right? So basically, the report generation is just a matter of exporting that data into a CSV,
79 00:10:33.720 ⇒ 00:10:43.729 Ashwini Sharma: file, right? But yeah, I think I would love if you could spend, or your team could spend some time verifying the correctness of the report.
80 00:10:43.990 ⇒ 00:10:48.660 Ashwini Sharma: Yeah, that’s… that’s on the active members report.
81 00:10:49.180 ⇒ 00:10:59.649 Ashwini Sharma: like, what, I’ll just take a minute over here to, you know, just deviate from the original topic, right? Sure. Which is,
82 00:10:59.720 ⇒ 00:11:14.869 Ashwini Sharma: I’ll need some work items for next week, right? So, basically, if you have some, you know, some of the KPIs, or some of the reporting needs, or, you know, that depends… right now, we all… all that we have is the Impexium data, right? The remembers data, so…
83 00:11:14.980 ⇒ 00:11:22.240 Ashwini Sharma: of that data set, if you have something that you would like us to work on, just, you know, flag it, and I can take it up.
84 00:11:23.860 ⇒ 00:11:36.410 Ashwini Sharma: The next item that we did was Snowflake setup, and which… this involves creating a bunch of databases for Snowflake, a couple of different roles, and then doing these grant assignments to different roles.
85 00:11:36.550 ⇒ 00:11:50.419 Ashwini Sharma: And creating some service accounts, using which Polytomic is going to add data to Snowflake, and then when we run dbt, it’s going to use one of the service accounts… oh, sorry, I’ll take that back. dbt runs from within Snowflake, so it’s not going to use the service account.
86 00:11:50.520 ⇒ 00:12:02.290 Ashwini Sharma: But let’s say tomorrow you use Sigma or any other BI tool to analyze that data or present dashboards, right? That’s going to happen using a service account that we have created.
87 00:12:03.280 ⇒ 00:12:10.870 Ashwini Sharma: The third thing that we did was dbt setup. This requires, you know, organizing the dbt project in such a way that
88 00:12:11.040 ⇒ 00:12:26.330 Ashwini Sharma: That, you know, as you grow with different data sources, the project scales accordingly, right? You don’t have to make architectural changes with the project, or refactor your models, significantly, wasting developer productivity, right?
89 00:12:27.840 ⇒ 00:12:47.259 Ashwini Sharma: Yeah, so that’s, what has been done, and then the dbt project that we create locally, like, in the training that’s going to happen in a few hours from now, I’ll explain, like, how we do it locally versus how we do it on a Snowflake instance, right? And yeah, so what we have done right now is…
90 00:12:47.280 ⇒ 00:12:56.939 Ashwini Sharma: I’ve used my GitHub personal access token, and then created a workspace within Snowflake, pulled the GitHub repository, and then running dbt commands on top of that.
91 00:12:57.030 ⇒ 00:13:11.849 Ashwini Sharma: But that needs change, and I think I’ll need some kind of a service account from you, which can be utilized to create secret and GitHub API integration, and then we can properly pull the repo into
92 00:13:12.010 ⇒ 00:13:14.760 Ashwini Sharma: Snowflake, and then use it, to run it from there.
93 00:13:15.740 ⇒ 00:13:24.590 Katherine Bayless: Yeah, I need to get that service account, that’s still on me. I do have an idea for the next, effort to undertake, but I can hold that for a parking lot if you want.
94 00:13:26.010 ⇒ 00:13:32.340 Uttam Kumaran: Yeah, we can go… we can go talk about that. I mean, the next few slides… I mean, I don’t know, Sweeney, do you want to just drive through, or we can…
95 00:13:32.740 ⇒ 00:13:35.870 Uttam Kumaran: We, you know, you can also talk about these later, or yeah.
96 00:13:36.150 ⇒ 00:13:51.340 Ashwini Sharma: Right, I can talk about it later also. This is just explaining, like, what was done for these things, right? After the basic one-on-one training is complete, I can walk through our codebase and then explain what are the various things that has been done, and how it should be done going forward.
97 00:13:51.340 ⇒ 00:13:52.290 Uttam Kumaran: Okay.
98 00:13:52.290 ⇒ 00:13:53.220 Ashwini Sharma: Yeah.
99 00:13:53.610 ⇒ 00:14:03.869 Uttam Kumaran: Maybe, yeah, let’s… maybe let’s spend some time talking about active members and then upcoming week, and… and then we can, you know, yeah, that would be great to hear, Catherine.
100 00:14:04.550 ⇒ 00:14:05.560 Katherine Bayless: have ideas.
101 00:14:06.720 ⇒ 00:14:07.050 Uttam Kumaran: Go for it.
102 00:14:07.050 ⇒ 00:14:12.170 Ashwini Sharma: So, active members, if you… do you want me to talk a little bit about this one, or should I pick it up.
103 00:14:12.170 ⇒ 00:14:16.839 Uttam Kumaran: Yeah, if you wanna… even if you wanna pull it up on your side, and we… yeah, let’s talk about the…
104 00:14:17.670 ⇒ 00:14:24.689 Kyle Wandel: Yeah, especially for Kyle, too, we can kind of share with him what we’ve done so far, what’s available in Snowflake. Maybe, Ashwini, if you want to share.
105 00:14:24.710 ⇒ 00:14:29.129 Uttam Kumaran: Where’d it go into Snowflake, and so the team can start running selects and stuff like that.
106 00:14:29.410 ⇒ 00:14:30.010 Ashwini Sharma: Right.
107 00:14:30.010 ⇒ 00:14:40.029 Kyle Wandel: I took a look at the active members report a little bit already, so it’s really, really nice, yeah, yeah. But I think what Captain’s gonna ask you feeds directly off that active members report, so it’ll be… it’ll be… it’s nice that it’s in there.
108 00:14:41.330 ⇒ 00:14:49.970 Ashwini Sharma: Sorry, one second, let me… Too many tabs open in my system, one second.
109 00:14:50.510 ⇒ 00:14:51.510 Ashwini Sharma: I’m losing track.
110 00:14:51.510 ⇒ 00:14:57.440 Katherine Bayless: I found out, because I do, I’ll, like, close my browser at the end of the day, so that I only have one day’s worth of tab chaos, but I.
111 00:14:57.440 ⇒ 00:14:58.470 Uttam Kumaran: Oh, yeah.
112 00:14:58.470 ⇒ 00:15:05.130 Katherine Bayless: developers on the IS team, she does it every hour. She’s like, no, otherwise I start to lose my mind. And I was like, wow, you’re really diligent.
113 00:15:06.040 ⇒ 00:15:14.799 Kyle Wandel: Yeah, I do it every day, but even as I switched over to this team, I’ve noticed, oh my gosh, I’ve noticed that I’ve also generated tabs, but nothing like that.
114 00:15:14.800 ⇒ 00:15:17.630 Uttam Kumaran: I’m more of a… I’m more Windows. I have a lot of Windows.
115 00:15:17.630 ⇒ 00:15:18.180 Katherine Bayless: nose.
116 00:15:18.180 ⇒ 00:15:19.620 Kyle Wandel: Yeah, I can’t do it.
117 00:15:19.620 ⇒ 00:15:22.829 Uttam Kumaran: My Mac Mini finally, like, collapsed yesterday.
118 00:15:22.830 ⇒ 00:15:24.300 Kyle Wandel: Oh, no.
119 00:15:25.630 ⇒ 00:15:27.260 Katherine Bayless: Context window.
120 00:15:27.260 ⇒ 00:15:27.820 Uttam Kumaran: Yeah.
121 00:15:27.820 ⇒ 00:15:30.309 Ashwini Sharma: Alright, okay. So,
122 00:15:31.020 ⇒ 00:15:38.909 Ashwini Sharma: So, what I was going to show is basically this thing, right? So, right now, everything is in depth, nothing is in broad.
123 00:15:39.150 ⇒ 00:15:41.899 Ashwini Sharma: If you’re looking into dev staging, right?
124 00:15:42.010 ⇒ 00:15:47.440 Ashwini Sharma: Now, this one has a bunch of different schema objects. Div staging is a database name.
125 00:15:47.580 ⇒ 00:15:56.180 Ashwini Sharma: Right? These are all high-level database, in case you have not worked with Snowflakes. These are the top-level objects, right? Everything is structured within these.
126 00:15:56.320 ⇒ 00:16:09.439 Ashwini Sharma: And dev staging is one of the objects, database objects, and a database object is going to have multiple schemas within it, right? It can have other objects also, but schemas is one of the primary objects that’s going to be in a database.
127 00:16:09.600 ⇒ 00:16:26.390 Ashwini Sharma: If you look at this information schema, this contains information about other schemas, right? What are the other schemas you have in the table? What are the table? What are the columns? What are the data types? Everything is there in the information schema. The other schemas are actually containers for tables and views and…
128 00:16:26.650 ⇒ 00:16:35.560 Ashwini Sharma: Stuff like that, right? Other DB objects. So the way that I’ve organized it is, right, you know, everything that’s in remembers.
129 00:16:35.850 ⇒ 00:16:44.340 Ashwini Sharma: and within remembers, a functional area, right? So, for example, like, remembers has accounting functional area, so all the tables for that goes into this.
130 00:16:44.830 ⇒ 00:16:45.850 Ashwini Sharma: As a view.
131 00:16:47.510 ⇒ 00:16:48.210 Ashwini Sharma: Right?
132 00:16:48.400 ⇒ 00:16:56.640 Ashwini Sharma: And this is how I’ve renamed it. For example, the accounts table is… Like, SDG remembers accounting accounts.
133 00:16:57.000 ⇒ 00:17:03.900 Ashwini Sharma: Sorry. And then for CRM, it follows the same structure.
134 00:17:06.420 ⇒ 00:17:12.560 Ashwini Sharma: So, in the CRM, like, I’ll go to the raw data because it’s easier to explain it from there onwards.
135 00:17:18.380 ⇒ 00:17:32.329 Ashwini Sharma: So, we have a bunch of tables over here. One of the primary tables is customer, right? And this customer table is sort of a parent table to organization and individual, right? So, you have…
136 00:17:35.160 ⇒ 00:17:42.610 Ashwini Sharma: I’m just going to put names over here, customer… And then we have, individual?
137 00:17:44.220 ⇒ 00:17:47.570 Ashwini Sharma: And also, we have something called organization.
138 00:17:49.780 ⇒ 00:17:50.670 Ashwini Sharma: So…
139 00:17:51.210 ⇒ 00:18:05.240 Ashwini Sharma: they contain additional data. Individual contains data about individual, who is a customer, organization contains data about an organization, which is also a customer, right? And the segregation over here is using a type, which is O and I,
140 00:18:05.420 ⇒ 00:18:11.360 Ashwini Sharma: And and then there is something called membership benefit.
141 00:18:13.380 ⇒ 00:18:24.770 Ashwini Sharma: This indicates what’s the relationship between different organizations, right? So, basically, there could be one organization, and it does not follow a strict relationship of
142 00:18:24.980 ⇒ 00:18:29.180 Ashwini Sharma: one parent, right? So, basically, your organization could have multiple parents.
143 00:18:29.470 ⇒ 00:18:39.039 Ashwini Sharma: In which case, it is indicated by multiple records in this table, and each record is having a different link to a parent organization. So, for example, like…
144 00:18:39.450 ⇒ 00:18:40.780 Ashwini Sharma: This one…
145 00:18:43.940 ⇒ 00:19:00.440 Ashwini Sharma: And here’s one trick about Snowflake, right? This is something that I learned quite later in my life. If you have multiple queries in Snowflake, you can always separate them by semicolons, and then you just run any one, like, Ctrl-Enter or Command-Enter, and it runs only that query.
146 00:19:00.790 ⇒ 00:19:07.640 Ashwini Sharma: And, you know, it’s so much easier, like, when you’re working with BigQuery or Databricks, it does not work this way.
147 00:19:08.530 ⇒ 00:19:09.030 Katherine Bayless: Oh, really?
148 00:19:09.450 ⇒ 00:19:10.180 Katherine Bayless: It’s funny.
149 00:19:10.180 ⇒ 00:19:12.120 Ashwini Sharma: Yeah. Let’s take a look at this data.
150 00:19:12.120 ⇒ 00:19:15.380 Katherine Bayless: I use works that way, so I’ve got… I’m just used to it, but yeah, interesting.
151 00:19:15.380 ⇒ 00:19:16.920 Uttam Kumaran: Same, yeah, same.
152 00:19:19.470 ⇒ 00:19:27.699 Ashwini Sharma: The… if you see this column, right, receive benefits from customer, this indicates what, this organization’s parent is.
153 00:19:27.870 ⇒ 00:19:28.620 Ashwini Sharma: Right?
154 00:19:28.730 ⇒ 00:19:30.230 Katherine Bayless: And.
155 00:19:30.230 ⇒ 00:19:35.820 Ashwini Sharma: This is the thing, and then what we do is we’ll use,
156 00:19:37.590 ⇒ 00:19:39.700 Ashwini Sharma: There is something called Related Contact.
157 00:19:43.270 ⇒ 00:19:51.740 Ashwini Sharma: this will tell you, like, how an organization is related to another organization or another individual, right? And then,
158 00:19:51.960 ⇒ 00:19:59.869 Ashwini Sharma: We also have the individual data over here, right? There is one more table, which is basically customer email.
159 00:20:01.680 ⇒ 00:20:04.899 Ashwini Sharma: So this contains the email of the customers.
160 00:20:06.060 ⇒ 00:20:10.119 Ashwini Sharma: So, let me go back to this one.
161 00:20:12.100 ⇒ 00:20:12.840 Ashwini Sharma: Yeah.
162 00:20:13.290 ⇒ 00:20:22.680 Ashwini Sharma: So what we have done here is created a dimension called individual, created a dimension called organization, and then created a fact table, which is basically
163 00:20:22.900 ⇒ 00:20:28.000 Ashwini Sharma: the membership records, right? How organizations are related to another organization.
164 00:20:28.200 ⇒ 00:20:33.229 Ashwini Sharma: And then out of these, this report, active member report, is derived.
165 00:20:33.470 ⇒ 00:20:39.870 Ashwini Sharma: Which will utilize all these different tables, do a bunch of joints, and then create a final report.
166 00:20:40.120 ⇒ 00:20:42.960 Ashwini Sharma: Now, like, once… once you…
167 00:20:43.310 ⇒ 00:20:50.040 Ashwini Sharma: Now, this is… this is based on, you know, my understanding of the… of the business, and after talking with Catherine, right?
168 00:20:50.250 ⇒ 00:21:02.569 Ashwini Sharma: In case there is something wrong with the report that I have created, maybe, like, we’ll have to look into what is wrong, and then figure out if another logic change is required in order to correctly create that report.
169 00:21:05.390 ⇒ 00:21:09.980 Ashwini Sharma: And, nope.
170 00:21:10.370 ⇒ 00:21:14.600 Ashwini Sharma: So, if you go here, and then let’s query that report.
171 00:21:39.450 ⇒ 00:21:42.689 Ashwini Sharma: Approximately 1,100, 1,200 records are.
172 00:21:43.210 ⇒ 00:21:45.800 Ashwini Sharma: Coming in over here, and then… You know?
173 00:21:47.070 ⇒ 00:21:51.899 Ashwini Sharma: I mean, exactly like the one that was shown in the spreadsheet, that is what I’ve tried to generate.
174 00:21:52.070 ⇒ 00:21:58.390 Ashwini Sharma: Yeah, that’s… That’s the ActiveM reports, everything about it.
175 00:21:58.850 ⇒ 00:22:03.899 Katherine Bayless: Yeah, that’s awesome. I mean, I know, it’s funny, it feels underwhelming, because we all are data people, right?
176 00:22:03.900 ⇒ 00:22:04.620 Uttam Kumaran: This is genuine.
177 00:22:04.620 ⇒ 00:22:05.560 Katherine Bayless: Generally huge.
178 00:22:05.560 ⇒ 00:22:13.010 Uttam Kumaran: But it’s here, it’s running, it can be run again, we know the logic, you know, like, that’s like, yeah, yeah. Exactly.
179 00:22:13.380 ⇒ 00:22:29.030 Katherine Bayless: Exactly. No, it’s cool. It’s really cool. And that logic should be correct. I mean, yeah, as we go through in QA, I think it’s more likely that we will find data that needs correcting, or maybe there’s just nuance that can be described differently, but, like, in terms of the joins, it should be right.
180 00:22:29.640 ⇒ 00:22:43.240 Uttam Kumaran: Yeah, and our typical, like, development process as we start to, like, do more is… it is sort of looks like this, where, as a developer, you should have the freedom to sort of, like, run SQL, create a table in dev.
181 00:22:43.560 ⇒ 00:22:55.559 Uttam Kumaran: people can poke at it, and then you’ll create a PR in GitHub. That PR will take your new script, run it in staging. That way, you… one, we test that everything compiles.
182 00:22:55.560 ⇒ 00:23:05.399 Uttam Kumaran: In an environment that’s basically a clone of production. At that point, you know, if we are gonna set up, like, staging dashboards or staging pipelines.
183 00:23:05.400 ⇒ 00:23:15.939 Uttam Kumaran: you basically want to be able to test the entire system in isolation outside of prod, right? Like, that’s the mentality. It can… it may be overkill for a short term, but, like, this is what
184 00:23:16.050 ⇒ 00:23:23.599 Uttam Kumaran: sort of this framework allows, and then people review and poke at it, and then we approve, and then it goes to production. Soon as it gets merged.
185 00:23:23.750 ⇒ 00:23:35.929 Uttam Kumaran: It gets picked up by dbt, and then it’ll get materialized, run, and then made available in the production schema. So people… people can select it from there, people in the BI tools will have access. So there’s this, like.
186 00:23:36.100 ⇒ 00:23:38.940 Uttam Kumaran: layering that’s possible. What is, like…
187 00:23:38.990 ⇒ 00:23:51.180 Uttam Kumaran: the opposite, or the usual system, is people just run things in production, there is just, like, no way to test or QA things, it’s a complete nightmare, so this, like, sets that up.
188 00:23:51.180 ⇒ 00:24:04.439 Uttam Kumaran: In this way, and then we hope that over time, like, we… more people are code owners, so you’ll get flagged and get up to review PRs, and, you know, a lot of our flows in Slack will be like, hey, I shipped a PR, can someone
189 00:24:04.550 ⇒ 00:24:14.179 Uttam Kumaran: take a look at the logic, poke at the… poke at the, you know, the output data, and yeah, like, that’ll be the sort of development process, I think, that we’ll get more people involved in.
190 00:24:14.610 ⇒ 00:24:22.680 Katherine Bayless: Yeah, like, I mean, a genuine workflow I see kind of evolving for some of the power users is going to be, like, you know, we’ll build something with
191 00:24:22.720 ⇒ 00:24:36.010 Katherine Bayless: like, you know, the data, and then they’ll come in and work with it with Cortex, and maybe they’ll get, like, a SQL query that’s recommended, right, for the question they’ve been asking, and they’re like, hey, actually, I want to ask this question all the time, can it become, you know, a report or a dashboard?
192 00:24:36.010 ⇒ 00:24:45.880 Katherine Bayless: let them put that, you know, PR in at the dev level, and then we can take a look at it and, like, you know, clean it up, put it into staging, and then push it into prod. So, like, I…
193 00:24:45.880 ⇒ 00:25:05.349 Katherine Bayless: I could see a world where people who don’t today write SQL are at least able to figure out, like, what are the questions I need to ask, how do I get that SQL code from AI, and then, you know, put it into our pipeline for review and publishing. It’ll take a little time to get there, but, like, I think there’s a lot of potential in the AI ability to translate human-speak into SQL.
194 00:25:06.170 ⇒ 00:25:13.829 Uttam Kumaran: Okay, perfect. Yeah, I exactly… I think that’s exactly, like, where we’re hoping it’s going. We had another customer ask us about, like.
195 00:25:13.990 ⇒ 00:25:31.960 Uttam Kumaran: oh, like, why… did we choose Snowflake? Like, why don’t we choose BigQuery because of Gemini and stuff like that? And I said, look, I think most of the stuff short-term in AI is gonna happen at, like, the BI layer, where you can actually loop in more context. I think Snowflake is doing a lot.
196 00:25:32.520 ⇒ 00:25:49.010 Uttam Kumaran: But, I don’t know, like, not a… a lot of people are not going to be consuming data directly out of Snowflake. I think where we’re going to start to use AI features, for example, is, like, instead of complicated case wins or, like, really crazy regex, you may be able to use a more probabilistic AI to categorize or bucket
197 00:25:49.010 ⇒ 00:26:00.239 Uttam Kumaran: that’s a good, great use case, I think, for AI. They do have some interesting, like, forecasting and prediction functions, but, like, I think that’s where we’re going to leverage the AI features directly in Snowflake.
198 00:26:00.720 ⇒ 00:26:02.010 Katherine Bayless: Yeah, yeah.
199 00:26:06.980 ⇒ 00:26:07.610 Uttam Kumaran: Cool.
200 00:26:07.950 ⇒ 00:26:09.960 Ashwini Sharma: Alright, I’ll stop the share.
201 00:26:10.510 ⇒ 00:26:14.679 Uttam Kumaran: We want to talk, Ashwini, about, like, you know, goals for next week.
202 00:26:16.020 ⇒ 00:26:20.070 Ashwini Sharma: Oh, yes, so, yeah, for next week,
203 00:26:20.680 ⇒ 00:26:35.950 Ashwini Sharma: I’m hoping that we could get some data from, Polyatomic, but I’m not really sure if they will finish the connector on time. But what I wanted was, like, we already have some data for remembers. We could utilize that for some other KPIs that you have.
204 00:26:35.980 ⇒ 00:26:42.680 Ashwini Sharma: And also, I remember right now, we talked about some of the flat files that, that were…
205 00:26:42.880 ⇒ 00:26:50.919 Ashwini Sharma: going to be exposed in S3, and then as a table in Snowflake, and then we generate some kind of reporting on top of that, right?
206 00:26:50.920 ⇒ 00:26:51.580 Katherine Bayless: Yeah, so.
207 00:26:51.580 ⇒ 00:26:52.880 Ashwini Sharma: Okay.
208 00:26:52.880 ⇒ 00:26:53.430 Katherine Bayless: Go ahead.
209 00:26:54.020 ⇒ 00:26:59.079 Ashwini Sharma: Yeah, yeah, so I just wanted to understand, like, what exactly are we going to do in that, yeah?
210 00:26:59.380 ⇒ 00:27:11.740 Katherine Bayless: Yeah, so, slight pivot, so the code that we had talked about, that, like, the stuff I’m putting up on the SFTP and whatnot, I think with Polyatomic maybe coming in ideally, hopefully, late next week.
211 00:27:12.160 ⇒ 00:27:24.469 Katherine Bayless: honestly, it’s not a big deal if I just keep running this thing. I’m also learning that as we get closer and closer to the show, I’m really only needing to deal with a couple of the files on a daily basis, rather than, like, all 10 that it was for a while.
212 00:27:24.470 ⇒ 00:27:41.930 Katherine Bayless: So, I think saving me 40 minutes a day for the next 3 weeks is really actually not the most prioritized, thing in my mind anymore. Okay. But, I have a better idea that’s not too dissimilar. So, to your point, Ashwini, now that we have the remembers data, I think starting to build stuff against it makes a lot of sense.
213 00:27:41.930 ⇒ 00:27:57.159 Katherine Bayless: We did get an ask from the membership team yesterday to pull together the member engagement report, which has not been updated in a while, and get it back into service, and so we would be able to use the remembers data for kind of, like, the backbone of that report.
214 00:27:57.520 ⇒ 00:28:03.239 Katherine Bayless: And then we would be able to deliver something to that team in Snowflake, like, now, which would be really cool.
215 00:28:03.690 ⇒ 00:28:11.240 Katherine Bayless: The trick is, there’s some additional data points that go into the report, which is kind of where my brain was starting to go into finger-steepling mode.
216 00:28:11.240 ⇒ 00:28:29.159 Katherine Bayless: So what it… the member engagement dashboard, maybe Kyle, if you’ve got it, you could put it up on the screen so folks can look at it, too. It includes stuff about, like, you know, the company, and how long they’ve been a member, and that kind of stuff, but then also, like, their other activities. So, have they exhibited? Have they attended our webinars? Have they bought research through Shopify?
217 00:28:29.160 ⇒ 00:28:32.910 Katherine Bayless: Actually, parking lot for the end. I do have some feedback on the scopes.
218 00:28:33.070 ⇒ 00:28:35.020 Katherine Bayless: So…
219 00:28:35.520 ⇒ 00:28:51.499 Katherine Bayless: That data is coming, sort of, two places right now, so all of the historical, sort of, data that has already happened and been captured is in that old marketing data warehouse, and then the new data is coming in from a variety of sources.
220 00:28:52.030 ⇒ 00:28:55.519 Katherine Bayless: I think, realistically, even if we could build this
221 00:28:55.520 ⇒ 00:29:12.650 Katherine Bayless: in Snowflake, just against the remembers data plus flat files that are exported from the old marketing warehouse and, you know, kind of marry the structure with the exports we’ll get from the systems. So, like, for example, Zoom, right? So we have an old table in the marketing data warehouse with all the historical webinar attendance.
222 00:29:12.650 ⇒ 00:29:15.909 Katherine Bayless: We just need to go grab the last, like, couple months and add it on top.
223 00:29:15.910 ⇒ 00:29:35.679 Katherine Bayless: Right, so if we could build this against that S3 integration that I’ve got in Snowflake, I think that would give us, A, something we could deliver to the membership team in addition to the active members report, B, it also starts the process of figuring out, well, what do we really want to do with all that old data? Because
224 00:29:35.750 ⇒ 00:29:48.409 Katherine Bayless: We are going to wind up wanting to incorporate the history into some of the new data sources that we’ll be building, and so I think this is a good sort of, like, first encounter with it, and, like, how do we want to structure this?
225 00:29:48.410 ⇒ 00:30:05.870 Katherine Bayless: I think also it gives us an opportunity to take that stage integration thing that I’ve set up with S3 and figure out, like, okay, what is the exact right way to have done this? Because I know I knitted it together and it works, but, like, I’m sure there’s a slightly more formal approach that could be taken to bringing the data in. So I think…
226 00:30:06.000 ⇒ 00:30:19.199 Katherine Bayless: If this is possible to be the, like, next kind of thing that we tackle, I think it’d be a huge win, and also not for nothing, but it would help Kyle, because rebuilding this, at the moment would be kind of a pain in the butt.
227 00:30:19.470 ⇒ 00:30:31.790 Kyle Wandel: I can show you the… the lovely semantic model, but, yeah, it’s a… it’s a beast, and I was literally just gonna, like, even, like, just, like, recreate it, and then just see what happens, but, if any help on it would be great.
228 00:30:33.590 ⇒ 00:30:55.659 Kyle Wandel: And so it all just feeds off this one main member company, or active member company table, and just all the different touchpoints that we have. And I can try to give you a list of that, but maybe the best way to approach it is to, like, look at the semantic model, and I’m gonna try to get you access to it, and then look at the tables from Postgres, SQL Server, and then kind of go from there. But…
229 00:30:56.350 ⇒ 00:31:08.970 Kyle Wandel: there are a lot of tables, obviously. Like, this can be cleaned up, for sure, and that’s what I was going to think about trying to do, but also at the same time, I want to make sure that the membership team has this as seamless transition as possible.
230 00:31:10.130 ⇒ 00:31:14.739 Katherine Bayless: I think, you know, to… just to kind of draw that out a little bit, too, like.
231 00:31:15.590 ⇒ 00:31:20.069 Katherine Bayless: With all of the things that we’ll be sort of reinventing, like.
232 00:31:20.250 ⇒ 00:31:39.329 Katherine Bayless: we do want it to feel familiar and equivalent, but also, you know, there is room for growth and reinterpretation, and so it’s kind of like, as long as we capture the spirit of the old report, and we don’t totally, you know, upend the way that they’re used to engaging with the information, I think that’s a good bridge.
233 00:31:40.270 ⇒ 00:31:45.610 Kyle Wandel: Yeah, we could simplify this from 21 relationships to, like, 5 to 10, that’d be great.
234 00:31:45.940 ⇒ 00:31:47.110 Katherine Bayless: Yeah.
235 00:31:50.150 ⇒ 00:31:51.500 Uttam Kumaran: Heck yeah, okay.
236 00:31:51.500 ⇒ 00:31:54.060 Katherine Bayless: Yeah, I mean, this would really, like, this would be huge.
237 00:31:54.060 ⇒ 00:32:10.710 Kyle Wandel: This would probably… yeah, and this feeds into the identity, like, management type of resolution that I think that you guys are gonna help us out with as well. But, this is probably the biggest ask in our organization, because it touches every single data point across our organization.
238 00:32:11.060 ⇒ 00:32:20.849 Kyle Wandel: And so, I 100% would be working with you guys to give you A access to all the data, and B, verify that that is the correct data to use.
239 00:32:21.000 ⇒ 00:32:22.280 Kyle Wandel: And going forward.
240 00:32:24.990 ⇒ 00:32:25.580 Ashwini Sharma: Cool.
241 00:32:26.610 ⇒ 00:32:32.860 Uttam Kumaran: So I don’t think… I don’t know, Kyle, if this is, like, continuing just to use this touchpoint we have on Thursdays is best. We can also…
242 00:32:33.180 ⇒ 00:32:39.210 Uttam Kumaran: do another kind of call earlier, you know, this week. This call was mainly gonna just be, like.
243 00:32:39.360 ⇒ 00:32:44.680 Uttam Kumaran: what got done, but I’m sure we’ll need, like, probably more working sessions, so maybe we can coordinate in Slack.
244 00:32:45.500 ⇒ 00:33:00.480 Kyle Wandel: Yeah, that’d be fine. I’m, I’m in all week tomorrow, besides, like, the holidays, and then I’m off most of the following week, so that week in between Christmas and January, but I’m definitely happy to jump on the call, if necessary.
245 00:33:01.100 ⇒ 00:33:01.929 Uttam Kumaran: Yeah, okay.
246 00:33:01.930 ⇒ 00:33:03.240 Katherine Bayless: Yeah, same.
247 00:33:05.920 ⇒ 00:33:06.590 Uttam Kumaran: Okay.
248 00:33:07.560 ⇒ 00:33:10.970 Uttam Kumaran: Ashwini, any questions here? I feel like I just gotta poke around this.
249 00:33:10.970 ⇒ 00:33:12.020 Katherine Bayless: Yeah, yeah.
250 00:33:12.240 ⇒ 00:33:21.230 Kyle Wandel: I mean, quite frankly, it’s a beast. So, any questions, please let me know, because I’ve always… I mean, I’ve been thinking about it for a week or two now in terms of, like, how to best tackle it.
251 00:33:24.210 ⇒ 00:33:26.109 Ashwini Sharma: I’ll be reaching out to you, Kyle.
252 00:33:27.120 ⇒ 00:33:28.080 Kyle Wandel: Wonderful.
253 00:33:28.190 ⇒ 00:33:29.190 Kyle Wandel: I think he is.
254 00:33:29.730 ⇒ 00:33:30.400 Katherine Bayless: Cool.
255 00:33:33.580 ⇒ 00:33:41.549 Katherine Bayless: In terms of the scopes for, like, the Okta and Shopify stuff, I can share a little bit of update on those if we want.
256 00:33:41.550 ⇒ 00:33:42.880 Uttam Kumaran: Okay, yeah, please.
257 00:33:43.250 ⇒ 00:33:51.769 Katherine Bayless: Yeah, yeah, okay. So the active one’s still kind of TBD. I think, Jay is kind of deciding how much, how fast, all the things, and like…
258 00:33:52.110 ⇒ 00:34:06.820 Katherine Bayless: who should really be in charge of that, and interestingly, Christina was kind of in, like, our COO was interested in understanding, like, how much of this could we, like, outsource the management of? So, like, not just necessarily the cleanup, but maybe even some of the, like.
259 00:34:06.820 ⇒ 00:34:14.550 Katherine Bayless: day-to-day administration. I know eventually that should be, you know, maybe a 10-minute check in the morning for somebody, right? But, like.
260 00:34:14.560 ⇒ 00:34:26.840 Katherine Bayless: if we are going to start bringing more platforms under enterprise governance, like, next year could be a lot of Okta work, and so having, like, a dedicated resource and a clean environment would probably make a big difference, so…
261 00:34:26.840 ⇒ 00:34:36.760 Uttam Kumaran: Yeah, I just hope that document sort of maybe just spurred the conversation even, or there’s, like, a lot of stuff in there that’s, like, okay, we can consider now. So, totally makes sense.
262 00:34:36.969 ⇒ 00:34:44.709 Katherine Bayless: Yeah, and then for the Shopify one, I actually got a chance to talk to, the person on the marketing team who’s technically kind of the owner for it at the moment.
263 00:34:44.709 ⇒ 00:35:01.299 Katherine Bayless: And I think what I said to her was, would you like it if Shopify didn’t suck so hard? And she’s like, yes! So I sent her the scope to look at, with the caveat of, like, this is, you know, my interpretation of the issue through the lens of the consultant, so, like, you know, poke all the holes and red line as you need.
264 00:35:01.299 ⇒ 00:35:07.739 Katherine Bayless: But she was totally on board with, like, yeah, let’s have somebody come in and try to kind of clean this up.
265 00:35:07.739 ⇒ 00:35:23.849 Katherine Bayless: And the idea of maybe not Shopify? I mean, she said the same thing as you guys said, like, it’s a huge lift to administer, it’s an expensive product, and at the end of the day, she said this, she’s like, we literally, we just need a way for people to, like, have a cart and check out on our website.
266 00:35:23.850 ⇒ 00:35:26.060 Uttam Kumaran: I know.
267 00:35:26.060 ⇒ 00:35:40.340 Katherine Bayless: So I was like, we can do this, yeah. So I’m gonna… I’ll circle back with her probably early next week, and see if she wants to push that one through. She’ll be out on parental leave, I think, in the early part of the year, so it might be something that I can handle while she’s out anyway.
268 00:35:40.340 ⇒ 00:35:42.319 Uttam Kumaran: Okay. I’ll keep you posted on that one.
269 00:35:42.500 ⇒ 00:36:00.649 Uttam Kumaran: Okay, great. Yeah, that’s… that’s helpful. I mean, I also think, in terms of both of those projects, the Okta one, like, from our perspective, is also the riskier one, and it’s just a huge beast. Like, I… in your shoes, I totally… I don’t… I don’t think we would have taken on the management of it at all. I think for us, I’m like.
270 00:36:00.750 ⇒ 00:36:05.029 Uttam Kumaran: Well, we could come in and do, like, a really rich discovery, find everything, and then…
271 00:36:05.160 ⇒ 00:36:17.959 Uttam Kumaran: I’ll tell you what we’re comfortable taking, what you should handle internally, or what options are externally, you know? And so, that’s helpful. And then on the Shopify side, much more confident that, like, whether it’s
272 00:36:18.070 ⇒ 00:36:26.639 Uttam Kumaran: there’s a wide variety of options, like, either some… you can… there are, like, Shopify-esque platforms to do digital asset sales.
273 00:36:26.770 ⇒ 00:36:35.330 Uttam Kumaran: you can also pretty easily build one just on Stripe or other options, so I’m… that is something we’ve done, you know, a bunch of times, so…
274 00:36:36.160 ⇒ 00:36:46.920 Katherine Bayless: Yeah, like, I think revisiting the Shopify just use case generally around, like, okay, what are the things we are selling via it currently, but then also knowing that we have this massive need for, like.
275 00:36:46.950 ⇒ 00:37:01.399 Katherine Bayless: governed data sharing with external partners, like, you know, we’re not asking them to pay for it via Shopify, right? But, like, could we deliver the data through whatever tool replaces Shopify, rather than having multiples, running around, or using SharePoint?
276 00:37:02.330 ⇒ 00:37:07.589 Kyle Wandel: I’m pretty sure we used to just deliver through, straight through our Impex website, so it was as simple as that.
277 00:37:07.780 ⇒ 00:37:12.189 Katherine Bayless: Yeah, she did say there was some… there was some reason…
278 00:37:12.190 ⇒ 00:37:12.770 Kyle Wandel: Right.
279 00:37:12.770 ⇒ 00:37:28.630 Katherine Bayless: Yeah, and I can’t remember what it was, but there was some reason they did move off of just using Impexium for it. It might have had to do just with the ability to, like, render a nice shopping cart on the website, honestly, because, I mean, Impexium’s webpages are… or Remembers’ web pages are not the most glamorous,
280 00:37:28.760 ⇒ 00:37:36.630 Kyle Wandel: Well, and I will say, around that time, we… everybody was kind of hating Impexium, so I think that also probably played into some of the politics of it.
281 00:37:36.630 ⇒ 00:37:39.930 Katherine Bayless: That’s true, yeah, probably. Probably. Yeah.
282 00:37:41.780 ⇒ 00:37:48.759 Uttam Kumaran: And then on our side, Catherine, I also have, like, a Q1 scope sort of ready for you. I sent it internally.
283 00:37:48.850 ⇒ 00:38:04.550 Uttam Kumaran: To our sales team, just to take a look at, but I can send it to you. I mean, again, it’s… it’s… it’s actually, I think, a lot sharper than the original one, because, we added a lot of, like, what we’ve talked about so far. Maybe if I can just flash that up, and I can just share…
284 00:38:04.680 ⇒ 00:38:07.009 Uttam Kumaran: At least the sort of scope.
285 00:38:07.600 ⇒ 00:38:12.480 Katherine Bayless: Yeah, because I did, I talked to, yeah, again, Christina yesterday, and was telling her, like, I think…
286 00:38:12.520 ⇒ 00:38:15.139 Katherine Bayless: Probably what we’ll do next year is… is…
287 00:38:15.140 ⇒ 00:38:33.409 Katherine Bayless: we’ll have SDG maybe take a pause and come back a little bit later once we have some more stuff kind of landed and ready to build against where their, you know, skill set kind of fits a little bit better, but keeping you guys continuous as much as possible, and she was totally on board. So once we have this in hand, I’ll submit it and try to get it worked through that chain.
288 00:38:33.410 ⇒ 00:38:36.360 Uttam Kumaran: Yeah, cause where I see it is, like, I mean…
289 00:38:36.760 ⇒ 00:38:43.880 Uttam Kumaran: right now, all of our focus is on, like, building marts, and so the kind of the scope, maybe I’ll even start at the top, is, like.
290 00:38:44.180 ⇒ 00:38:52.430 Uttam Kumaran: these are sort of, like, the core objectives. I don’t… it’s not clear right now if there is anything related to CES
291 00:38:52.520 ⇒ 00:39:12.349 Uttam Kumaran: But, again, if we do land any interesting data, or we can start to support, like, I definitely want to try to do that, because that’s, like, super, super timely. I think that’s where our team is really good, like, if there’s quick things we need to turn around. But really, the scope here is, like, continue to expand data sources.
292 00:39:12.660 ⇒ 00:39:24.709 Uttam Kumaran: So Salesforce remembers the S3 data, and then what, like, basically whatever Polyatomic can build, and then also any of the other workflows that exist.
293 00:39:24.710 ⇒ 00:39:47.959 Uttam Kumaran: then it’s all data modeling, and this is sort of where I explain, like, on our side, we’ll probably loop in one more person who just focuses on data modeling. So I know Kyle will definitely be, you know, a sort of stakeholder here, and maybe some… a few other folks, so I expect the requirements and the iteration cycles to grow there, so I think it’s really great that we’ve established an ETL tool.
294 00:39:47.960 ⇒ 00:39:51.909 Uttam Kumaran: You know, hopefully establishing utility. We have the data warehouse.
295 00:39:51.910 ⇒ 00:40:01.669 Uttam Kumaran: we have, like, sort of the mechanics of how we’re working together, and we’re gonna see that and learn a little bit of that from working directly with you, Kyle, and then…
296 00:40:02.020 ⇒ 00:40:06.240 Uttam Kumaran: as more stakeholders come in, it’ll just sort of scale there.
297 00:40:06.480 ⇒ 00:40:20.679 Uttam Kumaran: And so that… and then it’s… it’s really just, like, some of this is on, kind of, infra, so just… this is what I already talked about, so all the infra around ETL, orchestration, and then team enablement. So this is, like, onboarding folks.
298 00:40:20.740 ⇒ 00:40:29.599 Uttam Kumaran: you know, dbt and Snowflake. We definitely want to do, like, a cursor demo of, like, how to use cursor to speed up analysis, speed up data modeling.
299 00:40:30.180 ⇒ 00:40:47.289 Uttam Kumaran: questions of, like, how is this defined? How do we think about this? Like, it has, like, dramatically changed the way our team works, and, like, our clients benefit, because we just move way faster, and the quality of our stuff is way higher. I frankly…
300 00:40:47.700 ⇒ 00:41:02.660 Uttam Kumaran: am of the opinion that if your, like, consultant is not using cursor, like, what are you doing? It’s funny, though, because, I don’t know, I just think, like, for the big consulting companies, they sort of see it as, like, oh, it’s…
301 00:41:03.010 ⇒ 00:41:06.050 Uttam Kumaran: We don’t get paid if we, like, ship stuff faster.
302 00:41:06.170 ⇒ 00:41:16.940 Uttam Kumaran: I don’t know, you’re so, like, blinded by, like, this, like, notion of the billable hour, but I don’t know. We don’t think about that at all. I think for us, it’s just, like, actually getting
303 00:41:17.270 ⇒ 00:41:26.479 Uttam Kumaran: It’s actually more of, like, doing as much as we can, as fast as we can, with the highest quality, and using whatever the latest technology is to do that, so…
304 00:41:27.250 ⇒ 00:41:40.880 Katherine Bayless: Yeah, I know, it’s funny, I was, like, talking to an old friend last night, and, I say he was like, all of the, like, the data governance stuff, like, it’s not… it’s not bad by any means, but it’s good things to do, but it’s, like, it gets used as, like,
305 00:41:40.880 ⇒ 00:41:51.049 Katherine Bayless: like a, you know, like a red herring proximate objective. Like, people get all hung up on, like, oh, we just need to do that first. It’s like, you really don’t, actually. Maybe just start using what you have and figure out what matters.
306 00:41:51.050 ⇒ 00:41:52.150 Uttam Kumaran: There’s no nuance.
307 00:41:52.150 ⇒ 00:41:54.330 Katherine Bayless: govern that, yeah, exactly.
308 00:41:54.460 ⇒ 00:41:59.450 Uttam Kumaran: Everything has to have nuance and trade-offs and, like, risk assessment, but…
309 00:42:00.170 ⇒ 00:42:06.900 Uttam Kumaran: I don’t know, like, I think our company and our team is… we’re all, like, ex-product kind of engineers, like, internal, so…
310 00:42:07.130 ⇒ 00:42:08.480 Uttam Kumaran: we don’t…
311 00:42:08.690 ⇒ 00:42:27.480 Uttam Kumaran: like, our timelines that we’re all used to are just, like, move fast if we can, you know, versus… I think if you come from, like, a super formal consulting background, it can be, like, the opposite. It can be, like, no, no, no, like, don’t… like, don’t ship that, or, like, max as much hours as you can, and, like, stuff like that.
312 00:42:27.480 ⇒ 00:42:41.760 Uttam Kumaran: So, I don’t know, but back to what I was saying, we should definitely demo cursor to everybody, and get everybody involved. I think we’re… we’re finding new ways to help it aid with every single part of the data manufacturing
313 00:42:41.930 ⇒ 00:42:45.240 Uttam Kumaran: you know, process, so I think it could be awesome to… to share.
314 00:42:45.600 ⇒ 00:42:46.220 Katherine Bayless: Yeah.
315 00:42:49.770 ⇒ 00:43:04.809 Uttam Kumaran: Yeah, so, I mean, that’s, like, generally the scope for Q1, so I feel like on our side, we’ll probably end up looping in one more person, really, to support, like, on the modeling side. I think with me and Ashwini on the data engineering side, I’m not too worried.
316 00:43:04.810 ⇒ 00:43:20.399 Uttam Kumaran: And then… I think it’ll sort of depend if there… if… if really our goal is going to be supporting, I think, folks like Kyle, I think, like, I feel really comfortable. If there are areas where there’s not, like, an analyst or, like, analytic support.
317 00:43:20.520 ⇒ 00:43:39.150 Uttam Kumaran: that’s where maybe we want to loop in, like, an analyst from our team. I think additionally, Catherine, if you need help preparing for, like, executive stuff, or, like, you’re like, hey, I want to spend a week and, like, really, like, go deep on something, like, that’s where we would loop in, sort of, like, an analyst on our side.
318 00:43:39.410 ⇒ 00:43:42.619 Uttam Kumaran: But that’s, like, always an option. I think.
319 00:43:42.750 ⇒ 00:43:55.979 Uttam Kumaran: again, for us, like, we do, like, kind of… basically do, like, full-stack data. So, I think we… we do a lot of, sort of, executive board decks, investor-type stuff for… for people. We also do, like, really,
320 00:43:56.110 ⇒ 00:44:14.130 Uttam Kumaran: like, time box analysis, which is like, hey, go find, like, the insights in, like, this from the source, or hey, we’re thinking about changing pricing, like, can you run, like, some pricing analyses, or understand, like, the mechanics of how we price. We would do, like, a one to two week sprint, and, like, it’s just, like, really, really rich.
321 00:44:14.200 ⇒ 00:44:26.530 Uttam Kumaran: presentation that is, like, one, like, just, like, answers probably a lot of, like, questions that have been lingering, but, like, we just kind of go after a topic like that. So if any of that’s interesting, I think
322 00:44:26.750 ⇒ 00:44:27.470 Uttam Kumaran: back.
323 00:44:27.690 ⇒ 00:44:35.120 Uttam Kumaran: That we… we don’t… that will kind of come probably after we have stuff modeled, but that’s always an option, and we have some awesome people that
324 00:44:35.230 ⇒ 00:44:40.699 Uttam Kumaran: do, like, financial analysis, sales stuff. And then I also, again, even…
325 00:44:40.810 ⇒ 00:44:43.190 Uttam Kumaran: Even if you guys are interested in, like.
326 00:44:43.500 ⇒ 00:44:57.159 Uttam Kumaran: stuff on, like, the conversion on the website, and, like, more sort of, like, web analytics, product analytics. We just brought on a really great person, who I actually met through the Amplitude Slack channel, but we do a lot of Amplitude and mixed panel work.
327 00:44:57.620 ⇒ 00:45:04.960 Uttam Kumaran: And… yeah, we’re… it’s… That’s all sort of, like, funnel, conversion rate optimization, attribution type stuff.
328 00:45:05.620 ⇒ 00:45:21.879 Katherine Bayless: Yeah, I mean, truthfully, like, yes to all of it, right? So yeah, I mean, I think, like, this scope is perfect for Q1. I think, we’ll definitely start to surface, like, okay, what are the things we want to, you know, take advantage of the additional brainpower for,
329 00:45:21.990 ⇒ 00:45:22.899 Katherine Bayless: I think…
330 00:45:22.950 ⇒ 00:45:42.370 Katherine Bayless: I definitely foresee a lot of work on, like, the marketing data generally in terms of, like, timing, because those were a lot of questions we had this year, was like, did we move the paid advertising up too much during the free period where we lost revenue from people that would have registered later anyway, kind of thing? Like, so getting a really good understanding of, like.
331 00:45:42.400 ⇒ 00:45:48.909 Katherine Bayless: How our marketing is landing, what it’s driving, and what changes in different points in the pipeline, like, what impact they have.
332 00:45:49.070 ⇒ 00:46:04.109 Katherine Bayless: I think that’s gonna be a big one. I also think getting into, like, some small language models, right? Like, trying to figure out, like, what are the things that we actually do have, like, a strong use case and a good amount of data around that we should probably start exploring that.
333 00:46:04.110 ⇒ 00:46:21.869 Katherine Bayless: talking to one of the folks that’s newer on our, GLA team, and he’s, like, the AI guy from IBM’s legal team, and he said, like, you know, that’s their thing, they’re just all in on the small ones, like, we don’t need to mess around with the frontier stuff, we want to make, you know, really highly, or really powerful, like, small language models, and I’m like, yeah, I think
334 00:46:21.920 ⇒ 00:46:26.429 Katherine Bayless: Figuring out where we start working with those would be interesting.
335 00:46:26.520 ⇒ 00:46:42.509 Katherine Bayless: the finance stuff for sure, too. Like, there’s just… I mean, there are so many cool directions to go. So yeah, as soon as we start getting some more data in and playing with it… and I really do think that, like, tackling that member engagement report with the next week or two is gonna be a great way to start.
336 00:46:42.510 ⇒ 00:46:47.760 Uttam Kumaran: Seeing the possibilities of the data, because you guys will get a chance to see, like, what’s really in there, and, like.
337 00:46:47.760 ⇒ 00:46:53.859 Katherine Bayless: The types of things people have been looking at, and then how can we start to move them towards, you know, more robust metrics?
338 00:46:53.990 ⇒ 00:46:57.570 Katherine Bayless: The executive briefing thing, too, I think…
339 00:46:57.750 ⇒ 00:47:03.170 Katherine Bayless: I don’t know that we need to add it to this scope, because I don’t want to delay getting this one submitted, but, like.
340 00:47:03.940 ⇒ 00:47:18.750 Katherine Bayless: I think I will probably be making a budget request in March, so, like, once the post-show stuff is wrapped up, and so, like, because we didn’t get AWS ProServe through this year, and I didn’t necessarily budget for that for next year, so, like.
341 00:47:18.750 ⇒ 00:47:29.129 Katherine Bayless: basically going to the board in March and saying, like, hey, I would like to request an additional, you know, allocation of budget to acknowledge this work that we do still want to do, and some of that kind of stuff, and so getting
342 00:47:29.130 ⇒ 00:47:33.530 Katherine Bayless: prepared for that would be good, and I will happily lean on your expertise there.
343 00:47:33.530 ⇒ 00:47:38.930 Uttam Kumaran: Perfect. Yeah, we can… whatever decks or writing you need, like, we’re more than happy to do that.
344 00:47:38.930 ⇒ 00:47:39.450 Katherine Bayless: Yeah.
345 00:47:39.450 ⇒ 00:47:40.450 Uttam Kumaran: to use us.
346 00:47:40.660 ⇒ 00:47:54.429 Katherine Bayless: Yeah, and then I think, too, like, some work around… and I mean, honestly, this part kind of falls into the scope already, but, like, those metrics and KPIs that are the annual goals, I mean, as much as I’m hoping to eventually change the way we goal set, like, you know.
347 00:47:54.830 ⇒ 00:47:59.099 Katherine Bayless: Putting some solid, visualization and work around those in the near term.
348 00:47:59.560 ⇒ 00:48:01.739 Uttam Kumaran: Yeah, let me just make sure that ends up here.
349 00:48:03.080 ⇒ 00:48:22.940 Kyle Wandel: Yeah, and tying into that, I was gonna say, whether it be tied to, like, whatever visualization tool we go to, but if there’s a way we can kind of build… start building those out, I mean, Catherine and I probably need to talk about it, and it’d be good to give your guys’ opinion on it as well, but, whatever tools we do end up going with, it’d be nice to have them pretty much incorporated pretty seamlessly. I mean…
350 00:48:23.170 ⇒ 00:48:42.700 Kyle Wandel: so what I was… in my head, Kevin, what I was thinking initially, just to talk out loud, is that, like, membership and, like, lower… like, Anna’s can use the Cortex agent pretty easily, but then for any executive report or any report that even, like, Michael Brown and up uses, they would use Streamlit or Sigma, preferably Streamlit, because it’s all in one
351 00:48:42.950 ⇒ 00:48:46.449 Kyle Wandel: Native app, basically, but, and it would be fine.
352 00:48:47.020 ⇒ 00:48:48.450 Katherine Bayless: Yeah. Yeah.
353 00:48:48.450 ⇒ 00:48:53.590 Uttam Kumaran: Yeah, I think that’s where we just basically just develop a couple proof of concepts to get the feel, basically, of, like.
354 00:48:53.590 ⇒ 00:48:54.230 Kyle Wandel: Yeah.
355 00:48:54.230 ⇒ 00:48:55.190 Uttam Kumaran: Yeah, you know.
356 00:48:55.190 ⇒ 00:49:00.199 Kyle Wandel: Even just, like, simple tiles, honestly, is what the executive team likes to see.
357 00:49:00.200 ⇒ 00:49:01.300 Uttam Kumaran: Okay, okay.
358 00:49:01.800 ⇒ 00:49:02.660 Uttam Kumaran: Cool.
359 00:49:03.500 ⇒ 00:49:21.179 Uttam Kumaran: Okay, great. So I’ll just… yeah, I think I want to weave in… let me weave in some of the KPIs and goals sheet here, and I think that’ll be also… I want to just build that into this, which is, like, support as many of that as we kind of get this. And then, also, you know, on our side, I think, as a data team, for us to do, like, some standardization of, like.
360 00:49:21.350 ⇒ 00:49:27.399 Uttam Kumaran: metric definition. So I’ll also send, like, our standard,
361 00:49:27.780 ⇒ 00:49:30.579 Uttam Kumaran: we… just in a Google Sheet, but it’s kind of like…
362 00:49:30.760 ⇒ 00:49:35.309 Uttam Kumaran: Everything… sort of, like, building, like, a semantic understanding of every metric.
363 00:49:35.310 ⇒ 00:49:42.199 Kyle Wandel: So, like, who, like, where is the source, who’s it coming from, what’s the sort of, like, business definition, the data definition, and, like.
364 00:49:42.200 ⇒ 00:49:44.489 Uttam Kumaran: Building in a way where it’s usable to get sign-off.
365 00:49:44.500 ⇒ 00:50:01.609 Uttam Kumaran: Right? So there isn’t a sort of, like, formal definition for these, and maybe it’s a cross-functional group that we need to bring together to be like, this is how we define a member, an active member, stuff like that. And in that sheet, we’ll start by putting in, like, everything we know about the state of the world.
366 00:50:01.610 ⇒ 00:50:12.049 Uttam Kumaran: Like, what’s missing, and then we’ll also put in our commentary on, like, okay, here’s what, like, from what we’re seeing we should do, and then that allows us to work together to just drive
367 00:50:12.210 ⇒ 00:50:23.119 Uttam Kumaran: you know, towards, like, a firm understanding of, like, the business definition for things, and then how that gets codified in dbt, you know? And so that… then there’s, like, this, like, we’re just continuing to reduce
368 00:50:23.290 ⇒ 00:50:38.919 Uttam Kumaran: the confusion on what things mean, and then really nicely, once we layer this into a BI tool, or people are finding in the marts, it’s very, like, one-to-one with that, and there’s trying to reduce the confusion, you know, as much as possible. So I’ll send that, too.
369 00:50:39.160 ⇒ 00:50:44.569 Katherine Bayless: Yeah, and I think, just to make sure we call it out too, right, like, for that work, Kai is totally gonna be… Okay.
370 00:50:44.570 ⇒ 00:50:54.789 Uttam Kumaran: help, because that’s a lot of her background, is, you know, putting people in a room and saying, so what is a member, guys? Yeah, exactly. So she’s really strong on that kind of data governance.
371 00:50:55.260 ⇒ 00:50:56.139 Uttam Kumaran: Okay, perfect.
372 00:50:56.140 ⇒ 00:50:57.520 Katherine Bayless: Yeah, yeah.
373 00:50:58.940 ⇒ 00:51:04.620 Uttam Kumaran: Okay, great, so I will try to wrap this up, Catherine, today, and just get this to you, and then…
374 00:51:05.020 ⇒ 00:51:07.940 Uttam Kumaran: Like, I think, yeah, we should be in a good spot, so…
375 00:51:07.940 ⇒ 00:51:14.850 Kyle Wandel: But I mean, it all looks good, and I mean, I will 100% would love to help, like, I don’t want you guys to have all the fun and be able to build everything.
376 00:51:14.850 ⇒ 00:51:23.189 Uttam Kumaran: No, I’m not trying to have all the fun, believe me. So, this is where I think… this was the first week where we actually…
377 00:51:23.390 ⇒ 00:51:28.720 Uttam Kumaran: like, had the end-to-end set up for, like, one report, right? Because we’re kind of like…
378 00:51:29.340 ⇒ 00:51:35.109 Uttam Kumaran: Building the plane, and then flying to, like, this first destination, and, like, it all sort of, like, happened.
379 00:51:35.400 ⇒ 00:51:37.500 Uttam Kumaran: No, I mean… Yesterday, so…
380 00:51:37.740 ⇒ 00:52:02.730 Kyle Wandel: quite frankly, like, at the very beginning of it, I didn’t tell Catherine this, but at the very beginning, I was worried… not worried for you guys, but in terms of I just want to be able to do it myself. I want to learn myself, I want to be able to do it myself, but honestly, the past, like, 2 months have told me anything. Like, I’m just putting… we’re putting out fires just on a regular basis, and so I can’t even get access to, like, creating that document, which you talked about in Excel, and having that overarching look at, people… stakeholders can look at it, and kind of define those business rules, like, that’s something
381 00:52:02.730 ⇒ 00:52:07.400 Kyle Wandel: that… It’s extremely important, and something that I’ve been wanting to do, but haven’t had the time to.
382 00:52:07.540 ⇒ 00:52:17.529 Uttam Kumaran: Yeah, and like, that’s what we need from y’all, like, from the… from externally, we can’t drive that, and we can’t get the strategy right, and so… but also, totally, I think we are, like.
383 00:52:17.570 ⇒ 00:52:37.399 Uttam Kumaran: we try not to gatekeep on any of the ways that we do this, so I think what you’ll find is, like, we want as many people as possible to, of course, be in Snowflake, but, like, be able to review dbt PRs, write their own models, you know, and, like, that scales where we can go focus on, versus
384 00:52:37.400 ⇒ 00:52:42.590 Uttam Kumaran: You know, we will still… we can go one by one and create models, but where you want to utilize us is stuff that’s, like.
385 00:52:42.740 ⇒ 00:52:51.260 Uttam Kumaran: really messy, or, like, I don’t know what’s in there, figure that out, like, that’s where… and then we’ll start to just get everybody involved in this process, so…
386 00:52:54.440 ⇒ 00:53:04.710 Katherine Bayless: I think, I mean, I think it’ll evolve really organically, like, and definitely post-show, you know, things will be different, we won’t say better or worse, necessarily, but different, right? And so, yeah, like…
387 00:53:04.800 ⇒ 00:53:20.909 Katherine Bayless: I think, yeah, I’m very excited. Like, I’m very excited about the things that we’ve already got in here. I think the… the other thing that’s kind of on my mind, I know we’ve talked about this before, is, like, the RBAC and management, and just, like, how do we want to put some, you know, policy, again, like, mental policy.
388 00:53:20.910 ⇒ 00:53:21.420 Uttam Kumaran: the…
389 00:53:21.420 ⇒ 00:53:34.990 Katherine Bayless: And then also literal, but, like, what are the mental models we’re gonna use for assigning people to different things, and what are the things we’re looking for to give somebody the, like, alright, yeah, you can be writing models, like, we trust you in the dev stage, that kind of thing.
390 00:53:35.290 ⇒ 00:53:55.189 Uttam Kumaran: Yeah, exactly. So I think we do have the framework for that, Ashwini, and the way we set up roles, and then it’ll be for us to determine, like, who is going to be a developer versus just, like, read access to, like, a certain mart. And then also, like, we want this to layer into the BI, you know, layer, which is, like, people are just accessing certain things, so…
391 00:53:55.530 ⇒ 00:53:56.300 Katherine Bayless: Yeah.
392 00:53:56.780 ⇒ 00:54:05.869 Kyle Wandel: Probably further down the line, but if you could, like, define why the person, or that, like, the type of user that would use that type of stuff, that’d be great, because I…
393 00:54:06.310 ⇒ 00:54:22.150 Kyle Wandel: I’m a little hesitant, I’m not as optimistic as Catherine is in terms of people being able to be full-on developers, but it would also help us kind of establish to leadership that, hey, that maybe one department, or a department should have at least one data-heavy person on their team to understand it, basically.
394 00:54:22.150 ⇒ 00:54:34.279 Uttam Kumaran: to train them, like, a central group onboards and trains them and sort of supports, and then… but also, again, if there are sensitive data or PII, the way this typically goes is, like, you submit a request.
395 00:54:34.360 ⇒ 00:54:45.179 Uttam Kumaran: That way, there’s, like, an audit log, and then you get either temporary or some time-based access to a role. But the only thing is, it just takes resources to manage, you know, like.
396 00:54:45.180 ⇒ 00:54:50.880 Katherine Bayless: Yeah. That, so I think that’s the biggest thing for us as we start to get bigger, there’s more people involved.
397 00:54:50.890 ⇒ 00:54:53.839 Uttam Kumaran: To just see, like, what do we have bandwidth for as a crew?
398 00:54:55.230 ⇒ 00:54:55.730 Katherine Bayless: Yeah.
399 00:54:55.730 ⇒ 00:54:56.799 Uttam Kumaran: You know, so…
400 00:54:57.110 ⇒ 00:55:19.630 Katherine Bayless: I should also note, I chose to pause the hiring, because I was just like, I can’t do anything else. But, like, I will resume the recruitment for the additional data engineer in the spring, so this will be… I’m kind of, you know, the team’s kind of flat at the moment, but eventually, somebody that would report to Kyle, and in the near term would, you know, work very closely with. And so, looking for candidates and that kind of stuff, probably will lean on you guys a little bit for that, too.
401 00:55:19.630 ⇒ 00:55:27.439 Uttam Kumaran: Yeah, but if I can help with JDs or interviews, more than happy to, and then… yeah, I will think if you guys, once you…
402 00:55:27.830 ⇒ 00:55:32.040 Uttam Kumaran: Yeah, once that job posting’s up, let me know, I will send it to some people in that area, so…
403 00:55:32.040 ⇒ 00:55:39.180 Katherine Bayless: Okay, yeah, yeah. The trick is, they do have to be in Crystal City 3 days a week, which is kind of, like, alright, but yeah, yeah.
404 00:55:39.530 ⇒ 00:55:43.179 Katherine Bayless: There’s smart people there, so… Oh, totally, totally, yeah, yeah.
405 00:55:43.180 ⇒ 00:55:45.369 Uttam Kumaran: We’re all working at Amazon, though, I’m joking.
406 00:55:46.960 ⇒ 00:55:48.400 Uttam Kumaran: That’s true, that’s true.
407 00:55:48.400 ⇒ 00:55:49.693 Katherine Bayless: Right, right.
408 00:55:51.250 ⇒ 00:55:59.209 Uttam Kumaran: Okay, amazing. Okay, so I’ll get back a version of this to you today, Catherine, and then I think, Ashwini, you and Kyle are meeting later today.
409 00:55:59.270 ⇒ 00:56:00.330 Ashwini Sharma: Yeah.
410 00:56:00.330 ⇒ 00:56:08.860 Uttam Kumaran: If you can make sure, as me, we get, any of the links or resources from Kyle on the… on the new project, we can plan that out together.
411 00:56:09.180 ⇒ 00:56:16.420 Kyle Wandel: Yeah, we gotta get you added to Power BI, because I looked to try and add you now, and I couldn’t find your email, but I looked again and try to get you guys incorporated to those workspaces.
412 00:56:16.610 ⇒ 00:56:17.970 Uttam Kumaran: Okay, perfect.
413 00:56:18.800 ⇒ 00:56:21.990 Katherine Bayless: Yeah, I don’t… I honestly, I can’t remember if we had said.
414 00:56:21.990 ⇒ 00:56:22.370 Kyle Wandel: Yeah.
415 00:56:22.510 ⇒ 00:56:26.729 Kyle Wandel: with Power BI. I know we talked about it, but if we didn’t, we will, but yeah.
416 00:56:27.480 ⇒ 00:56:39.529 Katherine Bayless: And I will, I’ll modify the IAM permissions so that that bucket with the old data can come into Snowflake, and then however you want to bring it in, up to you, but yeah.
417 00:56:39.530 ⇒ 00:56:40.990 Uttam Kumaran: Okay. Perfect.
418 00:56:41.300 ⇒ 00:56:41.890 Katherine Bayless: Cool.
419 00:56:42.800 ⇒ 00:56:45.430 Uttam Kumaran: Alright, thank you guys, appreciate it.
420 00:56:45.430 ⇒ 00:56:45.990 Kyle Wandel: Thank you.
421 00:56:45.990 ⇒ 00:56:46.570 Uttam Kumaran: Thank you guys.
422 00:56:46.570 ⇒ 00:56:48.360 Ashwini Sharma: Thank you. I’ll see you guys in a little bit.
423 00:56:48.550 ⇒ 00:56:50.090 Uttam Kumaran: Goodbye.