2026-01-30_brainforge_x_cta_weekly_d2df76f7

Meeting Title: Brainforge x CTA: Weekly! Date: 2026-01-30 Meeting participants: Chi Quinn, Ashwini Sharma, Uttam Kumaran, Awaish Kumar, Katherine Bayless, Kyle Wandel

---

WEBVTT

1 00:00:18.910 ⇒ 00:00:20.000 Uttam Kumaran: Hello.

2 00:00:21.050 ⇒ 00:00:22.480 Chi Quinn: Hi, good morning!

3 00:00:22.480 ⇒ 00:00:24.569 Uttam Kumaran: Hey, good morning. How are you?

4 00:00:24.570 ⇒ 00:00:27.660 Chi Quinn: I am good, hanging in there, trying to stay warm.

5 00:00:28.070 ⇒ 00:00:30.709 Uttam Kumaran: Yeah, you’re… I assume you’re in DC as well?

6 00:00:30.710 ⇒ 00:00:32.430 Chi Quinn: Yeah, I’m in the area, so it’s…

7 00:00:32.430 ⇒ 00:00:32.880 Uttam Kumaran: Nice.

8 00:00:32.880 ⇒ 00:00:36.389 Chi Quinn: Still pretty icy, but it’s all good.

9 00:00:36.390 ⇒ 00:00:42.780 Uttam Kumaran: Right. Yeah, I’m glad. It’s still really cold here in Austin, which is, like, really surprising, I feel like.

10 00:00:42.880 ⇒ 00:00:46.379 Uttam Kumaran: Usually… It’s warm here, and it’s 30s.

11 00:00:46.910 ⇒ 00:00:51.010 Chi Quinn: Oh, wow, so I know that… that’s not good.

12 00:00:51.010 ⇒ 00:00:56.779 Uttam Kumaran: It just changes, like, the, like, life here. It’s just not the way it’s supposed to be.

13 00:00:56.780 ⇒ 00:01:07.430 Chi Quinn: I know, I feel like that’s pretty much all of within the U.S. right now, they’re just experiencing some kind of cold, and even somewhat in Florida, where I’m originally from, so…

14 00:01:07.430 ⇒ 00:01:10.860 Uttam Kumaran: Yeah, I thought… they said there’s part of it’s gonna get snow, right, in Florida?

15 00:01:10.860 ⇒ 00:01:15.079 Chi Quinn: That’s what they’re… I mean, they’re thinking, but I doubt it. It’s…

16 00:01:15.080 ⇒ 00:01:15.790 Uttam Kumaran: Okay.

17 00:01:15.790 ⇒ 00:01:20.429 Chi Quinn: Maybe, like, a little flurry or two, and then, you know, it’ll be all good.

18 00:01:21.590 ⇒ 00:01:24.030 Uttam Kumaran: That’s still rare, you know, I feel like.

19 00:01:26.290 ⇒ 00:01:31.749 Uttam Kumaran: Or Florida, yeah, even here, just to get, like, this much ice and stuff is, like, not common at all.

20 00:01:31.750 ⇒ 00:01:32.680 Chi Quinn: Yeah.

21 00:01:32.680 ⇒ 00:01:36.779 Katherine Bayless: That’s what I was gonna say, is, like, it’s the… the unfamiliarity that’s really the…

22 00:01:36.780 ⇒ 00:01:39.300 Uttam Kumaran: Yes. Yeah. Exactly.

23 00:01:40.800 ⇒ 00:01:41.400 Katherine Bayless: Yeah.

24 00:01:42.370 ⇒ 00:01:43.459 Katherine Bayless: Good morning.

25 00:01:43.460 ⇒ 00:01:44.300 Uttam Kumaran: Good morning!

26 00:01:44.300 ⇒ 00:01:44.750 Katherine Bayless: Funny.

27 00:01:44.750 ⇒ 00:01:46.359 Uttam Kumaran: How’s… how’s everything?

28 00:01:46.990 ⇒ 00:01:53.430 Katherine Bayless: It’s good. I feel like my brain is very ready for the weekend, like… Okay. Yeah, I’m like…

29 00:01:53.430 ⇒ 00:02:07.509 Uttam Kumaran: Really? I’m, like, catching, like, a second stride. I feel like I’m, like, we’re gonna be doing a bunch of modeling for y’all today. I’m, like, I’m fine, I’m, like, pumped. I’m, like, finally, like, I think we got through this first hump. So, I don’t know. I feel like my… my week…

30 00:02:07.840 ⇒ 00:02:11.000 Uttam Kumaran: Has been… middle of the week was tough, and then…

31 00:02:11.290 ⇒ 00:02:13.739 Uttam Kumaran: feel like I’m having… and then, it’s funny, because…

32 00:02:13.990 ⇒ 00:02:24.050 Uttam Kumaran: or there’s a guy on our AI team who I’ve been doing a lot in cursor, and I’m like, dude, we just need to spend, like, 2 hours together, because I’m doing a lot of stuff with data.

33 00:02:24.240 ⇒ 00:02:41.299 Uttam Kumaran: I think you probably know how to speed these up, but, like, you don’t know much about data, so… but, like, you do know a lot about AI. Can we just, like, chat for, like, 2 hours, and just, like, I’m gonna do some stuff, you… we can just brainstorm? So, that’s ultimately… I’m excited for that. He’s gonna stay up and work with me on some stuff.

34 00:02:41.660 ⇒ 00:02:44.470 Katherine Bayless: Nice. I, like, kind of jealous, actually, since my.

35 00:02:46.510 ⇒ 00:02:53.830 Uttam Kumaran: I know, I would open it up, but I’m like, I don’t even… it’s just gonna be like, hey, I wanted to try this, like, how should we do this? Like, what do you think?

36 00:02:53.830 ⇒ 00:02:54.170 Katherine Bayless: Fantastic.

37 00:02:54.170 ⇒ 00:03:00.880 Uttam Kumaran: You know, but we’re gonna talk a little bit about some of the data analysis thing that came out in that OpenAI article, for sure.

38 00:03:00.990 ⇒ 00:03:15.039 Uttam Kumaran: And then I think when I… I owe Jay a reply, but I think when I show y’all how to, like, kind of cursor, I think you’ll see kind of the goal of, like, having the shared context, having MCPs, having CLIs, and then…

39 00:03:15.630 ⇒ 00:03:23.389 Uttam Kumaran: basically being able to, like, iterate over, like… basically what it does is just goes in multiple places, strategically, like, pulls

40 00:03:23.710 ⇒ 00:03:32.489 Uttam Kumaran: codifies, throws into a prompt, then it thinks, and then it just kind of iterates, right? And that’s great, like, it’s actually really, really powerful, so…

41 00:03:33.280 ⇒ 00:03:45.379 Katherine Bayless: Yeah, no, I’m actually… I’m really excited to, do that, like, demo and overview next week. And yeah, I think, yeah, Jay, I think he’s interested in Cursor. I think he’s… his question is, like, like, hmm, tell me why you have.

42 00:03:45.380 ⇒ 00:03:52.300 Uttam Kumaran: Yeah, exactly. So yeah, I have a… I was thinking about it this morning, and I’m like, okay, I have a long message to share about how we arrived here.

43 00:03:53.460 ⇒ 00:03:54.560 Katherine Bayless: But, yeah.

44 00:03:55.080 ⇒ 00:04:08.279 Katherine Bayless: Yeah. Yeah, I mean, I think… I guess, actually, silly question, but I’m assuming, like, cursor is the kind where, like, you kind of, like, pick whichever models you want to work with over… or is it… it’s not, like, a cursor LLM, is it?

45 00:04:08.280 ⇒ 00:04:12.750 Uttam Kumaran: No, no, no. Cursor is, it’s like a VS Code-based IDE.

46 00:04:12.750 ⇒ 00:04:13.210 Katherine Bayless: Indeed.

47 00:04:13.210 ⇒ 00:04:17.690 Uttam Kumaran: So, similar ones are, like, windsurf, or Klein, or…

48 00:04:19.209 ⇒ 00:04:34.920 Uttam Kumaran: like, Google has one called Anti-Gravity now, but it’s… consider it just, like, a wrapper on VS Code. Yes, you can do a lot of the things, like switching the models you want, but actually, they’ve… they, one, they just shipped their own model called Composer that is really…

49 00:04:34.920 ⇒ 00:04:44.609 Uttam Kumaran: integrated well with the ID, and for example, like, I commonly will be working on things with 4 or 5 repos, like, hundreds and hundreds of files open.

50 00:04:44.660 ⇒ 00:04:50.430 Uttam Kumaran: And it quickly will index, and you’re able to discuss things. And then also, it’s also, you kind of set up

51 00:04:50.570 ⇒ 00:05:01.849 Uttam Kumaran: our… the repo in a way where anybody’s using cursor, you can put in rules, so you can basically use it to be like, hey, I want to learn more about this concept. There will be rules in the repo that anyone else’s cursor

52 00:05:01.920 ⇒ 00:05:11.949 Uttam Kumaran: can sort of… will know that it’s like a cursor MD file, basically, and it’ll read that, be like, oh yeah, if I get asked this type of question, here’s the way I’m supposed to respond, so you can kind of steer it

53 00:05:12.210 ⇒ 00:05:23.010 Uttam Kumaran: globally, in a nice way. I mean, the big thing is, like, for us, because most of our company is, like, all fairly technical, I feel like it’s the best…

54 00:05:23.300 ⇒ 00:05:38.970 Uttam Kumaran: ID… it’s the best way to kind of do a lot of knowledge work right now, but, some people still are going to prefer to use Claude for work, or use ChatGBT. And so, ideally, the goal there is, like, can we make sure that they have the right plugins?

55 00:05:39.220 ⇒ 00:05:40.930 Katherine Bayless: You know, available.

56 00:05:40.930 ⇒ 00:05:52.429 Uttam Kumaran: And that they’re able to, like, at least if they’re not able to have the customization that a cursor allows, they’re still able to get all the information, and it’s not just, like, an isolated ChatGPT instance, right? So…

57 00:05:52.850 ⇒ 00:06:00.530 Uttam Kumaran: that’s kind of how, but then again, like, even that layer, it’s like, Slack, like, we want to really democratize it in Slack.

58 00:06:01.100 ⇒ 00:06:01.430 Katherine Bayless: Yeah.

59 00:06:01.430 ⇒ 00:06:15.350 Uttam Kumaran: So, like, what is the ergonomics? Like, do people need to be able to choose, like, the model? Are we, like, gonna basically have, like, a series of questions that you can ask, and then if we encounter one that it’s not sure about, like, you’re kind of building this… this sort of, like.

60 00:06:15.640 ⇒ 00:06:18.299 Uttam Kumaran: This… this product, you know, a little bit.

61 00:06:18.410 ⇒ 00:06:22.780 Uttam Kumaran: So, yeah, it’s like a lot to unpack there.

62 00:06:23.260 ⇒ 00:06:24.710 Katherine Bayless: Yeah, that’s cool, though.

63 00:06:24.710 ⇒ 00:06:25.450 Uttam Kumaran: Yeah.

64 00:06:25.810 ⇒ 00:06:26.670 Katherine Bayless: Hey.

65 00:06:27.830 ⇒ 00:06:40.870 Uttam Kumaran: Cool, so today, really, I think I wanted to just focus time on sort of sharing, like, like, kind of, like, what’s in the archive data, and I think I heard your note, Catherine, loud and clear, like, I think we’re gonna try to drive to…

66 00:06:40.990 ⇒ 00:06:45.920 Uttam Kumaran: Get as much of the member engagement data model as we can.

67 00:06:46.230 ⇒ 00:06:48.720 Uttam Kumaran: So just to highlight,

68 00:06:49.390 ⇒ 00:06:53.920 Uttam Kumaran: where we are here. So, we have all the archived data loaded here.

69 00:06:54.120 ⇒ 00:06:58.129 Uttam Kumaran: Which is great. I think more than enough for us to…

70 00:06:58.440 ⇒ 00:07:02.650 Uttam Kumaran: Go after each of the sections of the Power BI report.

71 00:07:02.850 ⇒ 00:07:05.990 Uttam Kumaran: So today, I actually just want to, like.

72 00:07:06.160 ⇒ 00:07:15.500 Uttam Kumaran: go through the report one more time and just confirm a few questions before we sort of just, like, go heads down and try to build towards that. Additionally, I’m assuming that

73 00:07:15.650 ⇒ 00:07:18.770 Uttam Kumaran: are… Goal is just gonna be, like, hey.

74 00:07:19.240 ⇒ 00:07:25.479 Uttam Kumaran: These parts of the report you can source from this table or this… these joins.

75 00:07:25.740 ⇒ 00:07:27.530 Uttam Kumaran: And then try to deliver that

76 00:07:28.830 ⇒ 00:07:32.460 Uttam Kumaran: by some time where Kai and Kyle can take it from there.

77 00:07:33.040 ⇒ 00:07:57.009 Katherine Bayless: Yeah, totally. I think that’s, I think that’s exactly right, because then, we can… I mean, honestly, like, AI and Streamlit go really nicely together. So, like, the little, sort of, you know, coming soon type mock-up that I had put together, I mean, that was just, like, a one-shot, and it came out pretty well. Okay. So yeah, once I think the stuff that’s… the components of the report are in ProdMarts, then yeah, we can totally take it from there.

78 00:07:57.010 ⇒ 00:07:57.870 Katherine Bayless: And I actually…

79 00:07:57.870 ⇒ 00:08:03.969 Katherine Bayless: Kind of keen to, like, see how much of an appetite some of the membership folks might have for, like, you know, playing with it, too.

80 00:08:03.970 ⇒ 00:08:08.430 Uttam Kumaran: Yeah, I mean, totally. I feel like that’s honestly part of,

81 00:08:08.780 ⇒ 00:08:16.079 Uttam Kumaran: even the demo of whatever, you should share, like, how you built the demo. Maybe someone’s like, oh, I actually think I could, like, build

82 00:08:16.460 ⇒ 00:08:18.769 Uttam Kumaran: What we need faster, too, you know?

83 00:08:19.040 ⇒ 00:08:21.300 Katherine Bayless: Yeah, yeah, exactly, exactly.

84 00:08:21.870 ⇒ 00:08:27.239 Uttam Kumaran: So maybe I just want to ask a couple of questions about the report,

85 00:08:27.440 ⇒ 00:08:31.160 Uttam Kumaran: itself, so I think anywhere, that’s, like.

86 00:08:31.520 ⇒ 00:08:40.359 Uttam Kumaran: a fixed thing. I think I feel pretty good about most of this. I actually had a question. There’s some, like, cutoff on some of these, so I just want to confirm that, like.

87 00:08:40.530 ⇒ 00:08:42.570 Uttam Kumaran: These are all the columns here.

88 00:08:43.039 ⇒ 00:08:43.779 Katherine Bayless: Hmm.

89 00:08:44.049 ⇒ 00:08:50.879 Uttam Kumaran: And or, like, these are all of them, and this is just kind of the same thing, maybe it’s just, like, expanded here.

90 00:08:51.249 ⇒ 00:08:54.859 Uttam Kumaran: those are just the small things that I’m like, I just want to try to get as much

91 00:08:55.199 ⇒ 00:09:01.189 Uttam Kumaran: in the first pass as possible, to, like, be like, okay, cool, you can go to X table to find all of these items.

92 00:09:01.720 ⇒ 00:09:15.239 Katherine Bayless: Yeah, that’s a good point. Maybe, Kai can help track down the… actually, the other thing is, I have that Power BI, like, where I’d ask Claude Code to unpack the Power BI file, and give me the, like, inventory of all the fields that were.

93 00:09:15.240 ⇒ 00:09:18.220 Uttam Kumaran: Oh, yeah. So that’s it, that’s perfect.

94 00:09:18.220 ⇒ 00:09:33.099 Katherine Bayless: Yeah, maybe I can send that over to you. And then, I guess it might mean, even though you wouldn’t necessarily be, like, using Power BI Desktop or anything, like, we could still share the PBIX files, okay. They’re already in stuff that you have access to, but I’ll figure out where they’re hiding.

95 00:09:33.100 ⇒ 00:09:40.610 Uttam Kumaran: Okay, yeah, I mean, I think it’s good. I didn’t know that… yeah, until you mentioned it, I didn’t know that Claude could open those until you mentioned it earlier this week, so…

96 00:09:40.610 ⇒ 00:09:44.040 Katherine Bayless: Yeah. And I had no idea that it would be able to, I just was like, I was like.

97 00:09:44.040 ⇒ 00:09:48.609 Uttam Kumaran: Because if you open, like, a PDF, it’s, like, all just, like, it’s, like, unusable.

98 00:09:48.610 ⇒ 00:09:49.480 Katherine Bayless: Right. Text.

99 00:09:49.590 ⇒ 00:09:50.590 Uttam Kumaran: So…

100 00:09:50.590 ⇒ 00:09:54.749 Katherine Bayless: But this, I was like, oh no, I can totally read that. I was like, okay, go for it.

101 00:09:54.750 ⇒ 00:10:04.119 Uttam Kumaran: That would be great, because, you know, if we have those, I’ll put those in the repo, and I’ll actually… I’ll actually even ask Kirscher, just, like, once we develop the marts, to even be like, okay, we have…

102 00:10:04.440 ⇒ 00:10:05.460 Uttam Kumaran: is, like.

103 00:10:05.850 ⇒ 00:10:12.089 Uttam Kumaran: tell me if we’re missing anything, or, you know, help me construct the joins, or, like, how we can instruct to construct the joins, so…

104 00:10:12.500 ⇒ 00:10:13.760 Katherine Bayless: Okay, yeah.

105 00:10:14.450 ⇒ 00:10:21.890 Uttam Kumaran: And is this… is this… like, I think we’re gonna take Samsung as, like, our, like, QA… Okay, cool.

106 00:10:22.010 ⇒ 00:10:24.670 Katherine Bayless: I mean… I get picked on a lot as the QA.

107 00:10:24.670 ⇒ 00:10:25.310 Uttam Kumaran: Okay.

108 00:10:25.310 ⇒ 00:10:30.379 Katherine Bayless: I think partly because they just have such a sprawling footprint of activity with us, but yeah.

109 00:10:30.890 ⇒ 00:10:36.119 Kyle Wandel: They’re, like, the worst case, and sorry I’m late, but they’re, like, the worst case of the identity management resolution, so…

110 00:10:36.120 ⇒ 00:10:36.820 Uttam Kumaran: Okay, okay.

111 00:10:36.820 ⇒ 00:10:38.690 Kyle Wandel: It’s a really good opportunity for…

112 00:10:38.820 ⇒ 00:10:40.980 Kyle Wandel: Right. Basically, just trying to dedupe them.

113 00:10:42.260 ⇒ 00:10:54.680 Uttam Kumaran: Okay, great. Yeah, so among the three of us, we… today, we sort of, like, piecemeal this. I’m… I was like, let me… I wanted to take the identity management, like, the… the identity matching stuff.

114 00:10:54.990 ⇒ 00:10:56.920 Uttam Kumaran: And then, yeah, we’re just gonna try to…

115 00:10:57.740 ⇒ 00:11:05.849 Uttam Kumaran: bang as many of these out as we can. I think Monday or Tuesday, we’ll probably do an internal review, and then try to get a first pass of something out.

116 00:11:05.960 ⇒ 00:11:08.460 Uttam Kumaran: And then… at least…

117 00:11:08.850 ⇒ 00:11:17.859 Uttam Kumaran: give something that we can start to build for… I know Thursday is sort of, like, a deadline on just trying to get something over, so for me, it’s like, if I can get

118 00:11:18.260 ⇒ 00:11:21.779 Uttam Kumaran: If we can, like, put pencils down on some stuff on Tuesday.

119 00:11:22.030 ⇒ 00:11:28.709 Uttam Kumaran: And then that at least give some, like, Tuesday midday, at least gives some time to get feedback, and then collaborate Wednesday on trying to get something out.

120 00:11:29.590 ⇒ 00:11:33.819 Katherine Bayless: Yeah, I like that, I like that. And we did meet with,

121 00:11:33.930 ⇒ 00:11:46.590 Katherine Bayless: the team yesterday, and they were saying, so, like, of the components here, like, the CES data, so the attendee side, the registrations, the exhibitor side, the exhibits.

122 00:11:46.680 ⇒ 00:12:07.350 Katherine Bayless: And then the innovation awards and the committee stuff, like, those are kind of the big pieces that they’re, like, you know, most interested in having. Some of the smaller stuff, like the research downloads and the webinars, were less urgent, were, like, useful. So I think even if we can get through the registration, exhibit, awards, and committee stuff.

123 00:12:07.540 ⇒ 00:12:10.870 Katherine Bayless: I think if we deliver that on Thursday, they’ll be elated.

124 00:12:11.590 ⇒ 00:12:12.580 Uttam Kumaran: Okay, okay.

125 00:12:15.130 ⇒ 00:12:23.950 Kyle Wandel: And let me know if there’s anything I can help out with with the report in general. So I got the CES registration data working, but, just let me know if I can help out with anything else further with it.

126 00:12:24.060 ⇒ 00:12:25.910 Kyle Wandel: Yep.

127 00:12:27.760 ⇒ 00:12:28.320 Uttam Kumaran: Okay.

128 00:12:29.930 ⇒ 00:12:47.930 Kyle Wandel: I guess it kind of… I mean, it kind of goes into what I was doing while I was, like, these band SCAD scan data. And Catherine, this is a good question maybe for the entire team, but everybody… it looks like there potentially might be a way to fix some of the codes, but, like, it’d be a case-by-case situation, so…

129 00:12:47.930 ⇒ 00:12:51.019 Kyle Wandel: For example, I just got told by the conference team that

130 00:12:51.020 ⇒ 00:13:02.240 Kyle Wandel: one session, there’s no possible way I could have 2,000 people. And so, what I noticed is, one of the codes is Foundry02, the other code is Foundry 2.

131 00:13:02.570 ⇒ 00:13:08.850 Kyle Wandel: So one of them is an entry scanner, one of them is a sessions code scanner, and so it’s likely that…

132 00:13:09.500 ⇒ 00:13:14.870 Kyle Wandel: That’s likely one of the biggest issues in terms of what happened between these multiple-day scanners.

133 00:13:15.170 ⇒ 00:13:26.620 Kyle Wandel: So, I guess the question is, do I try and fix it for all of them using just basic logic of scan time and scan date, or do we do a case-by-case situation, or just leave as is?

134 00:13:27.230 ⇒ 00:13:28.080 Katherine Bayless: Hmm.

135 00:13:29.080 ⇒ 00:13:34.130 Katherine Bayless: I don’t know, I’m inclined to say…

136 00:13:35.530 ⇒ 00:13:41.329 Katherine Bayless: Because I know there are a lot of sessions in there. I feel like probably the answer is maybe if…

137 00:13:42.620 ⇒ 00:13:55.509 Katherine Bayless: if you could come up with a list of the ones that aren’t playing nice, and ask the conference’s team to just, like, run through it and correct them, like, just in a, you know, little CSV with the mappings, and then we could pipe that back in, but.

138 00:13:56.920 ⇒ 00:13:57.480 Kyle Wandel: Excellent.

139 00:13:57.480 ⇒ 00:14:00.479 Katherine Bayless: there are. I mean, if there’s, like, hundreds, then that’s not very nice, but…

140 00:14:00.840 ⇒ 00:14:06.020 Kyle Wandel: Yeah, I think there’s, like, 35 to 50, I think I just looked at it, but there’s, like, 40… yeah, like, 40-something.

141 00:14:06.310 ⇒ 00:14:13.099 Katherine Bayless: Yeah, so then I would say maybe you ask them if they would be able to just, like, give you the correct mappings for the codes, and then…

142 00:14:13.220 ⇒ 00:14:14.859 Katherine Bayless: Like, go from there.

143 00:14:16.310 ⇒ 00:14:16.880 Uttam Kumaran: Okay.

144 00:14:19.260 ⇒ 00:14:32.240 Katherine Bayless: the badge scan data generally, I don’t wanna… I’m like, yeah, I’m like, rabbit hole, because I’m also working on a different, like, lens on it, for the marketing team, and so I think, I think I have,

145 00:14:32.730 ⇒ 00:14:37.120 Katherine Bayless: I have some ideas, and I have some questions, but we can stay on the membership engagement report so that, you know.

146 00:14:37.120 ⇒ 00:14:37.750 Uttam Kumaran: Okay.

147 00:14:37.750 ⇒ 00:14:38.280 Katherine Bayless: the desperate.

148 00:14:38.280 ⇒ 00:14:42.460 Uttam Kumaran: Yeah, I mean, that was kind of all… that was really all I had there, so I think probably over the next, like.

149 00:14:43.150 ⇒ 00:14:48.300 Uttam Kumaran: two business days, we’ll just, like, kind of stream in questions and have stuff to QA in the warehouse.

150 00:14:48.440 ⇒ 00:14:49.659 Uttam Kumaran: Okay.

151 00:14:50.110 ⇒ 00:14:53.800 Uttam Kumaran: Yeah, I mean, that was… that’s the primary update, I just want us to really focus on that.

152 00:14:54.160 ⇒ 00:14:59.319 Uttam Kumaran: And then on our side, yeah, we’re starting… we…

153 00:14:59.550 ⇒ 00:15:14.530 Uttam Kumaran: within… we use linear for sort of planning, so we’ve broken everything down by project. We’re starting to create tickets, and then I’m gonna kind of move some of those over to Asana. So I’m gonna create… I’m gonna take the tickets that we currently have assigned for each of the piece here, and I’ll put those in Asana today.

154 00:15:14.810 ⇒ 00:15:15.849 Katherine Bayless: Okay, cool.

155 00:15:17.510 ⇒ 00:15:31.630 Katherine Bayless: Yeah, I’m actually… I’m very excited at the adoption of our Asana form internally, like, when folks have been reaching out over, like, Slack and email to ask for things, and I’m like, would you mind just putting this on Asana? And they’re like, sure! And so it’s like, it’s cool to actually see that start to work.

156 00:15:31.630 ⇒ 00:15:32.320 Uttam Kumaran: Nice.

157 00:15:32.690 ⇒ 00:15:33.420 Katherine Bayless: No.

158 00:15:38.700 ⇒ 00:15:49.369 Uttam Kumaran: Cool. I mean, that’s really what I wanted to kind of discuss today. I think the other piece we could do is we can confirm a time to kind of walk through Kersher for next week. I know there were some times sent earlier.

159 00:15:51.240 ⇒ 00:15:59.140 Katherine Bayless: Yeah. Okay, I’ve got some logistical questions, and then I think the badge scan stuff is worth maybe a little bit of a dive into.

160 00:15:59.140 ⇒ 00:15:59.860 Uttam Kumaran: Okay.

161 00:15:59.860 ⇒ 00:16:01.939 Katherine Bayless: Do we want to do cursor first?

162 00:16:01.940 ⇒ 00:16:03.099 Uttam Kumaran: Yes, let’s do that.

163 00:16:04.550 ⇒ 00:16:12.520 Katherine Bayless: I’ve got my calendar up and ready. I think the times I send… I forget exactly off the top of my head, but, like, Tuesday and Wednesday afternoons are pretty open next week.

164 00:16:13.290 ⇒ 00:16:18.120 Uttam Kumaran: Okay… Let’s…

165 00:16:19.550 ⇒ 00:16:22.810 Katherine Bayless: Actually… Here.

166 00:16:23.490 ⇒ 00:16:31.940 Uttam Kumaran: Yeah, Tuesday… Yeah, Tuesday, you mentioned 11 to 1.

167 00:16:34.210 ⇒ 00:16:36.089 Uttam Kumaran: I can also do afternoons.

168 00:16:37.070 ⇒ 00:16:45.839 Katherine Bayless: Yeah, yeah, yeah, I mean, honestly, yeah, Tuesday afternoon, I have a 2 to 4, but otherwise, I’m free.

169 00:16:46.690 ⇒ 00:16:48.919 Uttam Kumaran: Okay, I could do, like, 1 to 2.

170 00:16:49.270 ⇒ 00:16:50.449 Katherine Bayless: Yep, that works.

171 00:16:50.450 ⇒ 00:16:51.110 Uttam Kumaran: Okay.

172 00:16:52.500 ⇒ 00:16:56.040 Uttam Kumaran: And I don’t know if we’ll need the whole hour, but let’s just… we’ll just stop it.

173 00:16:56.570 ⇒ 00:16:57.190 Katherine Bayless: Okay.

174 00:16:57.410 ⇒ 00:16:59.260 Katherine Bayless: Does that work for you guys, Kyle? Kai?

175 00:16:59.640 ⇒ 00:17:00.899 Chi Quinn: Yeah, that’s fine.

176 00:17:01.400 ⇒ 00:17:01.930 Katherine Bayless: Okay.

177 00:17:02.160 ⇒ 00:17:03.540 Kyle Wandel: Yeah, Tuesday’s.

178 00:17:04.400 ⇒ 00:17:05.270 Katherine Bayless: Cool, cool.

179 00:17:07.020 ⇒ 00:17:11.850 Uttam Kumaran: Okay, and then, yeah, we can… I’ll just send this now, we can talk about any logistics things.

180 00:17:12.480 ⇒ 00:17:23.320 Katherine Bayless: Okay, so, one of them is, actually, I think the call is… let me look at my… since I’m looking at my calendar, yeah, it’s, like, right after this, 11.45 today,

181 00:17:23.319 ⇒ 00:17:33.459 Katherine Bayless: the Remembers team, I guess, they have made a change on their side to the data share, and so they’re gonna connect… they just wanted to hop on a call and have us, like.

182 00:17:33.460 ⇒ 00:17:34.130 Uttam Kumaran: Okay.

183 00:17:34.320 ⇒ 00:17:42.240 Katherine Bayless: the… to the new location. I’m assuming this is an easy thing. If somebody wants to join the call, they’re welcome to, but it seems pretty straightforward. Okay.

184 00:17:42.520 ⇒ 00:17:45.409 Katherine Bayless: But, so there’s that. And then,

185 00:17:45.620 ⇒ 00:18:02.440 Katherine Bayless: sort of similarly, the AWS ProServe work for the landing zone and control tower, I think we’re gonna move forward, which will mean setting up, like, additional, AWS accounts that we’ll, like, migrate stuff into.

186 00:18:02.730 ⇒ 00:18:20.629 Katherine Bayless: How awful is that gonna be, to move Snowflake over? Like, I wasn’t sure if I should give the remembers folks a heads up today that, like, at some point in the next couple months, we will also, like, have a new account identifier and need to remap it on our side, kind of thing, or if it’s not gonna be…

187 00:18:20.630 ⇒ 00:18:24.279 Katherine Bayless: Maybe it doesn’t work that way. I’ve never migrated Snowflake before.

188 00:18:24.280 ⇒ 00:18:27.080 Uttam Kumaran: Yeah, so I don’t know, Wish, have you done, like…

189 00:18:27.270 ⇒ 00:18:31.460 Uttam Kumaran: Have we… have you done a move, Snowflake, within one cloud provider before?

190 00:18:33.680 ⇒ 00:18:37.529 Awaish Kumar: Oh, we… Haven’t thought yet, yeah.

191 00:18:38.010 ⇒ 00:18:38.789 Awaish Kumar: I wonder if I’m.

192 00:18:39.190 ⇒ 00:18:44.650 Uttam Kumaran: I’m trying to think… I feel like I’ve done it once. I think it was… we just have to coordinate with support.

193 00:18:45.070 ⇒ 00:18:46.220 Uttam Kumaran: Yeah.

194 00:18:50.490 ⇒ 00:18:53.649 Uttam Kumaran: Yeah, maybe I can find out and see.

195 00:18:55.380 ⇒ 00:19:10.600 Katherine Bayless: Yeah, I mean, we’ve got time, honestly. Okay. We’re… even if I were to, you know, tell them right now that we want to do it, you know, still take a couple weeks to provision resources, and then they’ve got to do all the, like, you know, requirements gathering, and, you know, to build out the new environment, so it’ll be…

196 00:19:10.600 ⇒ 00:19:23.270 Katherine Bayless: several months before we would be moving the Snowflake instance. Okay. And we would have the option of choosing to keep this account, like, the AWS account that Snowflake’s attached to, and just reattach it to the new AWS.

197 00:19:23.270 ⇒ 00:19:23.760 Uttam Kumaran: Yeah.

198 00:19:24.850 ⇒ 00:19:41.259 Katherine Bayless: I think if I can get one of their solutions architects to kind of go through this account with a fine-tooth comb and make me, like, feel confident that it is configured well and worthy of reconnecting, then I’ll totally take that option, because it’ll be easier than, like, migrating all the contents over, but…

199 00:19:41.920 ⇒ 00:19:50.770 Uttam Kumaran: Yeah, I mean, on our side, it’s really, like… it… because we have dbt, it doesn’t really matter. Like, we can completely rebuild it, and we’ll just…

200 00:19:51.120 ⇒ 00:19:55.009 Uttam Kumaran: All we’ll… all we’ll do is take everything in raw and move it over.

201 00:19:56.720 ⇒ 00:19:57.450 Awaish Kumar: Wow.

202 00:19:57.750 ⇒ 00:20:04.229 Uttam Kumaran: We can talk to… we’ll have to talk to Polyatomic as well, it’s like, if they can point things to new tables.

203 00:20:04.420 ⇒ 00:20:08.489 Awaish Kumar: But, like, the ingestion part, we can actually do, like.

204 00:20:08.770 ⇒ 00:20:11.859 Awaish Kumar: From Snowflake, it is very easy to move to S3.

205 00:20:12.050 ⇒ 00:20:18.779 Awaish Kumar: like, through a stage, and then that stage, actually, then the Snowflake account can read from there, so it’s…

206 00:20:18.780 ⇒ 00:20:23.799 Uttam Kumaran: Yeah, so we can just move… we can actually just move everything to S3 also.

207 00:20:24.860 ⇒ 00:20:25.650 Uttam Kumaran: Yeah.

208 00:20:27.270 ⇒ 00:20:28.259 Katherine Bayless: Oh, I see what you’re saying.

209 00:20:28.260 ⇒ 00:20:37.920 Uttam Kumaran: Well, all of this is already kind of in S3. When we move from Polytomic, like, for example, Polytomic’s Salesforce, we can just land that directly in S3.

210 00:20:38.120 ⇒ 00:20:39.029 Katherine Bayless: Right. First.

211 00:20:39.200 ⇒ 00:20:44.370 Uttam Kumaran: As, like, the first layer, and then… Again, either use…

212 00:20:44.990 ⇒ 00:20:51.259 Uttam Kumaran: this, either use Snowflake’s inbuilt, or, again, you can use polysoming to move from S3 to Snowflake.

213 00:20:51.690 ⇒ 00:20:55.830 Uttam Kumaran: So, I think that was our goal, Wish, is to have everything

214 00:20:55.950 ⇒ 00:20:58.340 Uttam Kumaran: in S3 is, like, the lake first.

215 00:20:59.190 ⇒ 00:20:59.860 Katherine Bayless: Yeah.

216 00:21:00.290 ⇒ 00:21:01.610 Katherine Bayless: Yeah.

217 00:21:01.880 ⇒ 00:21:07.429 Katherine Bayless: Okay, that’s cool. I just wanted to make sure, before I, like, go barreling down this path, that you guys are gonna be like…

218 00:21:07.430 ⇒ 00:21:10.830 Uttam Kumaran: I mean, if they can keep this, then yeah, then there’s no… that’d be great.

219 00:21:10.830 ⇒ 00:21:12.309 Katherine Bayless: Yeah, yeah, yeah. Nice.

220 00:21:13.050 ⇒ 00:21:13.590 Katherine Bayless: Yeah.

221 00:21:14.760 ⇒ 00:21:23.079 Katherine Bayless: Okay, yeah, so those are the two things. So yeah, remapping it on the Remember side, and then potential AWS account move. Okay, I know there was one other thing.

222 00:21:27.200 ⇒ 00:21:50.060 Katherine Bayless: It’ll come back to me. Okay, let’s talk about these badge scans. Okay. So, I’m learning, based on, again, the delightful willingness to use the Asana forum, that people are really interested in this data, which makes sense. It’s really kind of the closest thing we have to behavioral data coming out of CES, right? And so…

223 00:21:50.720 ⇒ 00:21:55.479 Katherine Bayless: There’s a lot of, obviously, right, intention to use it for prospecting, and so…

224 00:21:55.980 ⇒ 00:22:02.630 Katherine Bayless: I’ve been trying to kind of pull together… so, marketing team needs a couple lists, like, today, kind of thing, so I’m like.

225 00:22:02.630 ⇒ 00:22:03.010 Uttam Kumaran: Yeah.

226 00:22:03.010 ⇒ 00:22:09.890 Katherine Bayless: pulling together a skeleton approach that I think maybe you can translate into something nicer later, but,

227 00:22:09.890 ⇒ 00:22:24.789 Katherine Bayless: So, like, the first pass is just kind of going through the companies and the email domains and seeing, like, do we already know, like, this company, you know, and are they a member, of course, using those two tables that I had done the little video for, that customer link and customer alias.

228 00:22:24.870 ⇒ 00:22:41.439 Katherine Bayless: But then the next step is, like, okay, if there wasn’t a match on company name or domain deterministically, you know, historically, I think, honestly, I think historically the data team did a lot of the legwork here, but the membership team, also does some of the research and stuff.

229 00:22:42.220 ⇒ 00:22:55.500 Katherine Bayless: to me, it seems like a great use case for an enrichment provider, like, instead of having people sort of manually going through these lists of, like, you know, new companies, like, I mean.

230 00:22:55.560 ⇒ 00:23:04.760 Katherine Bayless: surely we could make an API call out and get actual, like, enrichment data back, because it’s not something that Remembers does, like, it’s not like HubSpot, where you can kind of connect a.

231 00:23:07.200 ⇒ 00:23:09.260 Katherine Bayless: Oh, sorry, go ahead, somebody.

232 00:23:11.200 ⇒ 00:23:12.420 Uttam Kumaran: I think that it would be, yeah.

233 00:23:12.420 ⇒ 00:23:29.519 Katherine Bayless: Gotcha. Yeah, so anyway, so I was… I was starting down that kind of path, and, like, you know, I think… I also took a pass through just, like, having, run the, like, new companies’ names through, like, Haiku and Bedrock, just for, like, I mean, the training data in there is massive. There’s probably.

234 00:23:29.520 ⇒ 00:23:30.130 Uttam Kumaran: Yeah.

235 00:23:30.130 ⇒ 00:23:44.869 Katherine Bayless: hanging fruit, it just knows. But then beyond that, yeah, like, maybe looking to connect an enrichment provider. And so, just kind of curious to get your thoughts. I’m sure you’ve had more experience with them than I have. I mean, I’ve used, like, Dun & Bradstreet in the past.

236 00:23:44.870 ⇒ 00:23:54.730 Katherine Bayless: You know, probably one of the more expensive, kind of bougie enrichments. But, like, if you have recommendations around, like, yeah…

237 00:23:54.730 ⇒ 00:24:03.620 Katherine Bayless: platforms, ones that tend to do better, I don’t know. And maybe considering, like, our focus is the tech industry, maybe there are ones that are, like, sort of…

238 00:24:03.820 ⇒ 00:24:05.680 Katherine Bayless: Native to that space.

239 00:24:05.730 ⇒ 00:24:12.829 Katherine Bayless: I was also thinking there’s, like, publicly available stuff, too, right? Like, we could check SEC filings, patents, that kind of stuff.

240 00:24:12.870 ⇒ 00:24:27.789 Katherine Bayless: But yeah, I’m just like, anything we can do to give the membership team, like, a little bit of a leg up on this data, because I think they really do want to start digging into it, but I feel like it’s very raw in the current state.

241 00:24:27.990 ⇒ 00:24:33.050 Uttam Kumaran: Yeah, so there’s, like, in terms of the enrichment side, yeah, I mean, we’ve worked with all the major

242 00:24:33.230 ⇒ 00:24:46.739 Uttam Kumaran: provider. So, like, so again, it kind of… it basically splits depending on the, market, like, so, kind of, like, startup, mid-market, and then enterprise. So, at the enterprise level, you have, like.

243 00:24:46.870 ⇒ 00:24:56.920 Uttam Kumaran: Zoominfo, Apollo, Clearbit, there’s some more that are new, like Owler, and a few dollars.

244 00:25:01.950 ⇒ 00:25:03.440 Kyle Wandel: I think he froze, yeah.

245 00:25:12.910 ⇒ 00:25:17.440 Katherine Bayless: I mean, to be fair, it is 6 degrees outside. Freezing seems fair.

246 00:25:19.110 ⇒ 00:25:24.389 Awaish Kumar: Yeah, I think he was trying to say that, like, our AI team already works with a lot of these providers.

247 00:25:24.390 ⇒ 00:25:24.880 Katherine Bayless: Yeah.

248 00:25:24.950 ⇒ 00:25:30.650 Awaish Kumar: And, definitely, like, they can… they will be able to plug in

249 00:25:31.360 ⇒ 00:25:37.260 Awaish Kumar: This data, the company names, and whatever we have right now, and… and enrich it.

250 00:25:37.550 ⇒ 00:25:38.250 Katherine Bayless: Yeah.

251 00:25:38.410 ⇒ 00:25:42.440 Katherine Bayless: I guess, yeah, I was kind of thinking along the lines of, like, oh, you’re back.

252 00:25:42.440 ⇒ 00:25:45.179 Uttam Kumaran: Okay, that was weird, I don’t know, my Wi-Fi just cut it.

253 00:25:46.450 ⇒ 00:25:47.020 Katherine Bayless: Oh.

254 00:25:49.670 ⇒ 00:25:50.770 Awaish Kumar: Nothing.

255 00:25:56.350 ⇒ 00:25:56.930 Katherine Bayless: Thank.

256 00:25:59.370 ⇒ 00:26:05.139 Katherine Bayless: Alright, but yeah, I was kind of thinking along the lines of, like,

257 00:26:06.520 ⇒ 00:26:17.210 Katherine Bayless: Yeah, bringing in something like that, and then, ideally, like, as we establish new, like, you know, company name aliases and domains, we’re kind of capturing those, and…

258 00:26:17.220 ⇒ 00:26:25.900 Katherine Bayless: parking them, back into remembers so that that data then comes through for the next round of, matching. And we have pretty, like.

259 00:26:25.900 ⇒ 00:26:46.179 Katherine Bayless: I think LLM-able criteria for membership. It has to have a North American business address, like an office somewhere in North America, and has to be, you know, in the tech space or adjacent to it. It can’t be, like, a government university, that kind of thing. So, like, I feel like this feels pretty achievable as a pipeline.

260 00:26:46.180 ⇒ 00:26:59.810 Katherine Bayless: just kind of curious, like, cost-wise, what we’d be in for, because it isn’t necessarily in the budget, but at the same time, I feel like these services don’t tend to be, like, prohibitively expensive, usually. Again, Dun & Bradstreet being the exception.

261 00:27:00.650 ⇒ 00:27:06.489 Awaish Kumar: Yeah, like… like, I think, like, AIT much… will be much more…

262 00:27:06.610 ⇒ 00:27:15.909 Awaish Kumar: have more knowledge on it to give you concrete answers, but I think, yeah, like, we have explored a lot of these tools, and some of them are…

263 00:27:16.820 ⇒ 00:27:23.729 Awaish Kumar: Expensive, but, like, none of them are, like, per day in range, and you can utilize those to get this data.

264 00:27:24.480 ⇒ 00:27:30.760 Katherine Bayless: Okay. Yeah, because then I think what we could probably do, because, like, thinking ahead to the entity resolution stuff, too, is, like.

265 00:27:30.890 ⇒ 00:27:42.010 Katherine Bayless: if we had a sort of a listener kind of a concept, right? And so anytime a new company or email domain record is turning up in, you know, our data, then we’re just, like, by default, sending it out, running it, checking it.

266 00:27:42.180 ⇒ 00:27:52.509 Katherine Bayless: bringing it back. I mean, right now, there’s a need for these, like, massive files to be checked, but once those are done on an ongoing basis, I think, yeah, like, a listener makes sense to me.

267 00:27:53.150 ⇒ 00:27:56.640 Awaish Kumar: So, like, is… remember’s platform, like, do…

268 00:27:57.980 ⇒ 00:28:01.309 Awaish Kumar: Yeah, I think… is there an API for that?

269 00:28:02.320 ⇒ 00:28:24.989 Katherine Bayless: Yeah, actually, so, no. But they do have Power Automate, and so, yeah, like, piping data programmatically back into remembers is still a bit of a question mark for me to figure out. They do have some APIs, but they’re not very robust. Like, in fact, the reason they pushed us in the direction of the data share is because that’s the only way to really get everything. The APIs are more limited.

270 00:28:24.990 ⇒ 00:28:39.499 Katherine Bayless: But they do have this Power Automate connector, and if all else fails, there is, like, a manual upload kind of concept on the, like, UI side, where we could put CSVs into the database that way.

271 00:28:39.720 ⇒ 00:28:57.199 Awaish Kumar: I was asking because, like, we do a lot of these reverse ETL kind of thing, and we have the polyatomic now, so we could actually ask them to, like, if there is an API, we could ask them, let’s build a connector for us, so it can upload automatically from our system to the members.

272 00:28:57.240 ⇒ 00:29:03.549 Awaish Kumar: But yeah, if that is not the case, then yes, we have to… Do the manual first.

273 00:29:03.950 ⇒ 00:29:05.590 Katherine Bayless: Yeah, okay, okay.

274 00:29:06.060 ⇒ 00:29:06.820 Katherine Bayless: Okay.

275 00:29:07.250 ⇒ 00:29:16.910 Uttam Kumaran: Okay, I don’t know, sorry, I don’t know, my… the 5G band of my router is not working, something weird happened, but I guess I’ll just stay off video in case that doesn’t… that does anything here.

276 00:29:17.320 ⇒ 00:29:26.440 Uttam Kumaran: But, yeah, it kind of goes in and out, but I think, yeah, so there is a world of, of sort of vendor potential.

277 00:29:26.470 ⇒ 00:29:39.139 Uttam Kumaran: I think, Catherine, you’re kind of right in that there are expensive ones, like Dun & Bradstreet ZoomInfo are, like, sort of the most expensive. You also have ones that are sort of up and coming, and then you have ones that are really, really, like.

278 00:29:39.690 ⇒ 00:29:49.889 Uttam Kumaran: they’re focused on one type. So, in this mid-market, you have these guys, there’s, there’s, there’s, like, new ones like Captain Data, People Data Labs.

279 00:29:50.540 ⇒ 00:29:57.300 Uttam Kumaran: And so… and then in the startup world, this is really, like, Crunchbase…

280 00:29:57.460 ⇒ 00:30:07.609 Uttam Kumaran: And, like, there’s a company called Harmonic that also does really, really well. So, we actually… we work… we work with another company, and they’re,

281 00:30:07.610 ⇒ 00:30:17.639 Uttam Kumaran: They are an inbound lead, like, calendar scheduling tool, and they have a lot of enrichment, so we actually help them basically evaluate almost, like, 15 different vendors.

282 00:30:17.770 ⇒ 00:30:29.700 Uttam Kumaran: So, I do… we just finished this, like, last month, which is why I’m, like, able to rattle all these off, because these are, like… some of these are kind of obscure. So, we have the pricing on each of these, and…

283 00:30:30.550 ⇒ 00:30:37.240 Uttam Kumaran: We actually, for them, we… we created a control dataset, and we basically use all of them to run

284 00:30:37.460 ⇒ 00:30:43.520 Uttam Kumaran: On, like, a controlled data set of 400 companies that were in each of these segments to basically show the performance.

285 00:30:43.660 ⇒ 00:30:47.309 Uttam Kumaran: And so, some of them definitely stand out, depending, and, like.

286 00:30:47.670 ⇒ 00:30:53.160 Uttam Kumaran: For the most part, I don’t know if we… I guess I’ll… I’ll… I’ll kind of learn more as I go through

287 00:30:54.850 ⇒ 00:30:55.650 Uttam Kumaran: Red nut.

288 00:30:55.920 ⇒ 00:31:01.320 Uttam Kumaran: I assume we’re, like, it’s gonna be mo- it’s gonna be… Everywhere, maybe? I don’t know.

289 00:31:02.800 ⇒ 00:31:09.560 Katherine Bayless: Yeah, I mean, I think, at least in the sort of, like, you know, sandboxy approach I was looking at it, it’s like…

290 00:31:10.720 ⇒ 00:31:35.649 Katherine Bayless: because of our, like, the, like, Eureka Park and whatnot, like, there are a lot of startups that’s, like, you know, really generic company name and a Gmail address. Like, I think there’s some stuff we just are gonna have to, you know, accept defeat on, but then I think probably the ones that are sort of more established startups, we would probably be able to get data on this way. But yeah, I mean, like, some of them are, like, web technologies, and it’s like, contact me at gmail.com, you know, and you’re like.

291 00:31:35.650 ⇒ 00:31:40.080 Katherine Bayless: okay, well, I’m never gonna figure out who you are. Yes.

292 00:31:40.110 ⇒ 00:31:42.310 Katherine Bayless: But yeah. Because the other thing.

293 00:31:42.310 ⇒ 00:31:45.600 Uttam Kumaran: And do you know, like, what… Yeah, go, I got it.

294 00:31:45.600 ⇒ 00:31:46.440 Katherine Bayless: No, no, go ahead.

295 00:31:47.790 ⇒ 00:31:54.640 Uttam Kumaran: Do you know what fields, like, in particular they’re interested in? Like, is it… it’s, like, kind of classic, like, demographic fields?

296 00:31:54.820 ⇒ 00:31:58.340 Uttam Kumaran: Yeah, I’m just kind of curious, like, what some of the use cases could be.

297 00:31:58.570 ⇒ 00:32:16.640 Katherine Bayless: Yeah, so I think, like, the initial, like, baseline is just kind of, like, evaluating eligibility for membership, so they have to have a North American office, and they have to be either, like, in the tech industry or supporting it, so, like, we would want to be able to, like, weed out, like, you know, governments and universities and stuff like that.

298 00:32:16.640 ⇒ 00:32:39.790 Katherine Bayless: Hospitals are kind of in a gray area, because, like, healthcare, there are some lanes that they can be members in. But yeah, so I think, like, First Pass would just be using it to evaluate that North American presence and sort of, like, sector affiliation. But then, our revenue, or excuse me, our dues are calculated based on North American revenue, so, like, if any of the services can return, kind of, like, you know.

299 00:32:39.790 ⇒ 00:32:51.850 Katherine Bayless: ballpark figures in that arena, then I think, you know, the membership could obviously do a, like, sort by biggest companies and focus on them first, right? Actually, also, it would be good to,

300 00:32:51.850 ⇒ 00:33:01.569 Katherine Bayless: bring in the, like, entity list checking stuff. I know that’s a project Kyle’s bringing over from market research, because we obviously want to weed out anybody who’s on a bad list.

301 00:33:01.570 ⇒ 00:33:16.329 Katherine Bayless: But yeah, so it’s like, I don’t… I don’t know that we nearly need much beyond the North American address and a revenue range initially. I mean, there’s tons of things we can get from these, but I… I think if we start there, that would be still a game changer.

302 00:33:16.330 ⇒ 00:33:21.850 Kyle Wandel: Another big… I don’t know if you said a dataset that you can kind of access is PitchBook. PitchBook is another big one.

303 00:33:21.850 ⇒ 00:33:23.120 Uttam Kumaran: Yes, yeah.

304 00:33:23.500 ⇒ 00:33:24.910 Katherine Bayless: Yeah, reminder, yeah.

305 00:33:25.880 ⇒ 00:33:31.479 Kyle Wandel: We use that on the market research team for all of our index creations, basically.

306 00:33:32.240 ⇒ 00:33:33.760 Katherine Bayless: Yeah, that’s a good reminder.

307 00:33:35.500 ⇒ 00:33:43.449 Uttam Kumaran: Okay, then yeah, I mean, if some of these you already have access to, I mean, of course, I think even for just address and some of those, this… we should totally just leverage that.

308 00:33:43.450 ⇒ 00:33:56.250 Uttam Kumaran: And then we… I can sort of come back with some examples of some of these in our… maybe our… the control data set that we put together, because this will kind of show what data we can get. I mean, a lot of these you can get really quite advanced, so some of these…

309 00:33:56.510 ⇒ 00:34:04.359 Uttam Kumaran: For example, our client, they were interested in, like, we want to look at a person, and if they’ve changed jobs, like, what was their last job, and, like, what job did they change?

310 00:34:04.380 ⇒ 00:34:21.970 Uttam Kumaran: So some of these are very particular to, like, we look at people’s job changes, we’re looking at how many sale… like, has this company hired? What is their growth rate in their sales organization last month? Like, really sophisticated enrichment, and then some of these are just, like, how much money did they raise?

311 00:34:22.110 ⇒ 00:34:26.650 Uttam Kumaran: Yeah, where are they based, who is… who’s on the core executive team, things like that.

312 00:34:27.110 ⇒ 00:34:49.769 Katherine Bayless: Yeah, I would love to get to the, like, the personnel moves data at some point, because I, I mean, I have this hypothesis, right, that, like, there is probably a lot of horse trading that occurs at CES. It’s, like, the world’s largest unofficial job fair, right? And so I would love to track, like, growth in different, you know, organizations and teams, like, you know, before and after CES, like, if we could kind of figure out, like.

313 00:34:49.989 ⇒ 00:34:57.030 Katherine Bayless: you know, how do we move the needle on some of these, like, job roles? That’d be kind of cool. So, I’d like to get there.

314 00:34:57.030 ⇒ 00:34:57.610 Uttam Kumaran: Great.

315 00:34:57.760 ⇒ 00:34:58.880 Uttam Kumaran: Great, okay.

316 00:34:59.310 ⇒ 00:35:10.259 Uttam Kumaran: Yeah, so I think part of this is just gonna be us seeing, like, okay, like, there’s… if some of these are wrong, and, like, okay, can we… can we rely on, sort of, some type of waterfall enrichment, basically?

317 00:35:10.260 ⇒ 00:35:10.830 Katherine Bayless: Yeah.

318 00:35:10.830 ⇒ 00:35:20.139 Uttam Kumaran: So that’s typically what we do. A lot of folks will have their own data, as well as one or two of these, and we’ll kind of do, like, a waterfall, depending on the field.

319 00:35:23.320 ⇒ 00:35:25.279 Uttam Kumaran: So, that’s… that’s great. Okay.

320 00:35:25.520 ⇒ 00:35:35.800 Katherine Bayless: Okay, so then on the logistics of the actual data set, so, like, Kyle was talking about the… there’s a little bit of a challenge around some of the session codes not being quite in ideal shape.

321 00:35:35.800 ⇒ 00:36:00.789 Katherine Bayless: And there is a request that I am inclined to, honor, especially since we’re trying to, you know, get people excited about data, to integrate the, like, things that were not, like, necessarily, like, part of the conference’s stuff, but it happened at CES, like the LIT dinner registrations. Like, they want to see the LIT attendance coming through essentially as if it were badge scan days.

322 00:36:00.790 ⇒ 00:36:10.349 Katherine Bayless: Which I get. It’s… I mean, like, we think about this as badge scan data today, because that’s what it’s been, but, like, the Cvent stuff is also, like, attendance data at events at CES.

323 00:36:10.370 ⇒ 00:36:26.310 Katherine Bayless: I think, also, my hope is that next year we are collecting this data totally differently, and ideally all in one place, but I do think it’s worth trying to patch in the handful of things that were done in Cvent, so I think it’s probably a matter of just kind of, like.

324 00:36:26.330 ⇒ 00:36:34.469 Katherine Bayless: shaping it to look enough like this data, and then unioning it, but I will leave that to the, data wizards.

325 00:36:35.000 ⇒ 00:36:36.610 Uttam Kumaran: Okay, okay, perfect.

326 00:36:38.110 ⇒ 00:36:43.360 Katherine Bayless: Yeah, I think those are all the things that were on top of my head.

327 00:36:44.380 ⇒ 00:36:45.529 Katherine Bayless: For badges.

328 00:36:46.790 ⇒ 00:36:47.390 Uttam Kumaran: Okay.

329 00:36:49.080 ⇒ 00:36:49.630 Katherine Bayless: Yeah.

330 00:36:51.680 ⇒ 00:36:57.599 Uttam Kumaran: Cool. Any other, like, dataset we want to just, like, poke at before… Calling it?

331 00:37:01.940 ⇒ 00:37:08.249 Katherine Bayless: I mean, I guess, actually, I had a question around just, like, logistics for the,

332 00:37:08.520 ⇒ 00:37:15.509 Katherine Bayless: So, like, right now, the Remember stuff that’s been built, like, that has the DIM organization, right,

333 00:37:16.030 ⇒ 00:37:17.989 Katherine Bayless: Would we want to…

334 00:37:18.750 ⇒ 00:37:37.070 Katherine Bayless: So, like, those are going to be all of the organizations that are in remembers and coming out that way, but then, kind of, I guess, going back to the enrichment and prospecting piece, like, we also have a bunch of companies that aren’t yet in remembers, and that are going to be in these different CES files, and, like, do we want to use DIM organization for all of them, or do we want to have

335 00:37:37.070 ⇒ 00:37:44.170 Katherine Bayless: that sort of become, you know, just the Remembers one, and then there’s a different one where we’re capturing organizations coming out of CES data.

336 00:37:44.660 ⇒ 00:37:52.999 Uttam Kumaran: I think we’ll… yeah, I guess that’s a good question. We may end up with one, and then putting, like, a source field, so you could filter.

337 00:37:53.300 ⇒ 00:38:01.930 Uttam Kumaran: Ultimately, like, I think having one allows us to, like, leverage the data from both sides, and it’s… we just create a wide table with, like.

338 00:38:02.310 ⇒ 00:38:03.550 Uttam Kumaran: their metrics.

339 00:38:03.850 ⇒ 00:38:09.839 Uttam Kumaran: And then any sort of mapping table, we will also sort of create as, like, a map between

340 00:38:10.160 ⇒ 00:38:10.950 Uttam Kumaran: like…

341 00:38:11.230 ⇒ 00:38:22.050 Uttam Kumaran: different IDs or things like that, and that way, even if some of that has to get built manually, there’s a process where we can upload a CSV, and, like, some of those maps can get done.

342 00:38:22.710 ⇒ 00:38:30.189 Uttam Kumaran: And then, that way, over time, like, yeah, some of these things, maybe it’s not going to be clear, the mapping and the data, so we have to have a manual intervention.

343 00:38:30.320 ⇒ 00:38:33.699 Uttam Kumaran: But yeah, I feel like it’s best to have…

344 00:38:33.790 ⇒ 00:38:52.720 Uttam Kumaran: either we can go one or two ways. We can have two of them, and then they kind of… we basically merge into the, like, canonical DIMM organizations, and they’re built off of two other DIM organizations, or we just build one, and then you can just filter to the source and pull it out.

345 00:38:53.570 ⇒ 00:38:58.580 Uttam Kumaran: it’s kind of like just two flavors of ice cream, I think, basically.

346 00:38:59.180 ⇒ 00:39:07.369 Katherine Bayless: Yeah, I mean, to me, the combined approach probably makes the most sense, especially knowing the, like, entity resolution stuff that we wanted to.

347 00:39:07.370 ⇒ 00:39:08.140 Uttam Kumaran: Yeah.

348 00:39:08.460 ⇒ 00:39:12.530 Katherine Bayless: We’re all working off the same backbone, I think that makes the most sense.

349 00:39:12.720 ⇒ 00:39:13.370 Uttam Kumaran: Yes.

350 00:39:15.520 ⇒ 00:39:16.200 Katherine Bayless: Okay.

351 00:39:16.560 ⇒ 00:39:17.519 Uttam Kumaran: I think so, too.

352 00:39:17.520 ⇒ 00:39:20.489 Katherine Bayless: Yeah, and so… and that way, it’s all in one place, and then…

353 00:39:20.550 ⇒ 00:39:24.330 Uttam Kumaran: Ideally, it’ll be easier also to see, like, duplicates.

354 00:39:24.440 ⇒ 00:39:29.879 Uttam Kumaran: And then be like, okay, why is there a duplicate? Which… which is, like, okay, they’re two different addresses…

355 00:39:29.880 ⇒ 00:39:31.610 Katherine Bayless: Two different emails…

356 00:39:31.730 ⇒ 00:39:37.160 Uttam Kumaran: like, and then, I guess, seeing it all in that place allows us to sort of have those conversations.

357 00:39:38.120 ⇒ 00:39:58.320 Katherine Bayless: Yeah, I mean, yeah, I think… yeah, it’s… it’s gonna be a challenge across the board, but, like, I think the companies will be easier than the people will, like, once we start trying to actually figure out, like, who are these individuals, and are they, you know, same person with different, you know, information in different places, and then which information is more recent, and should we trust more, et cetera, et cetera? Like, yeah, I think the…

358 00:39:58.320 ⇒ 00:40:02.940 Katherine Bayless: Entity resolution on the humans is gonna be a deep rabbit hole.

359 00:40:02.940 ⇒ 00:40:03.600 Uttam Kumaran: Yes.

360 00:40:04.350 ⇒ 00:40:05.070 Katherine Bayless: Yeah.

361 00:40:05.490 ⇒ 00:40:08.200 Uttam Kumaran: Yeah, that’s gonna be a bit tougher.

362 00:40:09.960 ⇒ 00:40:10.860 Katherine Bayless: Yeah.

363 00:40:12.160 ⇒ 00:40:24.119 Katherine Bayless: But yeah, I mean, I guess if there’s any other… like you said, like, if there’s any other tables that you want to talk about in here, maybe… or actually, honestly, Kyle, too, if you’ve got any, like, oh, I know you’re gonna run into questions on this!

364 00:40:24.940 ⇒ 00:40:26.109 Kyle Wandel: Thank you.

365 00:40:26.110 ⇒ 00:40:27.200 Katherine Bayless: Jumping out at me.

366 00:40:27.200 ⇒ 00:40:45.769 Kyle Wandel: Yeah, nothing on the top of my head right now. I think the goal goes back to, like, A, how can I help? So, like, in terms of what tables or what models can I help build, or what logic do you guys need, or is it more so that you guys need logic for it? It’s like, how can I help, kind of, help the process in general? And then, honestly, just figuring out this goddang

367 00:40:45.970 ⇒ 00:40:48.110 Kyle Wandel: Badge scan stuff.

368 00:40:48.280 ⇒ 00:40:48.800 Uttam Kumaran: Yeah.

369 00:40:48.860 ⇒ 00:41:01.169 Kyle Wandel: Because I think it’s, like, one of those things where they want us to clean the data, like, what is the likely session code based on the session time and date? But even then, I mean, looking… I can try to look through, but even looking through, I don’t want to, like…

370 00:41:01.330 ⇒ 00:41:06.459 Kyle Wandel: miscategorize something that’s just not correct, but, we’ll see.

371 00:41:07.690 ⇒ 00:41:11.510 Uttam Kumaran: Yeah, I think we’ll probably have more for you later today or Monday on questions.

372 00:41:13.630 ⇒ 00:41:23.810 Uttam Kumaran: We’re gonna… like, part of this is also just, like, trying to build a foundation for some of this reporting, and then as we get some of these to be live, we’ll just slot in the live data.

373 00:41:23.970 ⇒ 00:41:29.400 Uttam Kumaran: So we’re gonna kind of build… start to build some of the marts around, like, Salesforce.

374 00:41:29.530 ⇒ 00:41:41.359 Uttam Kumaran: some of the CES work, and then building… and the first customer of it is gonna be the… this Power BI report, right? So, we’re trying not as much to not build it just bespoke for this.

375 00:41:41.620 ⇒ 00:41:49.999 Uttam Kumaran: And, like, go through the raw, intermediate marts, like, process. But, like, a lot… I think when… as we ship PRs, I think

376 00:41:50.110 ⇒ 00:41:53.799 Uttam Kumaran: Kyle, I’ll just send them to you to review, so you can take a look at the logic.

377 00:41:54.770 ⇒ 00:41:55.380 Kyle Wandel: Pretty good, yeah.

378 00:41:55.380 ⇒ 00:42:00.099 Uttam Kumaran: Once we… once they’re in review, they’ll get created,

379 00:42:00.570 ⇒ 00:42:08.879 Uttam Kumaran: they’ll get created here, so you can actually go look at the PR and then see them in here as well, so you can go, like, basically query and propose changes.

380 00:42:09.110 ⇒ 00:42:12.980 Kyle Wandel: Yep, and that’s in which… and that’s gonna be in the dev, one of the dev, right?

381 00:42:12.980 ⇒ 00:42:19.380 Uttam Kumaran: Yeah, so we’ll be… yeah, we’ll be working in dev, and then it should get… it’ll get created in.

382 00:42:19.380 ⇒ 00:42:20.220 Kyle Wandel: Staging.

383 00:42:20.220 ⇒ 00:42:21.660 Uttam Kumaran: Staging, I believe, yeah.

384 00:42:22.970 ⇒ 00:42:37.319 Kyle Wandel: Okay, cool. And then if you guys… I know Ashrini’s not on, but any documentation on terms of, like, I know I’ve asked this, like, 3 times, like, because I’m still trying to understand it, like, the between… the differences between STG, staging, underscore marts, dev, underscore march.

385 00:42:37.320 ⇒ 00:42:38.060 Uttam Kumaran: Yeah.

386 00:42:38.060 ⇒ 00:42:49.009 Kyle Wandel: any documentation on that, maybe, like, in the README for the GitHub would be great. Okay. That would help me a lot more, because I’ve… I know I’ve done some naming conventions that aren’t…

387 00:42:49.380 ⇒ 00:42:53.180 Kyle Wandel: I don’t want to make sure it’s all uniform, you know what I mean? Okay. So…

388 00:42:53.720 ⇒ 00:43:02.509 Kyle Wandel: that’s my biggest thing, is making sure the naming conventions, whenever I create a repo, whenever I create a schema, I don’t want to create something that’s not being used going forward.

389 00:43:02.860 ⇒ 00:43:08.490 Uttam Kumaran: Okay, I’ll… I’ll put that… I’ll just put… put our doc in the repo itself and send it to you.

390 00:43:09.080 ⇒ 00:43:23.810 Katherine Bayless: Yeah. Actually, I mean, honestly, I would… I would benefit from that too, because I was kind of like, okay, wait, what is the life cycle of something going from, like, we need this to we have this? And I was like, I got kind of amused by staging, staging, but yeah.

391 00:43:23.810 ⇒ 00:43:24.480 Kyle Wandel: Yeah.

392 00:43:26.110 ⇒ 00:43:28.859 Katherine Bayless: shoot, there was… oh, goddamn.

393 00:43:29.290 ⇒ 00:43:42.970 Katherine Bayless: Oh, I was just gonna say, yeah, like, once… so, like, the… the stuff that we’re building for the membership engagement report, like, once we have those pieces with the CES data, the exhibits, the committees, the innovation awards, like, those are the… the…

394 00:43:42.970 ⇒ 00:44:06.669 Katherine Bayless: biggest things that people are asking for. Like, I think, honestly, once they’re in place, we’ll be able to really fly through a lot of other stuff. Okay. Because it winds up just kind of being, like, you know, every team’s got a different, you know, sort of use case and lens, but they’re all asking for the same core data. They want to know, like, who was at CES, who had an exhibit booth, right? Like, that kind of stuff. And so, I think this is the

395 00:44:06.980 ⇒ 00:44:12.709 Katherine Bayless: The main components of the foundation, and then everything else will be us getting to kind of, like, add things on.

396 00:44:13.120 ⇒ 00:44:14.010 Uttam Kumaran: Okay, okay.

397 00:44:15.160 ⇒ 00:44:15.870 Uttam Kumaran: Okay.

398 00:44:16.170 ⇒ 00:44:21.379 Uttam Kumaran: Perfect. Yeah, I have, maybe I’ll just send this… I’ll just send you guys this,

399 00:44:22.200 ⇒ 00:44:33.020 Uttam Kumaran: this Notion site that we have, which has, like, kind of a little bit of information on how we structure dbt, and then I’m gonna… I’ll put in a, like, a CTA-focused one into the repo today.

400 00:44:33.460 ⇒ 00:44:34.660 Katherine Bayless: Okay.

401 00:44:34.660 ⇒ 00:44:39.149 Uttam Kumaran: So there’s something, Just so there’s something to read,

402 00:44:39.290 ⇒ 00:44:44.250 Uttam Kumaran: But, like, this is also kind of evolving as we do a bunch of these.

403 00:44:44.250 ⇒ 00:44:49.950 Katherine Bayless: So, we’re also having internal debates about, like, okay, do we need another layer? But this… this’ll give you a lot of.

404 00:44:50.030 ⇒ 00:44:54.460 Uttam Kumaran: Insight into, like, how we… Are doing the entire workflow.

405 00:44:54.860 ⇒ 00:45:05.149 Uttam Kumaran: And there are some… I’ll also share some links on… on just, like, how… basically how we arrived here. Like, there’s some other companies that we sort of followed to arrive here, so…

406 00:45:06.170 ⇒ 00:45:19.240 Katherine Bayless: Cool. The other thing on that same note is the… the RBAC stuff. So I did get Ian to add all the membership team folks that we’re working with in this sort of phase into Seablake, and so if you can let me know, like.

407 00:45:19.960 ⇒ 00:45:27.579 Katherine Bayless: in order for them to see a dashboard or Streamlit app that is built on top of ProdMarts, what permissions I need to assign them.

408 00:45:27.890 ⇒ 00:45:31.119 Uttam Kumaran: Great. So yeah, at the moment, you know, we have…

409 00:45:31.320 ⇒ 00:45:38.160 Uttam Kumaran: basically read-write permissions for each stage. Like, would you… would it be helpful to create, like, a…

410 00:45:38.410 ⇒ 00:45:43.960 Uttam Kumaran: We could create a streamlit Reader or streamlet viewer role.

411 00:45:44.300 ⇒ 00:45:51.400 Uttam Kumaran: that… Is able to see… basically any of the marts tables and Streamlit.

412 00:45:51.940 ⇒ 00:45:56.450 Uttam Kumaran: If that’s fine, that’s probably, like, what I would propose.

413 00:45:56.570 ⇒ 00:46:04.319 Uttam Kumaran: And then in terms of, like, where it lives, I guess there… I assume that it’s gonna be in Streamlit AppsDB.

414 00:46:05.040 ⇒ 00:46:11.690 Uttam Kumaran: That’s also fine, it doesn’t necessarily need to be in… in… in Prodmarts, like…

415 00:46:12.330 ⇒ 00:46:15.190 Katherine Bayless: I think I had set that up because.

416 00:46:15.190 ⇒ 00:46:15.620 Uttam Kumaran: Okay.

417 00:46:15.620 ⇒ 00:46:20.220 Katherine Bayless: trying to put that little Streamlit app together as that proof of concept. Feel free to delete it and…

418 00:46:20.220 ⇒ 00:46:29.749 Uttam Kumaran: Then I would suggest we put the Streamlit apps all in ProdMarts, because then we’ll give Reed access to this, and so people can not only go see the raw data.

419 00:46:29.860 ⇒ 00:46:36.469 Uttam Kumaran: They can also… go check out any of the Streamlit apps that live underneath here. So I can just…

420 00:46:37.190 ⇒ 00:46:39.299 Uttam Kumaran: I can basically just move these over.

421 00:46:40.330 ⇒ 00:46:42.919 Katherine Bayless: Okay. I mean, obviously, you could delete the little thing I built.

422 00:46:42.920 ⇒ 00:46:43.510 Uttam Kumaran: Okay.

423 00:46:43.510 ⇒ 00:46:45.900 Katherine Bayless: offend me, and just start fresh, but yeah.

424 00:46:45.900 ⇒ 00:46:46.460 Uttam Kumaran: Okay.

425 00:46:46.610 ⇒ 00:46:47.340 Katherine Bayless: Makes sense.

426 00:46:47.660 ⇒ 00:46:54.630 Uttam Kumaran: Okay, so maybe a waysh, on our side today, let’s just, like, let’s just clean some of this up, like, let’s get rid of some of the snowflake stuff, and then…

427 00:46:54.880 ⇒ 00:47:02.119 Uttam Kumaran: we can… we’ll create a role for… I guess at this point, Catherine, does, like, is access to certain data sets, like…

428 00:47:02.570 ⇒ 00:47:05.320 Uttam Kumaran: Is there any, like, risk?

429 00:47:05.480 ⇒ 00:47:08.870 Uttam Kumaran: That we’re… she’ll be aware of, or we can basically say, like, everyone…

430 00:47:09.800 ⇒ 00:47:11.819 Uttam Kumaran: Gets access to what we’re producing here.

431 00:47:12.650 ⇒ 00:47:28.020 Katherine Bayless: Yeah, I mean, I think at this point, everything in here is not, like, sensitive in a way that we would need to be more careful about. I think eventually we might start having stuff that’s, like, not sensitive, but just kind of, like, irrelevant to a lot of people. Like, I could see there being that kind of use case, but…

432 00:47:28.020 ⇒ 00:47:35.710 Uttam Kumaran: So in that situation, what we would do is we’d create, basically, like, role finance viewer, role sales viewer.

433 00:47:36.010 ⇒ 00:47:40.649 Uttam Kumaran: That… that would basically get certain read and write permissions.

434 00:47:40.850 ⇒ 00:47:54.440 Uttam Kumaran: And so we can continue to… we can layer that in. As long as, like, we sort of have these base roles, and then we’re… we’re… we may end up having, like, role finance read, right? And then we roll that up into, like.

435 00:47:54.820 ⇒ 00:47:59.700 Uttam Kumaran: us, or… Any, any sort of viewer, you know, things like that.

436 00:48:00.280 ⇒ 00:48:02.239 Katherine Bayless: Yeah, that makes sense. That makes sense.

437 00:48:02.240 ⇒ 00:48:04.680 Uttam Kumaran: We can put… I’ll put that into the,

438 00:48:05.340 ⇒ 00:48:15.860 Uttam Kumaran: I think we might… we already have our script moved in, or I’ll move in our, like, the RBAC script that we ran in the beginning into there, and then we can continue to update that.

439 00:48:16.050 ⇒ 00:48:17.210 Uttam Kumaran: As we go.

440 00:48:18.400 ⇒ 00:48:18.990 Katherine Bayless: Cool.

441 00:48:19.240 ⇒ 00:48:29.370 Katherine Bayless: And then in terms of, like, granting the permissions, I mean, I just, like, it’s going to the user and giving them that role, like, it doesn’t need to be, like, layered in any way.

442 00:48:31.740 ⇒ 00:48:36.169 Uttam Kumaran: Yeah… Awash, I don’t know, what do you have? Do you have any thoughts?

443 00:48:40.810 ⇒ 00:48:41.920 Katherine Bayless: Oh, you’re on mute.

444 00:48:44.640 ⇒ 00:48:50.049 Awaish Kumar: Yeah, I mean, like, we can create roles and assign to the people.

445 00:48:52.180 ⇒ 00:48:54.470 Awaish Kumar: I… I don’t think we need any…

446 00:48:55.610 ⇒ 00:48:58.069 Awaish Kumar: Like, extra thing here, like, it’s…

447 00:48:58.350 ⇒ 00:49:04.479 Awaish Kumar: Very easy to create some of the roles and just assign to the people who needs what access.

448 00:49:04.780 ⇒ 00:49:06.819 Awaish Kumar: Just based on that, yep.

449 00:49:07.240 ⇒ 00:49:22.270 Katherine Bayless: Okay, okay. Yeah, I wasn’t sure if there was, like, the, like, the both sides of the equation kind of thing, where it’s like, we need to, like, give the user the role, but then also say that prod marts can be viewed by that role. Like, I wasn’t sure if there was kind of, like, the double entry concept for Snowflake permissions.

450 00:49:22.950 ⇒ 00:49:33.470 Awaish Kumar: Yeah, like, that’s how it works, like, for a user, we just need to define the roles, like, what a finance… what a person in the finance team can access.

451 00:49:33.680 ⇒ 00:49:38.279 Awaish Kumar: We provide the access to those tables and databases to that role.

452 00:49:38.520 ⇒ 00:49:41.580 Awaish Kumar: And then we just assign each person to

453 00:49:41.890 ⇒ 00:49:47.370 Awaish Kumar: Each person from finance team to just that room only, so they can access that database.

454 00:49:48.000 ⇒ 00:49:55.520 Katherine Bayless: Okay, so if we start really simple, then we would just have the, like, roll prod read that we would give

455 00:49:55.950 ⇒ 00:49:57.689 Katherine Bayless: To these folks in membership, so that.

456 00:49:57.690 ⇒ 00:49:58.440 Uttam Kumaran: Yes.

457 00:49:59.040 ⇒ 00:50:00.490 Awaish Kumar: It’s a student, like…

458 00:50:01.250 ⇒ 00:50:02.209 Uttam Kumaran: Yeah, go ahead, Oish.

459 00:50:02.490 ⇒ 00:50:16.739 Awaish Kumar: Yeah, it’s still, like, this is what we set up initially, like, as part of, setting up Snowflake, but we can be more granular, because in prod, Mart, there could be a lot of things.

460 00:50:16.840 ⇒ 00:50:22.460 Awaish Kumar: They… we will create a mod for finance team, right?

461 00:50:22.660 ⇒ 00:50:30.350 Awaish Kumar: In the prod march, they have access to ProdMarts, but then only one schema, which is relevant to finance team.

462 00:50:31.050 ⇒ 00:50:43.300 Katherine Bayless: Right. Yeah, I think for now, we can stay broad. I don’t think we need to subspecify by departments. I think eventually we will maybe get there, but yeah, for the moment, I think prod is prod. It’s fine.

463 00:50:43.300 ⇒ 00:50:50.860 Uttam Kumaran: Yeah, and the reason to, like, we’re doing this layering is because otherwise, it’s gonna get, like, you may want to say, oh, actually, like.

464 00:50:51.060 ⇒ 00:50:56.870 Uttam Kumaran: finance teams should have access to what the sales teams have. Okay, let’s just, like, layer… the…

465 00:50:57.020 ⇒ 00:51:01.129 Uttam Kumaran: the Sales Mart reader writer roles into finance, right? And so.

466 00:51:01.320 ⇒ 00:51:16.789 Uttam Kumaran: The… keeping this, like, these, like, read-write primitives at the bottom allows you to just, like, then create, basically, groups at the top that are based on the business unit, versus, like, where this can get really difficult is if you’re like, oh, we have, like, finance

467 00:51:17.090 ⇒ 00:51:23.989 Uttam Kumaran: Finance team, and then you just run, like, oh, they can read this table, they can run… they can read this schema, they can…

468 00:51:24.110 ⇒ 00:51:33.829 Uttam Kumaran: add to this, and then it’s, like, it’s sort of horrible to manage. So, we’re, like, environment-based or MART-based reader-writer roles, and then grant those up.

469 00:51:34.120 ⇒ 00:51:35.610 Katherine Bayless: Okay. Yeah.

470 00:51:35.880 ⇒ 00:51:41.759 Uttam Kumaran: And then when… you can see in Snowflake, I forgot where it is, but you can see the Grant lineage.

471 00:51:41.940 ⇒ 00:51:43.070 Uttam Kumaran: Which, like…

472 00:51:43.280 ⇒ 00:51:58.609 Uttam Kumaran: for… for any of the governance fans in the building is, like, that’s, like, the really important thing, is just to make sure that you’re aware at any point, like, what someone has access to, and you’re ideally, like, creating a role for all access. You’re not grant… you never want to grant direct

473 00:51:58.860 ⇒ 00:51:59.960 Uttam Kumaran: Access.

474 00:52:00.090 ⇒ 00:52:03.550 Uttam Kumaran: to any… to any one person. It always should go through a role.

475 00:52:04.030 ⇒ 00:52:04.590 Katherine Bayless: Right.

476 00:52:04.590 ⇒ 00:52:08.820 Uttam Kumaran: Right, right, right. Yeah. So you’re granting something to the role, the role to a person, or a group.

477 00:52:09.140 ⇒ 00:52:20.369 Katherine Bayless: Right. Yeah, that makes sense, yeah. Yeah, you’ve correctly identified my PTSD, at my last place. Our Tableau environment was initially configured with, like, user-specific granular.

478 00:52:20.370 ⇒ 00:52:21.609 Uttam Kumaran: I’m like…

479 00:52:21.610 ⇒ 00:52:28.729 Katherine Bayless: they can see that the workbook exists, but they can’t open it, or they can open it, but they can’t see this sheet, and I was like, oh my god, you guys are trying to drive me insane. So yeah.

480 00:52:28.730 ⇒ 00:52:29.550 Uttam Kumaran: Ever.

481 00:52:29.550 ⇒ 00:52:36.909 Katherine Bayless: Yeah, we’re at a place now where, like, broad is fine, and we can add the complexity as it has a use case, but yeah, yeah.

482 00:52:37.010 ⇒ 00:52:38.739 Katherine Bayless: Keep it simple.

483 00:52:42.310 ⇒ 00:52:43.100 Awaish Kumar: Okay.

484 00:52:44.530 ⇒ 00:52:45.070 Uttam Kumaran: Okay.

485 00:52:45.560 ⇒ 00:52:56.489 Uttam Kumaran: So cool, I think a couple of, like, homework items on our side, so we’re gonna just… we’ll just clean through this. We’ll put in the dbt setup information, like the… meaning the…

486 00:52:56.810 ⇒ 00:53:07.959 Uttam Kumaran: different environments. We’ll put the, role-based asset control sort of plan also in there. I don’t know if that, Catherine, you wanted to do that today, but we can grant

487 00:53:08.460 ⇒ 00:53:21.830 Uttam Kumaran: those folks access to ProdMarts today, and then I can just… basically, the additional benefit of, like, assigning them a role is you can just assume the role, and you’re basically pseudoing them, because there’s no way in Snowflake to pseudo a person.

488 00:53:22.040 ⇒ 00:53:27.150 Uttam Kumaran: So, if they’re, like, they have that role, you can just be like, let me just assume the role and make sure that

489 00:53:27.280 ⇒ 00:53:28.969 Uttam Kumaran: I can also access everything.

490 00:53:29.770 ⇒ 00:53:37.520 Katherine Bayless: Yeah, that would be awesome. I mean, if we’ve got a little time, we could do an example one, but otherwise, yeah, whenever is fine.

491 00:53:37.520 ⇒ 00:53:38.140 Uttam Kumaran: Okay.

492 00:53:38.450 ⇒ 00:53:39.130 Katherine Bayless: Yeah.

493 00:53:39.210 ⇒ 00:53:39.930 Uttam Kumaran: Okay.

494 00:53:47.200 ⇒ 00:53:49.149 Katherine Bayless: What else is on everybody’s minds?

495 00:53:57.570 ⇒ 00:54:01.170 Kyle Wandel: Annoyance of cleaning this badge game, Dana, let’s be honest, so…

496 00:54:01.570 ⇒ 00:54:03.820 Katherine Bayless: Like, multiple days versus whatever, and…

497 00:54:03.930 ⇒ 00:54:07.049 Kyle Wandel: I don’t know if it can be clean, that’s what I’m trying to figure out, but we’ll see.

498 00:54:08.530 ⇒ 00:54:09.360 Katherine Bayless: Yeah.

499 00:54:10.080 ⇒ 00:54:24.060 Katherine Bayless: Yeah, I mean, honestly, Kyle, if you want to take some time this afternoon and just kind of, like, pair on it, I mean, maybe we can just, like, two heads bash away at it, but I think we probably will have to have the conference’s team figure out some of the edge cases, just so we are confident that we got it right.

500 00:54:25.340 ⇒ 00:54:31.890 Katherine Bayless: It’s funny to me, though, because I’m like, there’s gotta be a source for this? Like, where’s the… where’s the authoritative mappings? I’m like… but…

501 00:54:32.460 ⇒ 00:54:33.220 Katherine Bayless: Oh, well.

502 00:54:34.410 ⇒ 00:54:40.879 Kyle Wandel: Well, I mean, so, like, I think what happened is they used the wrong scanner. So they used the wrong scanner for…

503 00:54:41.070 ⇒ 00:54:41.680 Kyle Wandel: Certain…

504 00:54:41.680 ⇒ 00:54:42.510 Katherine Bayless: session.

505 00:54:42.510 ⇒ 00:54:59.260 Kyle Wandel: So, like, what I think is happening, at least for this particular… this one particular case, is that the session Foundry 2 was using an entries code scanner, because there’s a lot of entry codes for Foundry 02, basically, which is the next day.

506 00:54:59.830 ⇒ 00:55:08.700 Kyle Wandel: And so Foundry 2 session occurred on the first day, and then those Foundry 02 sessions only occur between this particular time slot.

507 00:55:08.940 ⇒ 00:55:12.470 Kyle Wandel: Which is not the time slot of Foundry 02.

508 00:55:13.050 ⇒ 00:55:13.880 Katherine Bayless: Yeah.

509 00:55:13.880 ⇒ 00:55:24.679 Kyle Wandel: And so, it’s like, do we try to figure out if you can just use the time slot of the session listing, and using the listing.

510 00:55:25.000 ⇒ 00:55:29.319 Kyle Wandel: And see if you can only… if you only grab people who scanned.

511 00:55:30.660 ⇒ 00:55:46.230 Kyle Wandel: at Fountain Blue, I guess, in terms of just, like, location, you pick the… that’s B1 filter, and then you kind of recode the people into that session, basically, based on the scan date and the location, but then you’re going to be picking up people who

512 00:55:46.510 ⇒ 00:55:47.730 Kyle Wandel: scanned.

513 00:55:48.640 ⇒ 00:55:56.190 Kyle Wandel: at other sessions during… either A during that time, or if they just did an entry scanner at that time, in between that time slot.

514 00:55:56.550 ⇒ 00:55:57.490 Katherine Bayless: Yeah.

515 00:55:57.830 ⇒ 00:56:03.129 Kyle Wandel: So I don’t really know, like, the best way to figure this out.

516 00:56:03.760 ⇒ 00:56:13.129 Katherine Bayless: Yeah, okay. Yeah, I apologize, I did not realize when you explained it before, I thought it was more just, like, that the codes weren’t mapping, but yeah, people using the wrong scanners at the wrong place.

517 00:56:13.130 ⇒ 00:56:27.350 Kyle Wandel: Yeah. And so, it… basically what we talked about the other day, in terms of, like, the multiple day stuff, that’s probably what ended up happening, is my guess, is that they just picked up the wrong scanner, they didn’t refresh it, or something… I don’t know how the scanners work, so that’s another thing.

518 00:56:27.350 ⇒ 00:56:36.330 Kyle Wandel: But my guess is they didn’t refresh it or whatever, and so it just… it just maintained that session from the previous day, or the previous last time that thing was used, basically.

519 00:56:36.340 ⇒ 00:56:42.290 Kyle Wandel: And I don’t know the best way to clean that from our side, because there’s really…

520 00:56:43.030 ⇒ 00:56:44.840 Kyle Wandel: There’s not a great way to do it.

521 00:56:45.540 ⇒ 00:56:46.410 Katherine Bayless: -

522 00:56:47.300 ⇒ 00:57:08.069 Katherine Bayless: Well, and it’s funny, because I, similarly, I observed and participated in… I scanned people, so, like, I got to, like, use the scanners and see that part, but I didn’t see the, like, check-in, check-out process for them, so I, yeah, like, whose job was it to make sure it was set up for the right session, and were people just like, oh, just grab a loose one and using it, like, oh god.

523 00:57:08.130 ⇒ 00:57:09.760 Katherine Bayless: I did not. Okay.

524 00:57:10.530 ⇒ 00:57:19.789 Kyle Wandel: So yeah, I don’t really know what the best way to respond to conferences is, because they did say, like, you could use a scan session time, which they’re not wrong, but it doesn’t…

525 00:57:20.940 ⇒ 00:57:30.639 Kyle Wandel: fix the situation, because then it’ll just misclassify other records that may not be… like, the only way you can do it is, like, this is a likely person, basically.

526 00:57:31.140 ⇒ 00:57:33.729 Katherine Bayless: Right. I mean, I think, yeah, like…

527 00:57:33.850 ⇒ 00:57:43.710 Katherine Bayless: probably a waterfall logic where, yeah, like, anything where, like, the code and the time and the scan are all matched up and correct, those are fine, and then it’s, like.

528 00:57:44.040 ⇒ 00:57:46.609 Katherine Bayless: What are all the other outliers?

529 00:57:47.100 ⇒ 00:57:55.370 Kyle Wandel: Yeah, and then you have to pick a time, so, like, you have to pick, like, within 10 minutes of the time slot before and after. Like, that’s probably…

530 00:57:56.280 ⇒ 00:57:57.220 Katherine Bayless: Right.

531 00:57:57.220 ⇒ 00:57:59.440 Kyle Wandel: That’s probably that session, but again, you have to say.

532 00:57:59.440 ⇒ 00:57:59.800 Uttam Kumaran: Yeah.

533 00:57:59.830 ⇒ 00:58:07.709 Kyle Wandel: session and not the actual session, which, again, goes to a data quality issue. I know you and I talked about this, but I almost just really want to…

534 00:58:08.130 ⇒ 00:58:18.830 Kyle Wandel: not really push back, but just be like, this is the data that we were kind of given, and we’ll clean up the best we can, but, it just more goes back to… this gain of practice is gonna be better, but I don’t know the best way to handle that.

535 00:58:20.620 ⇒ 00:58:21.160 Katherine Bayless: Yeah.

536 00:58:21.370 ⇒ 00:58:26.939 Chi Quinn: Well, I was gonna say, as far as the reporting end, is that something I should hold off now? Because…

537 00:58:27.350 ⇒ 00:58:31.979 Chi Quinn: That’s… that’s the same scan… the badge scanner, data, correct?

538 00:58:32.130 ⇒ 00:58:32.730 Kyle Wandel: Yeah, yeah.

539 00:58:32.730 ⇒ 00:58:40.520 Chi Quinn: Oh, okay, so should I at least hold off for now, or, like, sending it off, or just send it and…

540 00:58:42.110 ⇒ 00:58:49.190 Chi Quinn: I guess, because they might… I don’t know, someone would notice, like, hey, there’s something wrong, or…

541 00:58:49.190 ⇒ 00:59:03.079 Kyle Wandel: Oh, I mean, yeah, people are noticing, that’s why… that’s why I’m getting info… or emails or messages about it from, conference people, because there’s, for example, there’s no way… they comment there’s no way for one particular

542 00:59:03.720 ⇒ 00:59:11.579 Kyle Wandel: Session to have… 1,000 peop… or not, yeah, almost 2,000 people, which I agree.

543 00:59:11.900 ⇒ 00:59:29.509 Kyle Wandel: But it goes back to, okay, so how many people should be that? So, like, if it’s, like… in that time window, if I just use that time window of that particular person and Fontaine Blue, it’s, like, 166 people. So my guess is that even that’s incorrect. But that would be the logic that you would do, which would be the location the session is happening.

544 00:59:29.510 ⇒ 00:59:35.899 Kyle Wandel: within scans that happen within that particular, time slot, basically.

545 00:59:37.990 ⇒ 00:59:47.799 Kyle Wandel: And that’s, like, really, like, the two main qualifiers I feel like you have to use, like, the… because if… I don’t know if there’s any more that you can use, basically.

546 00:59:48.370 ⇒ 00:59:51.370 Katherine Bayless: Yeah, well, and I think… yeah, I mean…

547 00:59:51.850 ⇒ 00:59:58.350 Katherine Bayless: I, yeah, I agree, I think that’s all the logic that really is available to us, but then I think, yeah, like, to Kai’s point, it’s like.

548 00:59:59.930 ⇒ 01:00:21.909 Katherine Bayless: Yeah, for internal, like, you know, prospecting and analysis, like, you know, I’d be very comfortable being like, look, hey man, this is the data we got, this is what you got to work with. If you want to help, you know, tidy it up, fine, but, like, we’re not going to go through line by line. But then the trick is the sponsors that we have, like, committed to share the data with, like, I don’t want us to send…

549 01:00:21.910 ⇒ 01:00:27.119 Katherine Bayless: people’s contact information somewhere it shouldn’t go, and so I’m like, oh god.

550 01:00:27.850 ⇒ 01:00:45.730 Kyle Wandel: could you… could we maybe just, like, I don’t know, the integrity way, the data integrity way to do it is to flag those people that don’t fit in a time slot, really, like, don’t match, like, period, like, just don’t match. And then, or at least on multiple days. And so, like, they just don’t match, and then you just say that there are X number of unmatched.

551 01:00:46.080 ⇒ 01:00:48.690 Katherine Bayless: scans, but I don’t really know the best way to do that.

552 01:00:49.760 ⇒ 01:00:50.610 Katherine Bayless: Yeah.

553 01:00:52.120 ⇒ 01:00:58.390 Kyle Wandel: I definitely agree, I think you want it from a data integrity issue, but we’re just missing a large part of the data, because…

554 01:00:59.700 ⇒ 01:01:01.939 Kyle Wandel: Of scanner issues in general.

555 01:01:03.950 ⇒ 01:01:08.689 Katherine Bayless: Suddenly, I’m like, oh, I see now why the old data team said it would take 6 weeks to put this data together.

556 01:01:10.900 ⇒ 01:01:19.239 Kyle Wandel: And I think it’s not… like, the data itself is not bad. It’s the… it’s the input of the data from the very, very beginning that is just not…

557 01:01:19.640 ⇒ 01:01:23.599 Kyle Wandel: Done very well, and it’s hard to do when you have probably what?

558 01:01:23.750 ⇒ 01:01:30.820 Kyle Wandel: I don’t know how many scanners we have, but you probably have, like, 100… like, 100 plus different people who are scanning in different ways, and so it’s just like, that’s…

559 01:01:31.540 ⇒ 01:01:34.199 Kyle Wandel: It’s difficult to maintain that process.

560 01:01:34.710 ⇒ 01:01:35.270 Katherine Bayless: Yeah.

561 01:01:35.720 ⇒ 01:01:49.529 Katherine Bayless: Yeah. Okay, yeah, I mean, I think maybe, I know since we’re at time for this call, but maybe, Kyle, if you’re free after I get off this Remembers thing, let’s maybe put heads together and figure out, because, yeah, I think we need to…

562 01:01:49.700 ⇒ 01:01:58.169 Katherine Bayless: have a… I like your idea of, like, you know, cordoning off the bad scams, and then maybe kind of going from there and saying, like…

563 01:01:58.900 ⇒ 01:02:09.149 Katherine Bayless: we only want to deliver the data to those sponsors that we’re positive is correct, and then, yeah, yeah, yeah, yeah. Are you around this afternoon?

564 01:02:09.150 ⇒ 01:02:11.210 Kyle Wandel: Yeah, I think s- yes, I am.

565 01:02:12.810 ⇒ 01:02:15.570 Katherine Bayless: I’ll ping you after I get done with remembers, then, because…

566 01:02:16.240 ⇒ 01:02:16.980 Kyle Wandel: Yeah.

567 01:02:17.200 ⇒ 01:02:19.009 Kyle Wandel: What time is that at? Like, 1… like, noon?

568 01:02:19.170 ⇒ 01:02:23.250 Katherine Bayless: It’s right after this, it’s, I should be done by 12.15.

569 01:02:23.630 ⇒ 01:02:30.169 Kyle Wandel: Okay, cool, yeah. I’m just gonna eat lunch and it’ll be great. But yeah, that’s, like, I mean, the basking’s my biggest thing, and I just… it’s just annoying.

570 01:02:30.350 ⇒ 01:02:31.980 Katherine Bayless: Yeah. Yeah, yeah.

571 01:02:32.520 ⇒ 01:02:48.799 Katherine Bayless: Well, I mean, it’s funny too, then, because I’m, like, I’m over here trying to, like, put together prospecting lists, so that marketing can send emails out, and I’m like, okay, well, I should make sure that the scans that I drew from for those are actually correct scans before we email people being like, I hope you enjoyed this event, and then they’re like, I didn’t go to that event.

572 01:02:49.220 ⇒ 01:02:51.970 Kyle Wandel: I do think that that’s where… yeah.

573 01:02:52.320 ⇒ 01:02:59.950 Kyle Wandel: where the issue becomes. So, I think it goes, like, what you were saying, we probably do need to just match it based on what we know.

574 01:03:00.200 ⇒ 01:03:01.939 Katherine Bayless: And, like, the best we can.

575 01:03:01.940 ⇒ 01:03:09.440 Kyle Wandel: Basically, and then… If it’s not within that time slot, then it probably is not the person.

576 01:03:09.690 ⇒ 01:03:10.300 Katherine Bayless: Right.

577 01:03:10.820 ⇒ 01:03:15.779 Katherine Bayless: I wonder how many false positives there are, too, but yeah, yeah, yeah, okay, alright, yeah, let’s go.

578 01:03:16.080 ⇒ 01:03:16.580 Katherine Bayless: We’ll die.

579 01:03:16.580 ⇒ 01:03:17.689 Kyle Wandel: Probably a lot.

580 01:03:17.690 ⇒ 01:03:18.990 Katherine Bayless: Yeah, I know.

581 01:03:20.150 ⇒ 01:03:24.039 Kyle Wandel: I mean, like I said, just… so even just, like, narrowing down to…

582 01:03:24.360 ⇒ 01:03:27.470 Kyle Wandel: the Foundry O2, like, that particular code.

583 01:03:27.600 ⇒ 01:03:30.950 Kyle Wandel: It said, like, there were 167 people within that.

584 01:03:31.140 ⇒ 01:03:39.280 Kyle Wandel: time slot, basically. And, like, my guess is that’s only supposed to be, like, 18, is my guess, like, less than 20, so…

585 01:03:41.150 ⇒ 01:03:44.200 Katherine Bayless: Yeah, okay. Well, we’ll see what we can come up with.

586 01:03:45.770 ⇒ 01:03:46.970 Katherine Bayless: Good catch, though.

587 01:03:47.510 ⇒ 01:03:48.240 Katherine Bayless: Personally.

588 01:03:49.310 ⇒ 01:03:50.520 Kyle Wandel: Damn, yeah.

589 01:03:51.890 ⇒ 01:03:59.040 Katherine Bayless: I mean, good news is, we’re… I’ve been given the mandate to figure out a better way to do this, so, just…

590 01:03:59.040 ⇒ 01:03:59.730 Kyle Wandel: from…

591 01:03:59.730 ⇒ 01:04:00.320 Katherine Bayless: Yeah.

592 01:04:00.700 ⇒ 01:04:12.239 Kyle Wandel: From your perspective, is it… is it, like, is it an okay data integrity practice to just be like, we, this is ones that we have… we are positive on, and we have X number of records that we’re unsure about, or how would you guys handle it?

593 01:04:14.310 ⇒ 01:04:16.489 Uttam Kumaran: Yeah, it’s often…

594 01:04:17.930 ⇒ 01:04:24.500 Uttam Kumaran: Some of it is like, okay, our finding and our data is gonna inform how we collect it in the future, so…

595 01:04:25.680 ⇒ 01:04:40.109 Uttam Kumaran: in my eyes, like, the end customer may or may not be happy either way, as long as we know, like, we have, like, a recommendation on, like, what is the fix gonna be, because even if we’re like, hey, we’re confident about these, not confident about these, they may say.

596 01:04:40.230 ⇒ 01:04:46.250 Uttam Kumaran: Well, like, just give us everything, because we previously reported on everything, and…

597 01:04:46.520 ⇒ 01:05:05.940 Uttam Kumaran: what I always like to do is to just share, like, what we found, and then say, in or… like, moving forward, in order for this to be accurate, it has to be in a certain way. And then, on some pieces, look, if we’re able to give folks subsets of data, and maybe they are able to say, like, this should be included, this shouldn’t, then they can do that, but…

598 01:05:08.880 ⇒ 01:05:20.500 Uttam Kumaran: I don’t know, it’s sort of tricky. It’s also tricky not knowing, like, who the folks are on your guy’s end, just, like, if they’re gonna be okay with that, or if, like, they sort of, like, just give us something, whatever it is.

599 01:05:21.560 ⇒ 01:05:43.960 Katherine Bayless: Well, I mean, it’s… so these are, like, sponsor, like, comp… I mean, so obviously, also, there’s the internal analytics and stuff like that, but, like, the piece that is on my mind of, like, oh god, is the sponsors that, like, we basically… we’ve collected these scans, and we give them to them, and we attest that, like, you know, this person went to your event, and they’ve consented to receive marketing by virtue of having been scanned, kind of thing, and so…

600 01:05:43.960 ⇒ 01:05:52.090 Katherine Bayless: like, if we’re sending sponsors people that weren’t at their event because the wrong scanner was used. I mean, I think to that end, like.

601 01:05:52.090 ⇒ 01:06:09.570 Katherine Bayless: to Kyle’s point, if they’re inside the time window, we just kind of have to trust that they were there. If they’re outside the time window, I think we can maybe reasonably set them aside. And in theory, the sponsor wouldn’t be asking for data that doesn’t exist, right? If they know only 20 people are at their session, we give them a.

602 01:06:09.570 ⇒ 01:06:10.330 Uttam Kumaran: Yes.

603 01:06:10.330 ⇒ 01:06:29.279 Katherine Bayless: Like, they’d be happy, but incorrect. But yeah, I think I agree, the salient point, I think, is, like, having documentation around, okay, this was, you know, the issue around scanners being used incorrectly, this was the data that we were, you know, working with, and these were the choices we made to, like.

604 01:06:29.280 ⇒ 01:06:31.030 Katherine Bayless: Deliver the final results.

605 01:06:31.030 ⇒ 01:06:35.270 Katherine Bayless: So that at least it’s captured somewhere, what we decided to do.

606 01:06:35.560 ⇒ 01:06:36.170 Uttam Kumaran: Yes.

607 01:06:37.370 ⇒ 01:06:39.090 Katherine Bayless: Yeah, and…

608 01:06:39.200 ⇒ 01:06:55.690 Katherine Bayless: I’ll think about whether or not we should let legal know. So there is also… there’s a… there is a different issue with the scanner stuff, which is there are a few sessions where the sponsor, you know, paid for the session and was expecting to get the list, but then nobody was scanning at the session, which is a totally different issue.

609 01:06:55.690 ⇒ 01:06:56.190 Uttam Kumaran: Yeah.

610 01:06:56.190 ⇒ 01:07:04.580 Katherine Bayless: Like, we’re already in conversations with our legal team around, like, well, can we give them, like, I don’t know, just, like, a random 100 people who were, like, nearby?

611 01:07:06.930 ⇒ 01:07:07.480 Katherine Bayless: Look at this.

612 01:07:07.480 ⇒ 01:07:12.080 Uttam Kumaran: That’s how I was gonna… That’s up to you.

613 01:07:13.030 ⇒ 01:07:16.560 Uttam Kumaran: Catherine, I have no… Yeah.

614 01:07:16.760 ⇒ 01:07:20.210 Uttam Kumaran: Then we’re just painting, with a paintbrush, whatever.

615 01:07:21.040 ⇒ 01:07:22.200 Katherine Bayless: Exactly.

616 01:07:22.200 ⇒ 01:07:29.759 Kyle Wandel: That’s… and that’s what I want, personally. That’s personally what I like to try to avoid, but I also understand the need to have the correct data list, but, like.

617 01:07:30.780 ⇒ 01:07:47.199 Kyle Wandel: I can only… we can only do so much cleaning to the point where it’s like, it could be… it could be true, but I don’t really want to report on it could be true. I’d say… I’d rather just say it’s true, and, like, we need to be better, basically. Or, like, well, not my team, but the other teams need to be better.

618 01:07:47.760 ⇒ 01:07:56.489 Katherine Bayless: Yeah, I mean, yeah, like, at the end of the day, we just need a different solution for attendance capture at these events. Like, I mean, the badge scanners.

619 01:07:56.780 ⇒ 01:08:03.280 Katherine Bayless: Yeah. I mean, they’re janky. They’re just janky. They’re not great devices, and I feel like ARFID badges make more sense.

620 01:08:03.440 ⇒ 01:08:08.190 Katherine Bayless: So yeah, we’re gonna look into those options, but… but yeah. Yeah.

621 01:08:10.030 ⇒ 01:08:10.720 Katherine Bayless: Yeah.

622 01:08:11.820 ⇒ 01:08:12.580 Katherine Bayless: Fun.

623 01:08:15.520 ⇒ 01:08:16.180 Uttam Kumaran: Okay.

624 01:08:16.790 ⇒ 01:08:18.270 Uttam Kumaran: So yeah, I think we’ll be…

625 01:08:18.420 ⇒ 01:08:23.330 Uttam Kumaran: A little bit on modeling stuff today and on Monday, so should have some follow-up updates there.

626 01:08:23.580 ⇒ 01:08:31.300 Kyle Wandel: Yep, just keep me posted on anything you need, for the join logic, or anything you… any pipelines you want me to create, and I’ll just keep…

627 01:08:31.580 ⇒ 01:08:33.080 Kyle Wandel: I’m annoyed this bad stuff.

628 01:08:33.990 ⇒ 01:08:34.590 Uttam Kumaran: Okay.

629 01:08:35.350 ⇒ 01:08:35.910 Awaish Kumar: Great.

630 01:08:37.200 ⇒ 01:08:37.880 Katherine Bayless: Alright.

631 01:08:37.880 ⇒ 01:08:38.680 Uttam Kumaran: Perfect.

632 01:08:38.920 ⇒ 01:08:40.309 Katherine Bayless: See y’all later, then.

633 01:08:40.550 ⇒ 01:08:41.090 Uttam Kumaran: Thank you.

634 01:08:41.920 ⇒ 01:08:42.710 Kyle Wandel: Thanks, guys.

635 01:08:42.710 ⇒ 01:08:44.120 Chi Quinn: Bye. Thank you. Bye.