2026-03-05_default_project_dashboard_sync

Meeting Title: Default Project Dashboard Sync Date: 2026-03-05 Meeting participants: Scratchpad Notetaker, Uttam Kumaran, Nandika Jhunjhunwala, Demilade Agboola, Mustafa Raja, Brylle Girang, Caitlyn Vaughn

WEBVTT

1 00:00:25.410 ⇒ 00:00:26.610 Uttam Kumaran: Hello!

2 00:00:26.820 ⇒ 00:00:27.790 Nandika Jhunjhunwala: Bye!

3 00:00:28.530 ⇒ 00:00:31.289 Uttam Kumaran: Oh, my jacket looks really ugly in the back, hold on.

4 00:00:48.530 ⇒ 00:00:54.720 Uttam Kumaran: Great. How’s Omni been, Ondica? I don’t know if you’ve, like, played around at all. Or with the AI piece.

5 00:00:56.400 ⇒ 00:00:57.539 Nandika Jhunjhunwala: Sorry, can you say that again?

6 00:00:57.540 ⇒ 00:01:00.090 Uttam Kumaran: How has, like, Omni been playing around with, like, the AI?

7 00:01:00.090 ⇒ 00:01:09.679 Nandika Jhunjhunwala: Yeah, I’ve been very hands-on with growth this week, so plan for me is to, like, pivot back into, like, analytics and, like.

8 00:01:09.810 ⇒ 00:01:25.550 Nandika Jhunjhunwala: Omni, like, today, later, and tomorrow. Okay. But yeah, I haven’t played around much. I had a session with Demi, which was helpful. A lot of the views that we need to create the dashboards that we need, aren’t fully there yet.

9 00:01:25.750 ⇒ 00:01:30.009 Nandika Jhunjhunwala: So I was trying to create a view of my own,

10 00:01:30.140 ⇒ 00:01:32.150 Nandika Jhunjhunwala: I’m just having some trouble with…

11 00:01:32.500 ⇒ 00:01:41.830 Nandika Jhunjhunwala: accessing data, I think I was, like, looking in the wrong, like, data source. But yeah, I think we’ll get there soon. It’s a lot of trial and error on my end, yeah.

12 00:01:42.350 ⇒ 00:01:42.890 Uttam Kumaran: Cool.

13 00:01:43.030 ⇒ 00:01:48.369 Demilade Agboola: Yeah, if you, like, for the schemas, and I know there’s a lot going on there.

14 00:01:49.410 ⇒ 00:01:50.660 Nandika Jhunjhunwala: Personally, like.

15 00:01:50.920 ⇒ 00:01:54.229 Demilade Agboola: you can always… we can always have, like, a call where I can, like, walk you through.

16 00:01:54.230 ⇒ 00:01:54.630 Nandika Jhunjhunwala: Yeah.

17 00:01:54.630 ⇒ 00:01:56.779 Demilade Agboola: Yes, that’s not… that’s definitely not a problem.

18 00:01:58.790 ⇒ 00:02:04.580 Demilade Agboola: And if you do need things, like, expedited based on an ad hoc request, please let me know, like, that’s not…

19 00:02:04.580 ⇒ 00:02:05.110 Nandika Jhunjhunwala: prefer.

20 00:02:05.110 ⇒ 00:02:05.660 Demilade Agboola: And kept.

21 00:02:05.660 ⇒ 00:02:12.670 Nandika Jhunjhunwala: I think, yeah, like, there’s a huge push internally to have, like, sales activity dashboards, like.

22 00:02:13.100 ⇒ 00:02:13.590 Uttam Kumaran: Yeah.

23 00:02:13.590 ⇒ 00:02:22.660 Nandika Jhunjhunwala: it’s first data, maybe you join with Hyperline, if possible, or QuickBooks. Yeah, so that is definitely…

24 00:02:22.660 ⇒ 00:02:23.920 Uttam Kumaran: We’re gonna talk about it today.

25 00:02:23.920 ⇒ 00:02:24.450 Nandika Jhunjhunwala: Yeah.

26 00:02:24.450 ⇒ 00:02:26.459 Uttam Kumaran: I think we have some good progress on that, too.

27 00:02:27.600 ⇒ 00:02:28.670 Demilade Agboola: Yeah, like…

28 00:02:28.980 ⇒ 00:02:43.680 Demilade Agboola: some of that, because, again, we’ll share some of that information, like, there is quite a bit of data in the warehouse. You might say too much, but there is a lot, just a lot of stuff going on there. And the plan is just being able to…

29 00:02:43.800 ⇒ 00:02:46.409 Demilade Agboola: Whittle it down to what’s the most important thing.

30 00:02:46.630 ⇒ 00:02:50.910 Demilade Agboola: And just help you have that so you can quickly do whatever you need to do with it.

31 00:02:50.910 ⇒ 00:03:06.349 Nandika Jhunjhunwala: Yeah, I think the schemas are more so also for the rest of the team, so that they can see, like, the data that they expect to be available in the warehouse is available or not, and in the form that they’re sort of visualizing it to be. I think there’s a lot of, like.

32 00:03:06.640 ⇒ 00:03:18.890 Nandika Jhunjhunwala: gaps in, like, context that we need to bridge for the modeling to be, completed the way everybody’s, like, hoping it will be, so I think the schemas would help us, like.

33 00:03:19.060 ⇒ 00:03:25.050 Nandika Jhunjhunwala: Like, identify, like, what information we haven’t, you know, like, transferred over to you guys to, like, do that.

34 00:03:25.150 ⇒ 00:03:41.560 Nandika Jhunjhunwala: Because I think, like, when I dug into, like, Salesforce data, I found a bunch of, like, tables that were missing, and then, like, Demi very helpfully, like, connected them, and that was super, super nice. So I think just, like, doing sort of that sort of, like, check with every internal stakeholder that owns that piece of data would…

35 00:03:42.060 ⇒ 00:03:43.740 Nandika Jhunjhunwala: Would be great for us.

36 00:03:44.830 ⇒ 00:03:51.870 Demilade Agboola: Yeah, that’s definitely not a problem. Yeah, I know, like, because Salesforce has, like, what, almost, like, over 800 different.

37 00:03:51.870 ⇒ 00:03:53.750 Nandika Jhunjhunwala: Yeah, absolutely.

38 00:03:53.750 ⇒ 00:03:57.400 Demilade Agboola: I’m not adjusting every single thing, so just being able to, like, show what’s available.

39 00:03:57.540 ⇒ 00:04:01.960 Demilade Agboola: will be very helpful. So, yeah, we’ll just, like, that will just be… we’ll do a raw schema.

40 00:04:02.680 ⇒ 00:04:07.850 Nandika Jhunjhunwala: And just, like, the name, like, the names of the columns in each of the schemas, so that’ll be, like, the information as well.

41 00:04:08.130 ⇒ 00:04:08.870 Nandika Jhunjhunwala: Thank you.

42 00:04:09.180 ⇒ 00:04:10.220 Demilade Agboola: Not a problem.

43 00:04:15.460 ⇒ 00:04:24.540 Demilade Agboola: Okay, so for… This week, So we have, like, our, like, executive summary.

44 00:04:26.530 ⇒ 00:04:29.149 Demilade Agboola: We plan to, like, go through the different,

45 00:04:30.110 ⇒ 00:04:33.800 Demilade Agboola: Data platforms, the dashboard progress, timelines.

46 00:04:34.180 ⇒ 00:04:36.440 Demilade Agboola: Also talk about product analytics.

47 00:04:36.940 ⇒ 00:04:39.880 Demilade Agboola: about the… reverse ETL.

48 00:04:42.540 ⇒ 00:04:46.599 Demilade Agboola: talk about, like, ID matching between data sources, because that’s come up.

49 00:04:46.810 ⇒ 00:04:47.820 Demilade Agboola: as well.

50 00:04:48.030 ⇒ 00:04:54.379 Demilade Agboola: So, like, the list of fields and metrics, which kind of ties to one article just mentioned.

51 00:04:54.520 ⇒ 00:04:59.190 Demilade Agboola: As well as…

52 00:05:01.580 ⇒ 00:05:09.279 Demilade Agboola: As well as, sensitive, sorry, sensitive dashboards, new requests, and risks and mitigations, so let’s just get into it.

53 00:05:09.530 ⇒ 00:05:19.530 Demilade Agboola: So from a high level, this week, we’ve been able to start building out the, like, build out the V1 Army dashboard, and they’re ready for QA.

54 00:05:19.740 ⇒ 00:05:22.000 Demilade Agboola: So this week was…

55 00:05:22.190 ⇒ 00:05:33.990 Demilade Agboola: just fully creating dbt models. We’re not creating models, like, testing and pushing them over and over again, just so that the numbers were good. Testing dbt runs, rolling out the dashboards.

56 00:05:34.100 ⇒ 00:05:45.269 Demilade Agboola: And also, like, just, we’re going to use stakeholder feedback to just QA the data models and see where the possible gaps exist in how we currently are looking at the data. And so we have a call with Laura today.

57 00:05:46.080 ⇒ 00:05:50.510 Demilade Agboola: On the product analytics side, we have, like, post hoc events.

58 00:05:51.090 ⇒ 00:05:56.759 Demilade Agboola: And using properties that they continue to be implemented, and Greg and Nadica are working closely on this.

59 00:05:57.370 ⇒ 00:06:04.489 Demilade Agboola: Ultimately, the next steps will be, on the product analytics side, is that we want to build our charts

60 00:06:04.700 ⇒ 00:06:13.179 Demilade Agboola: For events, and we have started tracking and, are building out the user property and event tracking implementation.

61 00:06:13.620 ⇒ 00:06:26.639 Demilade Agboola: On the data platform and analytics side, we want to continue to roll out more dashboards. The goal is to have two dashboards out this week, the LoRa dashboard and the Lauren dashboard out this week.

62 00:06:26.770 ⇒ 00:06:33.789 Demilade Agboola: And continuing to use the feedback that we’re getting off these dashboards, for… QAing and, and…

63 00:06:33.990 ⇒ 00:06:36.170 Demilade Agboola: Modifying our modeling.

64 00:06:37.410 ⇒ 00:06:40.150 Demilade Agboola: Alright, so, in terms of the key wins this week.

65 00:06:41.280 ⇒ 00:06:45.010 Demilade Agboola: the Omni dashboard is live,

66 00:06:46.210 ⇒ 00:06:50.100 Demilade Agboola: we’re not… we’ve not yet fully handed it over to the stakeholder, which is Laura.

67 00:06:50.510 ⇒ 00:06:55.129 Demilade Agboola: But the idea is we will start to start handing over these dashboards.

68 00:06:55.470 ⇒ 00:07:04.070 Demilade Agboola: The… there have been some delays due to, like, modeling, that caused some miscommunication.

69 00:07:04.320 ⇒ 00:07:11.739 Demilade Agboola: So basically… Some of the numbers were ready, quote-unquote, like, 2 weeks ago, but, based off of

70 00:07:11.970 ⇒ 00:07:16.999 Demilade Agboola: the new information that came out in terms of getting access to the EQALS dashboard late last week.

71 00:07:17.350 ⇒ 00:07:22.039 Demilade Agboola: Some of those numbers needed to be reconstructed and built again from scratch.

72 00:07:22.190 ⇒ 00:07:26.460 Demilade Agboola: And so that caused a little delay in being able to push out the dashboard fully.

73 00:07:26.760 ⇒ 00:07:31.329 Demilade Agboola: But in terms of the impact, we have our first integrated view of the business data.

74 00:07:31.750 ⇒ 00:07:36.169 Demilade Agboola: And then the feedback loop for iteration as needed is…

75 00:07:37.000 ⇒ 00:07:48.199 Demilade Agboola: what we’re going to start now, so it’s basically, hey, here’s your dashboard on a day-to-day. If you have questions, or some things just don’t seem to align with the definition, do let us know, we will fix that.

76 00:07:50.200 ⇒ 00:07:56.760 Demilade Agboola: And so the idea today is also tied to the second win. We’re in the phase where we can start our dashboard QA and iteration.

77 00:07:57.300 ⇒ 00:08:00.889 Demilade Agboola: And we want to be able to have stakeholders use it.

78 00:08:01.070 ⇒ 00:08:14.179 Demilade Agboola: have questions, see ways in which they can integrate new data in there, because again, we have data from Salesforce, Hyperline, QuickBooks, so potentially, like, Laura might actually want to see other things as well in that dashboard.

79 00:08:14.320 ⇒ 00:08:17.190 Demilade Agboola: And so we can look at how we can integrate it in there.

80 00:08:17.460 ⇒ 00:08:25.410 Demilade Agboola: And also, we set up, like, dbt runs on, like, GitHub Actions, so the idea now is every 6 hours, dbt would run, and just to ensure that

81 00:08:26.420 ⇒ 00:08:29.350 Demilade Agboola: The new transformations are done.

82 00:08:29.540 ⇒ 00:08:47.569 Demilade Agboola: So potentially what we can do with that is… I know, like, data comes into the warehouse every night, so it’s only once every 24 hours. Potentially, we can actually look at increasing the frequency on that end of things. So now the data that we’re looking at is not just a one

83 00:08:47.880 ⇒ 00:08:58.670 Demilade Agboola: time, like, once in the last 24-hour snapshot, we can start to have, like, once every 6-hour snapshots of the data based off of the ingestion and the dbt run pipelines.

84 00:08:58.830 ⇒ 00:09:04.669 Demilade Agboola: So that way, we can have our dashboard stay up to date without having to, like, manually run it.

85 00:09:08.220 ⇒ 00:09:15.620 Uttam Kumaran: Yeah, can we show… yeah, you want to pull this up, the actual dashboard? Because, you know, I was looking at this yesterday and today, and I think this isn’t a lot…

86 00:09:16.480 ⇒ 00:09:18.619 Uttam Kumaran: I feel like we’ve made a lot of progress.

87 00:09:18.910 ⇒ 00:09:24.400 Uttam Kumaran: sort of, it’s all, like, kind of happening at once, so I feel like, without sort of spying on…

88 00:09:24.700 ⇒ 00:09:31.340 Uttam Kumaran: what’s in GitHub. I feel like, we worked on a lot of modeling that sort of brought these pieces together. In particular, if you just…

89 00:09:31.550 ⇒ 00:09:41.589 Uttam Kumaran: go at the top of this dashboard, Demi. Like, the biggest piece here is when you talk about, like a, you know, subscription SaaS business.

90 00:09:41.650 ⇒ 00:09:55.400 Uttam Kumaran: It’s really about flows of revenue between… within categories, right? So, if you think about a dollar, a dollar may flow in from a new customer, that customer may upgrade, that customer may then downgrade, that customer may then churn.

91 00:09:55.400 ⇒ 00:10:02.300 Uttam Kumaran: that customer may then come back, like, after a churn period. So all of those different flows, we’re trying to

92 00:10:02.480 ⇒ 00:10:16.950 Uttam Kumaran: give that, because typically the way SaaS businesses are orchestrated is you have teams that are associated with one of those flows, right? You have, like, a retention team, a growth team, you have customer success, and so it’s actually helpful to say, like, yes, we have 3 million of ARR, but we actually

93 00:10:17.000 ⇒ 00:10:35.959 Uttam Kumaran: $500, 000 o f t ha t i s n e t n e w, b u tw e ha d 600, 000 o f c h u r n, w e ha d an o t h er$ 300,000 of expansion, and so, yes, you do want the number to go up and to the right, but looking at the pieces, basically, like, the contribution is what is, like, essential here, and so if you just click on the components on the right side.

94 00:10:36.120 ⇒ 00:10:54.410 Uttam Kumaran: Demi, you can see that we’ve sort of categorized how much ARR was churned, how much was due to expansion, how much is net new, and so you now have the ability, and these are just some of the sort of views we put together, but typically the way we’ve seen this done is, like.

95 00:10:54.710 ⇒ 00:11:06.800 Uttam Kumaran: okay, I want to… well, of course, you want to mitigate churn, but really, it’s like, you’re gonna expect some amount of churn. So, churn is a percentage of ARR. What is our goal for that, and are we in line with that? And so.

96 00:11:07.320 ⇒ 00:11:11.759 Uttam Kumaran: really, I think, like, the… the graph at the bottom is the, like.

97 00:11:12.000 ⇒ 00:11:27.350 Uttam Kumaran: is really, I think, the screenshot of, like, okay, what are all the pieces that we’re trying to nail? And then, basically, as default grows, what are the contributing pieces? Did we expand a new client? Did we mitigate churn? Did we get a lot of new business coming in?

98 00:11:27.610 ⇒ 00:11:31.609 Uttam Kumaran: things like that. So, this is still in QA, and I think we’re working with

99 00:11:32.100 ⇒ 00:11:45.759 Uttam Kumaran: Laura, like, a couple things we’ll highlight is, like, what is our definition of churn and equals? It was a little bit different than what we were defining it as, and so there’s, like, some pieces, so I know there’s some zeros and things like that, but, like, this is the view we’re driving towards.

100 00:11:46.430 ⇒ 00:11:54.029 Caitlyn Vaughn: Okay, wait, this is so awesome. This is so helpful to see all of this. Something else that I…

101 00:11:54.380 ⇒ 00:11:58.540 Caitlyn Vaughn: saw from… Brow… I don’t know if you know BrowserBase.

102 00:11:58.540 ⇒ 00:11:59.140 Uttam Kumaran: Yes.

103 00:11:59.140 ⇒ 00:12:13.640 Caitlyn Vaughn: they just sent their investor update, and on their email, they had a really cool breakdown of their revenue per month and where it came from. So for them, it was, like, self-serve versus sales-led versus.

104 00:12:13.640 ⇒ 00:12:14.380 Uttam Kumaran: Okay.

105 00:12:14.380 ⇒ 00:12:21.449 Caitlyn Vaughn: expansion versus churn or whatever, which I thought was really helpful, and we’re gonna have…

106 00:12:21.560 ⇒ 00:12:40.549 Caitlyn Vaughn: maybe this is, like, a separate conversation, but we’re essentially gonna have, like, a handful of different revenue streams, including self-serve, velocity, aka mid-market, and enterprise. Probably, like, our biggest three, but there’s, like, a few ways that we could probably break this down that would be super helpful for us.

107 00:12:40.550 ⇒ 00:12:45.519 Uttam Kumaran: If you have a screenshot of that, or if you even want to sketch it out, like, we’ll just… because we have all the components now.

108 00:12:46.480 ⇒ 00:13:00.750 Caitlyn Vaughn: Very cool. We have a, like, a financial goals sheet that Laura built out, so maybe on the next call, Demi, we can, like, go through that sheet and break down all the, like, different components of how we’re, recognizing revenue.

109 00:13:01.290 ⇒ 00:13:08.780 Demilade Agboola: Okay, yeah, definitely. I know that Laura also has some of these numbers, like, available to her, and that’s part of why I want to have the QA call with her.

110 00:13:08.940 ⇒ 00:13:18.799 Demilade Agboola: So we can analyze, like, hey, this… because I noticed, for instance, certain customers… were… appeared to be… churned.

111 00:13:20.210 ⇒ 00:13:29.140 Demilade Agboola: But, like, my model didn’t counter much churn. I went into Hyperline, it’s… they were on churn. And, like, those sort of, like, small definitions here and there.

112 00:13:29.280 ⇒ 00:13:37.589 Demilade Agboola: So just being able to just align and be clear on, like, okay, so what is the ultimate source of truth on, like, churn? You know, do we…

113 00:13:38.370 ⇒ 00:13:47.709 Demilade Agboola: like, how do we want to look at it? I know we’re trying to use a lot of, like, Salesforce data as the source of truth for a lot of these things, but I also noticed that, like.

114 00:13:48.550 ⇒ 00:13:52.929 Demilade Agboola: Hyperline seem to have more precise data in terms of, like.

115 00:13:52.930 ⇒ 00:13:53.660 Caitlyn Vaughn: Hmm.

116 00:13:53.660 ⇒ 00:13:59.410 Demilade Agboola: People’s subscriptions, because if someone has canceled the subscription pipeline, it shows that this has been canceled.

117 00:13:59.730 ⇒ 00:14:00.090 Caitlyn Vaughn: Yeah.

118 00:14:00.090 ⇒ 00:14:02.289 Demilade Agboola: We are reflected in Salesforce.

119 00:14:02.290 ⇒ 00:14:16.509 Caitlyn Vaughn: Salesforce. Yeah, I think that’s a good shout. I feel like that’s probably a good thing for us to push for. The only issue… well, I imagine the pushback is going to be, how can we get that data reflected better into Salesforce? Because everybody has access.

120 00:14:16.510 ⇒ 00:14:17.100 Uttam Kumaran: Yeah.

121 00:14:17.100 ⇒ 00:14:17.570 Caitlyn Vaughn: First.

122 00:14:17.570 ⇒ 00:14:22.439 Uttam Kumaran: So where this typically happens is, like, there’s always gonna be, difference

123 00:14:22.710 ⇒ 00:14:27.239 Uttam Kumaran: I’m gonna expect the difference between close a book for finance versus operational.

124 00:14:27.240 ⇒ 00:14:30.879 Caitlyn Vaughn: Right? Like, you need to do stuff day-to-day, but, like, finance is gonna be, like.

125 00:14:30.940 ⇒ 00:14:34.570 Uttam Kumaran: Until 2 weeks after the month, where we close everything, that month…

126 00:14:34.750 ⇒ 00:14:38.570 Uttam Kumaran: is not reflected. So this is, like, an expected bifurcation.

127 00:14:38.570 ⇒ 00:14:40.119 Caitlyn Vaughn: Okay. And it’s gonna be something that…

128 00:14:40.120 ⇒ 00:14:48.349 Uttam Kumaran: constantly comes up, so… I think I just want to warn everybody that this is just, like, what happens, is finance is gonna say, like, they’re gonna have an accounting definition.

129 00:14:48.460 ⇒ 00:14:58.529 Uttam Kumaran: Of, like, booked revenue and booked expense, versus we’re gonna need to see money come in and, like, be able to make operational decisions on a daily basis, even if it’s not, like.

130 00:14:58.650 ⇒ 00:15:03.350 Uttam Kumaran: gap accounted close of books, right? Because then we’d be waiting 6 weeks to even understand.

131 00:15:03.350 ⇒ 00:15:04.270 Caitlyn Vaughn: So… Yeah.

132 00:15:04.270 ⇒ 00:15:20.010 Uttam Kumaran: that’s the thing I want to separate, is that, like, there will be some teams for which, yes, because they’re not only bonused on Salesforce, that is their source of truth, they live there, they’re gonna be like, I don’t care what happens in Hyperline, like, that’s not… that’s, like, Laura’s thing, and vice versa. Laura’s gonna say.

133 00:15:20.010 ⇒ 00:15:24.290 Uttam Kumaran: Well, Salesforce, like, we don’t get paid if someone… like, it’s based on the contract, so…

134 00:15:24.360 ⇒ 00:15:33.319 Uttam Kumaran: that’s gonna happen. So I think the two things we want to just understand is, like, for these dashboards, who is the stakeholder? Right? So if this is something that is going…

135 00:15:33.730 ⇒ 00:15:47.169 Uttam Kumaran: to both, like, the entire company, then we need to make sure both of those parties and us are all on the same page, that, like, Laura is comfortable with the fact that these are operational metrics that are going to be used to run the business day-to-day.

136 00:15:47.240 ⇒ 00:15:57.419 Uttam Kumaran: And then we’re also working on dashboards for her that are purely reflective of QuickBooks, right? And QuickBooks is gonna be the source of truth for, like, finance close of… you know, closing things out.

137 00:15:57.420 ⇒ 00:15:59.350 Caitlyn Vaughn: Yeah. And we’ve satisfied…

138 00:15:59.350 ⇒ 00:16:16.490 Uttam Kumaran: that she’s both okay with the fact that people are making decisions based on, like, non-booked revenue, but they have to. And then for sales, like, yes, they’re gonna always ask for Salesforce as the source of truth, and so it’s gonna be less of, like, pick one, it’s actually, like, you have a dashboard that serves your need.

139 00:16:16.490 ⇒ 00:16:26.250 Uttam Kumaran: and you are just aware of the fact that we are supplying sales with dashboard that has money on it, and Laura is, like, comfortable with that. It’s like.

140 00:16:26.790 ⇒ 00:16:28.449 Uttam Kumaran: That’s how to address that.

141 00:16:29.040 ⇒ 00:16:32.590 Uttam Kumaran: because that’s… this is just, like, the way it works, right? Like…

142 00:16:32.590 ⇒ 00:16:33.370 Caitlyn Vaughn: Yeah.

143 00:16:35.580 ⇒ 00:16:41.410 Caitlyn Vaughn: That’s interesting. Yeah, we have, obviously, like, paying out sales commission based on meetings booked and revenue.

144 00:16:41.410 ⇒ 00:16:42.170 Uttam Kumaran: Yes.

145 00:16:42.170 ⇒ 00:16:50.829 Caitlyn Vaughn: But then the way that we actually recognize revenue from an accounting standpoint has changed some of the way that we have built out the product, so…

146 00:16:50.830 ⇒ 00:16:51.650 Uttam Kumaran: Yeah.

147 00:16:51.650 ⇒ 00:16:52.960 Caitlyn Vaughn: two very different…

148 00:16:53.730 ⇒ 00:17:01.750 Uttam Kumaran: And this is always, like, and again, finance is gonna be very, very cons… these are two, like, very opposite parties, like, a very conservative, close the book, everything needs to tie up, versus sales is, like.

149 00:17:02.930 ⇒ 00:17:05.890 Uttam Kumaran: put the opportunity in, it’s like, booked the… right, so…

150 00:17:05.890 ⇒ 00:17:06.579 Caitlyn Vaughn: Endless one.

151 00:17:06.589 ⇒ 00:17:13.459 Uttam Kumaran: We are the people… but see, of course, like, this conversation didn’t even happen until we’re starting to be like, let’s look at everything.

152 00:17:14.000 ⇒ 00:17:33.429 Uttam Kumaran: So, I think the way we’re going at it is, like, we’re giving Laura a lot of space to dictate, like, her dashboards, but also indicating… I think we need to just indicate to her that we are going to put ARR metrics, but for the sales and for the broad company, they should know that this is not, like, the accounting certified number.

153 00:17:33.440 ⇒ 00:17:36.010 Caitlyn Vaughn: That’s what I want to be really cautious of, is like.

154 00:17:36.140 ⇒ 00:17:44.359 Uttam Kumaran: these are things that are gonna help the business, but Laura and her reporting is the gatekeeper for, like, actually how much money hit the bank account and got processed.

155 00:17:45.330 ⇒ 00:17:48.600 Uttam Kumaran: like, I will tell you that in working in SaaS for a while.

156 00:17:49.070 ⇒ 00:17:53.069 Uttam Kumaran: people report on just, like, what they’re comfortable with. So, like.

157 00:17:53.070 ⇒ 00:17:53.450 Caitlyn Vaughn: Yeah.

158 00:17:53.450 ⇒ 00:18:03.960 Uttam Kumaran: for example, like, what you’re seeing in the browser base, you don’t… it’s not clear whether those are, like, things that they’ve actually, like, GAP certified, or those are just, like, they added up everything in Salesforce. It’s not… there’s not, like, a standard.

159 00:18:04.690 ⇒ 00:18:13.720 Uttam Kumaran: It gets closer and closer as the company gets bigger, and because things start to, like, slow down, you have time, and the close of book doesn’t take as long, but this is, like, how it typically is, you know?

160 00:18:14.780 ⇒ 00:18:23.169 Uttam Kumaran: So, we will go, I think, for the… for anything that involves Salesforce, we will start to use Salesforce to match expected ARR,

161 00:18:23.350 ⇒ 00:18:26.570 Uttam Kumaran: 2, and, like, look at all the components of usage.

162 00:18:26.570 ⇒ 00:18:28.169 Caitlyn Vaughn: Yeah. And we will also…

163 00:18:28.170 ⇒ 00:18:31.819 Uttam Kumaran: for Laura, have her sign off on the fact that, like.

164 00:18:32.520 ⇒ 00:18:37.099 Caitlyn Vaughn: We are implementing the definitions the same way, like, what is churn, things like that.

165 00:18:37.100 ⇒ 00:18:45.209 Uttam Kumaran: And then she’s also gonna have a dashboard that has all the expenses and revenue from QuickBooks, which is the finance source of truth for accounting, you know.

166 00:18:45.240 ⇒ 00:18:46.889 Caitlyn Vaughn: Yeah, that makes sense.

167 00:18:50.390 ⇒ 00:18:57.329 Uttam Kumaran: Yeah, I agree. And then if you want to scroll down, I think, Demi, you can just share, like, basically we’re…

168 00:18:57.640 ⇒ 00:19:12.560 Uttam Kumaran: you then get a lot of fun on, like, what we could do with the customer data. So you can show, like, which customers, like, how our ACV is changing, our ACV is new customers. I think if you scroll slightly up, that green and red chart is actually also really good.

169 00:19:12.660 ⇒ 00:19:14.780 Caitlyn Vaughn: So you can start to see, like.

170 00:19:14.780 ⇒ 00:19:19.900 Uttam Kumaran: Churn… how many customers churned in a month versus restarted versus our net new?

171 00:19:20.020 ⇒ 00:19:22.939 Caitlyn Vaughn: And so, again, like, really, these are, like, the flows.

172 00:19:23.100 ⇒ 00:19:30.240 Uttam Kumaran: I think all the lines of default are going up in the right, so really the nuance is in the flows, you know?

173 00:19:30.960 ⇒ 00:19:37.779 Uttam Kumaran: And so, like, yes, sales sold a bunch of stuff, but then they were all crap, because they all churned really quickly, right? So this is the things that allow you to…

174 00:19:38.580 ⇒ 00:19:44.769 Uttam Kumaran: to see that. And I think a lot of the stuff that we were originally planning with Deanna on is really the, like, account management, like.

175 00:19:45.040 ⇒ 00:19:55.150 Uttam Kumaran: like, we want to… eventually, we can start doing things like cohorting. Is the average default customer who started this month versus 6 months ago worth more over time?

176 00:19:55.230 ⇒ 00:20:07.209 Uttam Kumaran: So, like, are we expanding them quicker? Like, are they… are they using more things in the first 3 months than they did 6 months? And this is where, Nandika, this actually gets… this is the link, then, to the product analytics work, you know?

177 00:20:07.360 ⇒ 00:20:09.070 Nandika Jhunjhunwala: For sure, yes.

178 00:20:09.880 ⇒ 00:20:34.629 Nandika Jhunjhunwala: like, that was, like, something that I also wanted to talk about, and these are conversations I’ve been having with Greg. Like, he’s been bringing up, like, super important points that, like, has been forcing me to think, like, what data do we want available from post-hoc and Omni, if any, and then do you want any data from Omni into Post Hog, again, if any. So, I think those are questions that we definitely need to, like, answer, eventually, or, like, decide on, like, how we want to, like.

179 00:20:35.300 ⇒ 00:20:37.860 Nandika Jhunjhunwala: Create those, like, dashboards and, like, what…

180 00:20:38.040 ⇒ 00:20:43.440 Nandika Jhunjhunwala: Where, like, joining these two data sources would create leverage for us, for sure.

181 00:20:44.410 ⇒ 00:20:46.880 Uttam Kumaran: Yeah, I agree. And I think this is where, like.

182 00:20:47.010 ⇒ 00:20:51.720 Uttam Kumaran: I would recommend doing as much as you can in posthog, and then bringing the pieces you need.

183 00:20:51.800 ⇒ 00:20:53.479 Nandika Jhunjhunwala: Right? To connect.

184 00:20:53.480 ⇒ 00:20:55.090 Uttam Kumaran: And I think that’s probably what Greg…

185 00:20:55.910 ⇒ 00:20:59.879 Uttam Kumaran: It’s probably what Greg said, too, I think so.

186 00:21:02.460 ⇒ 00:21:09.560 Demilade Agboola: Yeah, so, like, again, like, work is definitely going on behind the scenes to get these numbers out to you.

187 00:21:10.840 ⇒ 00:21:21.110 Demilade Agboola: And the idea, again, of, like, the QA sessions, once the dashboards are ready, is, we want to be able to get feedback on how these numbers work for each of the individual stakeholders.

188 00:21:21.210 ⇒ 00:21:27.929 Demilade Agboola: So we’re sure that whatever they’re using, they feel comfortable with it, and they can trust it on a day-to-day basis.

189 00:21:28.060 ⇒ 00:21:32.650 Demilade Agboola: And… That is…

190 00:21:32.930 ⇒ 00:21:38.459 Demilade Agboola: so in terms of other, like, just, like, timelines and how things have been so far.

191 00:21:39.380 ⇒ 00:21:46.529 Demilade Agboola: We have work… Like, we’re currently working in Laura’s flow, as well as Lauren’s flow.

192 00:21:46.920 ⇒ 00:21:50.830 Demilade Agboola: So right now, that involves…

193 00:21:50.930 ⇒ 00:21:52.829 Demilade Agboola: Building out the dashboards for them.

194 00:21:53.080 ⇒ 00:21:57.410 Demilade Agboola: I know Lauren’s dashboard should be out by tomorrow, all things being equal.

195 00:21:57.650 ⇒ 00:22:00.139 Demilade Agboola: So we can have, like, a call with…

196 00:22:00.260 ⇒ 00:22:03.759 Demilade Agboola: Stakeholders early next week, and have an idea of that.

197 00:22:03.930 ⇒ 00:22:10.979 Demilade Agboola: In terms of, like, the expected delivery of Laura’s flow, it was expected that we got that out earlier this week.

198 00:22:12.280 ⇒ 00:22:16.959 Demilade Agboola: Yes, so we can get that.

199 00:22:17.870 ⇒ 00:22:21.380 Demilade Agboola: in terms of, like, the rest that was supposed to be gotten out earlier this week.

200 00:22:21.710 ⇒ 00:22:29.719 Demilade Agboola: I think one of the huge delays to that was the, access to equals was gotten, like, last week, Friday, where, like, some of the definitions were.

201 00:22:30.020 ⇒ 00:22:44.100 Demilade Agboola: So, like, early this week was just basically trying to ensure the models aligned with expectations of what churn is. I also had a call with Laura last week where we talked about what renewal counts as. So, for instance, the renewal dates

202 00:22:44.620 ⇒ 00:22:59.509 Demilade Agboola: isn’t actually the renewal date, it’s 6 weeks after, because sometimes people are in the process of renewing, and the renewal date goes over, so that’s not necessarily a renewal, that’s just them, you know, continuing on with the same plan.

203 00:22:59.750 ⇒ 00:23:02.030 Demilade Agboola: And then…

204 00:23:02.290 ⇒ 00:23:12.629 Demilade Agboola: Just being able to align on that, and so restart, churn, and all those definitions created, like, a little extra delay in terms of being able to get out the dashboard.

205 00:23:13.000 ⇒ 00:23:19.619 Demilade Agboola: But then, you know, takeaway from this is that, ultimately, the modeling will take quite a bit of time.

206 00:23:19.860 ⇒ 00:23:24.789 Demilade Agboola: potentially… 3 weeks, 3 to 4 weeks per phase.

207 00:23:25.040 ⇒ 00:23:28.249 Demilade Agboola: We will try to get across as quickly as possible.

208 00:23:28.460 ⇒ 00:23:37.359 Demilade Agboola: But sometimes these things will just take time, especially based off of feedback from the stakeholders on the numbers they see, how they align with what they want to do.

209 00:23:37.520 ⇒ 00:23:45.069 Demilade Agboola: And so, like, modeling and dashboarding could take, like, 3 weeks, 3 to 4 weeks, and ultimately, it’s one of those things where…

210 00:23:45.610 ⇒ 00:24:00.510 Demilade Agboola: when the numbers get closer, you start to get more feedback, and that more feedback creates more work, right? Like, I’m pretty sure when I leave the call with Laura today, we’ll get quite a bit of feedback on certain things she would like to see, or, like, us to change definitions of.

211 00:24:00.700 ⇒ 00:24:12.480 Demilade Agboola: And that would also create more work for, like, next week in terms of, you know, delivering things. So just being able to hold those multiple, things all at once was just very important as we’re going forward in this… in this work.

212 00:24:13.850 ⇒ 00:24:25.780 Caitlyn Vaughn: Cool. So, question on the first piece here, metric definition and alignment. Was that supposed to be all of the schemas for the different dashboards with sign-offs from each stakeholder?

213 00:24:26.320 ⇒ 00:24:32.640 Demilade Agboola: So ideally, yeah, it’s a definition… it’s not necessarily, like, the schemas, per se, it’s more of the,

214 00:24:32.850 ⇒ 00:24:35.880 Demilade Agboola: what is in… your dashboard?

215 00:24:36.890 ⇒ 00:24:44.070 Demilade Agboola: possible, like, what do you need to see in your dashboard? And as best as possible, how would you define that? Ultimately, some of these things are harder to define without, like.

216 00:24:44.630 ⇒ 00:24:46.290 Demilade Agboola: crystal examples.

217 00:24:46.490 ⇒ 00:25:00.259 Demilade Agboola: And so it’s possible, like, hey, I want to see churn, I want to see this, I want to see that. But until the numbers start coming out, where you’re like, oh, this isn’t how I want to see churn, or this isn’t, like, what a renewal looks like.

218 00:25:01.930 ⇒ 00:25:04.749 Demilade Agboola: That, like, process starts to evolve, where you have.

219 00:25:05.570 ⇒ 00:25:07.690 Demilade Agboola: You know, it’s an iterative process.

220 00:25:07.880 ⇒ 00:25:19.840 Demilade Agboola: So yeah, so, like, a high-level thing of, like, hey, this is what the dashboard will look like, these are the numbers you need to see, this is… I want to see, you know, burn rate, I want to see revenue, I want to see this, I want to see ARR, like, those are, like, the high-level things.

221 00:25:19.840 ⇒ 00:25:22.209 Uttam Kumaran: So is this what we put… is this the document we put?

222 00:25:22.430 ⇒ 00:25:25.690 Uttam Kumaran: with… put together with Laura, like, the Google Doc.

223 00:25:26.210 ⇒ 00:25:28.880 Demilade Agboola: Yeah, so that’s… that would be… that would be the high level.

224 00:25:29.300 ⇒ 00:25:36.569 Demilade Agboola: Dean, so… We can, like, we can model, so an example would be, we have booked revenue.

225 00:25:37.100 ⇒ 00:25:40.569 Demilade Agboola: you can model the booked revenue, turn it across to Laura.

226 00:25:40.740 ⇒ 00:25:53.350 Demilade Agboola: Laura looks at it, tries to understand what’s going on, and says, no, actually, you’re ignoring this category, or there’s a certain thing you might need to integrate into the booked revenue. So that can potentially…

227 00:25:55.070 ⇒ 00:26:05.799 Demilade Agboola: Take… allow us to take a little extra time, because we’re, like, trying to figure out how to ensure that the numbers that everyone gets are what they need on a day-to-day, and integrate every single thing across board.

228 00:26:06.100 ⇒ 00:26:15.379 Caitlyn Vaughn: Yeah, I totally get that. I think the metric, definition, and alignment, when I was first looking at this, I was assuming that it was, like, at the start of each

229 00:26:16.680 ⇒ 00:26:35.240 Caitlyn Vaughn: dashboard creation, but there’s actually just one at the top. So now I’m reading this as, like, all of the dashboards should have been defined, and then feedback should have been given on them, right? Because we marked that as completed, but there’s only two out right now. There’s CS, and there’s the LoRa financial one.

230 00:26:35.950 ⇒ 00:26:37.090 Demilade Agboola: Yeah, so…

231 00:26:37.350 ⇒ 00:26:52.280 Demilade Agboola: Ideally, yeah, we’re trying to… so the metric alignment and definition is… is going to happen for each dashboard. Ultimately, the idea, and why it’s not necessarily explicitly put for every other single flow, which it’s going to be a part of, is that

232 00:26:53.640 ⇒ 00:27:09.790 Demilade Agboola: as that’s happening, the concept of, like, we’re modeling here, that can happen in parallel, quote-unquote, without initially being an explicit thing, around it. But yes, also something to integrate in that sense of, we’ll be getting that, and that’s why.

233 00:27:09.790 ⇒ 00:27:11.519 Uttam Kumaran: Yeah, I think that’s where I’m also, like.

234 00:27:11.580 ⇒ 00:27:25.129 Uttam Kumaran: this is where I think, Mustafa, we just need to reflect it. So if there is work going on for each of these work streams around alignment, like, I kind of separate it into two things. Like, I think, Demi, you went and met everybody, like, you have a relationship with Stan, with Lev.

235 00:27:25.130 ⇒ 00:27:32.199 Uttam Kumaran: with Laura now, right? With all the core people we’re building towards. So part of that is, like, yes, metric definition, there’s also just, like.

236 00:27:32.390 ⇒ 00:27:43.149 Uttam Kumaran: you need… they need to know that, like, kind of, like, what we’re doing here, and, like, that you’re gonna be approaching them. For each of these, though, I think we need to bake in two changes. One, I think I want to make it clear if, like.

237 00:27:43.360 ⇒ 00:28:03.210 Uttam Kumaran: if we are gonna do more, like, work to put together that type of document for every single one, we should indicate it on the chart. And then second, I think we just, like, over… I think we were just overzealous on how long, I think, the data modeling is going to take for each. Because if each of these folks comes back with.

238 00:28:03.390 ⇒ 00:28:17.319 Uttam Kumaran: hey, actually, like, now that I looked at the data, it’s basically, like, a now that I looked at the data, actually, like, I want to redefine it type situation. Yeah. So, we try to phase it out, where it’s like, can we drive towards a first dashboard? Yes, there are still, like, I need new views, or…

239 00:28:17.350 ⇒ 00:28:26.209 Uttam Kumaran: I’m not perfectly comfortable, but I need some time to think about it. We can, like… okay, that’s, like, phase two. That’s what we’re trying to drive towards, and I think we’re gonna be there

240 00:28:27.400 ⇒ 00:28:40.310 Uttam Kumaran: if you want to just… can you just demigo? I don’t know if we have this on the next slide, but actually, we can say here. Basically, we’re gonna try to drive towards finance… being there for finance, and customer, which is, like, basically Salesforce stuff for Lev.

241 00:28:40.450 ⇒ 00:28:41.619 Uttam Kumaran: by next week.

242 00:28:41.780 ⇒ 00:28:47.209 Uttam Kumaran: Binance, looking at kind of the point today, and I think looking at how far we’ve gone, I think

243 00:28:47.380 ⇒ 00:28:53.680 Uttam Kumaran: we’re gonna basically be able to have a V1 of that dash tomorrow. There’s just formatting and a couple of tweaks that we need to do.

244 00:28:53.870 ⇒ 00:28:55.609 Uttam Kumaran: And then the Salesforce data.

245 00:28:55.760 ⇒ 00:29:05.380 Uttam Kumaran: really, the point there is, like, we’re just gonna use the one-time exports from the product data, marry that with Salesforce, and then drive towards the first version of

246 00:29:05.490 ⇒ 00:29:06.870 Uttam Kumaran: of that dashboard, too.

247 00:29:06.870 ⇒ 00:29:07.319 Caitlyn Vaughn: the, the.

248 00:29:07.320 ⇒ 00:29:10.180 Uttam Kumaran: customer side, that’s really the ask for, like, Lev.

249 00:29:10.360 ⇒ 00:29:23.339 Caitlyn Vaughn: Yeah, I think that, like, I know you guys are blocked on a couple of things, but there is still plenty of things that could be done in the meantime, you know, like, work on other work streams. Yeah.

250 00:29:23.340 ⇒ 00:29:32.500 Caitlyn Vaughn: And same with even as I’m thinking about this, like, metric definition and alignment, like, we talked about it last week, it’s, like, one of those things where if something else is blocked, like.

251 00:29:32.500 ⇒ 00:29:43.230 Caitlyn Vaughn: we should have all of the high-level dashboards from every person of, like, this is what we’re going to build, does this look good? Even if they’re like, yes, and then later they’re like, oh, actually, I need to change some stuff, like…

252 00:29:43.230 ⇒ 00:29:53.319 Caitlyn Vaughn: that would be really helpful, and even getting to the first draft of the dashboard that you guys have basically created for Laura, like, that is super helpful. To me, that’s, like.

253 00:29:53.950 ⇒ 00:30:12.369 Caitlyn Vaughn: you know, the first version out, where people are like, okay, you’re doing work, you’re building towards the thing that I thought that you were, and now I see the value out of it, and it’s totally fine if they want to, like, change it, and you guys need to, you know, spend more time doing that on a… on a single dashboard, but I think those are, like, the two key places where it’s, like.

254 00:30:12.490 ⇒ 00:30:25.359 Caitlyn Vaughn: customer-facing, aka us-facing, right? That would be, like, helpful metrics for us to see, like, now we see the progress being made, because you guys are going and doing all this stuff in the back end, which we aren’t seeing, right?

255 00:30:25.360 ⇒ 00:30:25.790 Uttam Kumaran: Yeah, yeah, yeah.

256 00:30:25.790 ⇒ 00:30:32.610 Caitlyn Vaughn: Probably well we ever, hopefully. But, yeah, yeah, those are, like, the two places where we.

257 00:30:32.610 ⇒ 00:30:37.690 Uttam Kumaran: Yeah, so on that point, yeah, I think we’re… when we say blocked, it’s not like, okay, we’re just…

258 00:30:37.740 ⇒ 00:30:50.690 Uttam Kumaran: we’re chillin’. I think we are going to the next thing. I think we can… we should do a better job of showing, like, what that is. Yeah. And I think on the data modeling side, really what that is, is, like, I sort of chatted with Demi, and I was like.

259 00:30:50.740 ⇒ 00:31:08.760 Uttam Kumaran: we… so just… we’re just gonna drive… like, I think the polyatomic and ingestion, we’re gonna… it’ll figure… it’ll figure itself out, but in the meantime, like, I want to just drive with one-time exports wherever we can do, so we can get dashboards running. I think also, Caitlin, that’s just gonna raise the noise on, like, making that possible, because

260 00:31:09.060 ⇒ 00:31:13.029 Uttam Kumaran: The more pipelines we can build, the faster those can refresh.

261 00:31:13.630 ⇒ 00:31:21.439 Uttam Kumaran: So, I basically said, look, I don’t want us waiting on imports or the perfect pipelines to just limit getting these dashboards out.

262 00:31:21.440 ⇒ 00:31:23.830 Caitlyn Vaughn: Yeah. Even for, for Email Bison.

263 00:31:23.830 ⇒ 00:31:41.600 Uttam Kumaran: for any of our new sources, just get the data, model it, and then get the dashboard out, and then we will say, like, hey, the fact that, like, we don’t have a pipeline means we can’t update this daily, or, like, means that there’s, like, some limitations. And that’s, like, I think a good place to then have that negotiation, right? Versus, like…

264 00:31:41.640 ⇒ 00:31:44.720 Uttam Kumaran: So far early, where there’s not something you could see, you know?

265 00:31:44.720 ⇒ 00:32:00.380 Caitlyn Vaughn: Yeah. Yeah, exactly. The other thing is, we are probably gonna churn on Catalyst, so I don’t think we need the reverse ETL stuff. I also think we should, like, hunt reverse ETL until we’re finished with this project, so all of those, like.

266 00:32:00.570 ⇒ 00:32:09.629 Caitlyn Vaughn: associated hours with that can be put back into, like, finishing dashboards and making sure we’re, like, getting the foundation of this project done.

267 00:32:10.340 ⇒ 00:32:14.720 Caitlyn Vaughn: Yeah, I think we briefly spoke about the reverse ETL stuff, which I think.

268 00:32:14.720 ⇒ 00:32:22.510 Uttam Kumaran: Yeah, we have a slide we’ll talk about today, so maybe, Demi, I think… I think if we’re good on this, I think we have some changes on the Gantt to make that we can…

269 00:32:22.850 ⇒ 00:32:24.509 Uttam Kumaran: Make it get back to you.

270 00:32:25.610 ⇒ 00:32:38.270 Nandika Jhunjhunwala: I had a quick question on, the workflow… workstream you just discussed. So, are you talking about doing one-time exports again off of, like, Salesforce, Hyperline, and so on for…

271 00:32:38.270 ⇒ 00:32:39.630 Demilade Agboola: sources we already have, so for.

272 00:32:39.630 ⇒ 00:32:39.980 Nandika Jhunjhunwala: Okay.

273 00:32:39.980 ⇒ 00:32:47.890 Demilade Agboola: we have, that’s fine. Yeah. But things, like, we don’t have, like, Postgres access, and, like, that’s blocking the view to put things together.

274 00:32:47.890 ⇒ 00:32:49.929 Uttam Kumaran: And Thomas did that for us before, so we’ll just…

275 00:32:50.080 ⇒ 00:32:53.560 Uttam Kumaran: Just redo that, you know, we’ll just re-get that and, yeah, keep pushing.

276 00:32:54.710 ⇒ 00:33:12.899 Nandika Jhunjhunwala: Then my question then would be, like, once we have those dashboards off of one-time exports, like, I know there’s, like, some Salesforce topics in Omni that already exist off of the one-time exports. How easy is it for you, or, like, how seamless is it to, like, move from, like, those exports to, like, the pipeline?

277 00:33:12.900 ⇒ 00:33:16.500 Uttam Kumaran: Good follow-up question. Sort of depends, I think.

278 00:33:16.500 ⇒ 00:33:16.950 Demilade Agboola: Yeah.

279 00:33:16.950 ⇒ 00:33:19.969 Uttam Kumaran: Like, I think it’s just gonna depend on how…

280 00:33:20.790 ⇒ 00:33:31.249 Uttam Kumaran: Like, again, things like renaming columns is one thing, but if the overall structure is different, or it’s like, we now move from one table to, like, which then join 10 tables.

281 00:33:31.430 ⇒ 00:33:41.539 Uttam Kumaran: Either way, it’s gonna be work. I think, like, we’re… we’re gonna try our best to, like, do that sort of switchover, but still, I think given how fast default is moving, it’s… in our lens.

282 00:33:41.730 ⇒ 00:33:47.290 Uttam Kumaran: Like, if we guys were moving slower, then we’d be like, okay, let’s… Try to nail that first.

283 00:33:47.410 ⇒ 00:33:54.879 Uttam Kumaran: I think we… we are sitting on this data, we need to get these dashboards out, it’s worth that risk, and it’s not like…

284 00:33:55.090 ⇒ 00:33:57.810 Uttam Kumaran: oh, it’s gonna take 2 months, like, I think it’ll just add, like.

285 00:33:58.090 ⇒ 00:34:02.819 Uttam Kumaran: Probably another, like, 10 hours of work to just transition it over once we get the data landed, you know?

286 00:34:03.280 ⇒ 00:34:14.280 Demilade Agboola: Yeah, I think it just depends on the source. So, like, for Postgres, I would expect them to be quite similar, the exports, because it’s the same tables, but once you’re dealing with things like that.

287 00:34:14.330 ⇒ 00:34:27.880 Demilade Agboola: like, Salesforce versus an API, the data just varies quite a bit, so it’s not just, like, the same names and the same columns, you’ll need to do more joins to be able to get the same sort of table you had before that was in a one-time export, so…

288 00:34:27.880 ⇒ 00:34:31.049 Nandika Jhunjhunwala: So… Then, like, the other question…

289 00:34:31.050 ⇒ 00:34:50.809 Nandika Jhunjhunwala: So, like, when I looked at the Gantt chart, and I saw, like, metric definition and alignment, the way I think that that information was given to you was, like, a document by Laura, where she, like, spoke to different stakeholders for those data sources, and sort of made, like, educated guesses on, like, what those metrics would look like. But I think, like, where I see discrepancy is, like.

290 00:34:50.810 ⇒ 00:34:55.370 Nandika Jhunjhunwala: What metrics they expect to be available, and then what metrics are available.

291 00:34:55.370 ⇒ 00:35:07.270 Nandika Jhunjhunwala: And then what joins we need to create to make those metrics available for each of those dashboards. So I think, like, it would really help to, like, again, inform all these stakeholders that this is the data that

292 00:35:07.270 ⇒ 00:35:17.469 Nandika Jhunjhunwala: We currently have, so we can get closer to the ground truth of, like, what’s possible now, and then we can, you know, ex… like, adjust our expectations and requirements accordingly, and…

293 00:35:17.540 ⇒ 00:35:39.200 Nandika Jhunjhunwala: really, like, align more quickly on, like, what’s possible, and get those dashboards out, because when I speak to the sales team, you know, they have all these ideas of, like, what these dashboards are gonna look like, and then I’m always like, wait a second, like, that’s not… that’s not gonna happen, like, anytime soon. So I think, like, there is, like, a gap there that I, like, encounter quite frequently,

294 00:35:39.200 ⇒ 00:35:48.630 Nandika Jhunjhunwala: Which is why, like, I’ve been asking for those schemas, because I think, like, them having access to those schemas, like, helps them know as well, okay, this is the data that we currently have.

295 00:35:48.630 ⇒ 00:35:49.839 Nandika Jhunjhunwala: In this state.

296 00:35:49.920 ⇒ 00:36:02.089 Nandika Jhunjhunwala: And it’s, like, maybe a pipeline or an export, and they can, like, then wrap their heads around, like, what… where we are, and, like, where we can head, and, like, what they need to do on their end to, like, help you guys get… get over there.

297 00:36:02.090 ⇒ 00:36:08.320 Uttam Kumaran: Yeah, so to kind of give you a sense of where we, like, where… how we do it, we basically do our… we’ve bottled Salesforce.

298 00:36:08.800 ⇒ 00:36:14.169 Uttam Kumaran: And, like, subscription metrics, like, a ton. So we give our first, basically, guess at, like, the…

299 00:36:14.230 ⇒ 00:36:24.879 Uttam Kumaran: benchmark suite of, like, metrics that everybody’s asked for, and then there’s probably some that are like, okay, this works, and then there’s probably a few more to add. So, I would say, like.

300 00:36:24.930 ⇒ 00:36:40.589 Uttam Kumaran: the risk is probably a bit lower than, like, we’re going this direction, and they need all these other metrics. We, for the most part, give our first best pass, knowing what we know about y’all, and then there’s, like, a few more to add, and so one thing we’ll do is we’ll totally show you what’s in Salesforce, but I think

301 00:36:40.760 ⇒ 00:36:53.569 Uttam Kumaran: what’s also helpful is to just… once the topic is there in Omni, you can basically share that. I actually think it’s worthwhile, while you’re meeting with everybody, to just note those down, because I actually think some of those we were probably just, like.

302 00:36:53.680 ⇒ 00:37:12.190 Uttam Kumaran: you know, a few queries away from enabling. And even if we can’t enable in, like, a dashboard, we can still write the query for you to get a one-time export. So I think it’s helpful, actually, to keep just writing, like, so-and-so had this question, is that in the dashboard? Can I answer that via existing model of topic? Or is that something…

303 00:37:12.600 ⇒ 00:37:19.160 Uttam Kumaran: we want to just get out as a query, and then we can move into the model. So, like, that’s how this, like, whole flow is gonna work.

304 00:37:19.520 ⇒ 00:37:31.939 Nandika Jhunjhunwala: That makes sense. Okay. I think, like, specifically for Laura, that would be a good exercise, because I think she’s the one that needs multiple disparate data sources together, more out of everybody else.

305 00:37:34.230 ⇒ 00:37:34.950 Demilade Agboola: Okay.

306 00:37:38.350 ⇒ 00:37:40.650 Demilade Agboola: Otam, do you want to talk about the product analytics?

307 00:37:41.060 ⇒ 00:37:57.579 Uttam Kumaran: Yeah, I guess I feel like I spoke with Greg yesterday before he went off. I think he’s on, like, a Michigan trip with his fam, but he said this is all going super well. I mean, I think… I’m glad overall, like, I’ve been sort of following and talking to him daily, so I’m glad that, like, you know, we started originally, Caitlin.

308 00:37:57.580 ⇒ 00:38:06.880 Caitlyn Vaughn: when I came to New York with, like, Amplitudes, sort of just sitting there and, like, not having this, and I think we’re in a much, much better place. Yeah. I think also, Nandika, part of our goal originally, like, last month was, like.

309 00:38:06.900 ⇒ 00:38:12.490 Uttam Kumaran: not just, like, ripping this together is, like, doing this with you, so I think it’s awesome seeing, like.

310 00:38:12.910 ⇒ 00:38:14.889 Uttam Kumaran: Like, you actually seeing the events.

311 00:38:15.010 ⇒ 00:38:20.499 Uttam Kumaran: And then I think Greg is sort of at your disposal, like, I think he’s… he’s both really knowledgeable about, like.

312 00:38:20.540 ⇒ 00:38:39.179 Uttam Kumaran: what to go after, but there also is just, like, visibility broadly into, like, what people are doing in the platform. I know we’re kind of, like, developing it for default currently, and then Phoenix when it comes out, so that’s always, like, when I call him and ask him, like, he’s like, that’s the friction. I said, that’s just what it is.

313 00:38:39.540 ⇒ 00:38:44.619 Uttam Kumaran: So I think continuing to just make sure that your existing product is tracked.

314 00:38:44.790 ⇒ 00:38:48.799 Uttam Kumaran: And then, basically, understanding what it will take as features come out.

315 00:38:48.920 ⇒ 00:38:57.360 Uttam Kumaran: So I feel like we’re on track there. I think, yeah, the next thing is basically whatever you need out of PostDog to end up joining within Omni to stuff.

316 00:38:57.710 ⇒ 00:39:00.360 Uttam Kumaran: is what I’ll be looking out for, so…

317 00:39:01.290 ⇒ 00:39:12.870 Nandika Jhunjhunwala: Yeah, I think, like, that’s what I’ve been, like, thinking about, is, like, what… do we need, like, additional data in post-hoc? I know that Posthog makes it really easy to, like, make other data available in the platform.

318 00:39:12.870 ⇒ 00:39:13.410 Uttam Kumaran: Yeah.

319 00:39:13.410 ⇒ 00:39:15.270 Nandika Jhunjhunwala: actors as well.

320 00:39:15.270 ⇒ 00:39:17.209 Uttam Kumaran: Typically, it’s like, if you need…

321 00:39:17.410 ⇒ 00:39:22.190 Uttam Kumaran: dollar metrics, or if you need, like, IDs, Or, like, team names?

322 00:39:22.820 ⇒ 00:39:29.169 Uttam Kumaran: So, I think that’s… that’s kind of what… sometimes you’re gonna make a decision, like, should we put this in the event, or should I get it from…

323 00:39:29.430 ⇒ 00:39:31.330 Nandika Jhunjhunwala: Yep, exactly.

324 00:39:31.620 ⇒ 00:39:32.180 Uttam Kumaran: -Oh.

325 00:39:32.820 ⇒ 00:39:36.559 Nandika Jhunjhunwala: Yeah, and to clarify, Phoenix is instrumented.

326 00:39:36.560 ⇒ 00:39:37.180 Uttam Kumaran: Cool.

327 00:39:37.180 ⇒ 00:39:40.220 Nandika Jhunjhunwala: Yeah, our old product is not instrumented at all.

328 00:39:40.220 ⇒ 00:39:41.510 Uttam Kumaran: Okay, okay, okay, cool.

329 00:39:41.510 ⇒ 00:39:42.950 Nandika Jhunjhunwala: Yeah, yeah.

330 00:39:44.130 ⇒ 00:39:46.830 Demilade Agboola: Okay…

331 00:39:47.030 ⇒ 00:39:58.420 Demilade Agboola: So, I mean, we have this in the reverse ETL, and how we’re thinking about, like, defining frequency for Reverse ETL and potentially building a custom connector based on the specifications, but since you’re…

332 00:39:59.760 ⇒ 00:40:06.550 Demilade Agboola: since you’re thinking of being in that for now, I think this slide has less relevance and relevance as things stand right now.

333 00:40:08.180 ⇒ 00:40:19.829 Uttam Kumaran: Can we talk about the reverse ETL? So, I think, Caitlin, we discussed this briefly. Can I just get a set… like, I think for us, I want to know, there’s… if it’s a… if it’s a small, narrow thing, I think we can build this.

334 00:40:20.220 ⇒ 00:40:27.639 Uttam Kumaran: So I do want to know, like, I know you mentioned depriving the whole thing, I think Catalyst is off the table, but there still are things in Salesforce where you’re like.

335 00:40:27.850 ⇒ 00:40:29.329 Uttam Kumaran: We just need a field.

336 00:40:30.650 ⇒ 00:40:41.099 Uttam Kumaran: like, I would love to just hear, like, what the scope is, and then maybe if, like, we can pare it down. Because, for example, do… like, making sure something from Mother Duck gets into Salesforce.

337 00:40:41.300 ⇒ 00:40:49.169 Uttam Kumaran: like, hourly or daily is not hard. What is hard is, like, if there are, like, 50 metrics, and it needs to go to, like, 4 systems, then I’m like…

338 00:40:49.870 ⇒ 00:40:53.409 Uttam Kumaran: Okay, that is now, like, a… that has to become a line on the Gantt versus, like.

339 00:40:53.410 ⇒ 00:40:54.340 Caitlyn Vaughn: Right.

340 00:40:54.340 ⇒ 00:40:58.669 Uttam Kumaran: an afternoon project to just get, like, one field synced, right? So…

341 00:40:58.670 ⇒ 00:41:14.589 Caitlyn Vaughn: It’s, I think a little bit murky still, which is why I would say we should, like, punt this till after the project is done anyway, because the real stakeholders on this would be, like, Ryan and Lev, and when Ryan, like.

342 00:41:15.020 ⇒ 00:41:26.019 Caitlyn Vaughn: he, like, put a calendar, a meeting on my calendar, and was like, what data can we get into Salesforce? And I was like, no, that’s not how this works. Like, you tell me what you need, and I will tell you.

343 00:41:26.020 ⇒ 00:41:26.470 Uttam Kumaran: I like it.

344 00:41:26.470 ⇒ 00:41:41.390 Caitlyn Vaughn: anything’s possible, right? Yeah. So, yeah, I think… let me get some clarity on… on him, on, like, a specific ask, because otherwise it’s not worth anybody’s time to hook it up. But it’s just gonna be Salesforce, we’re kicking Catalyst, and then…

345 00:41:41.390 ⇒ 00:41:41.740 Uttam Kumaran: Cool.

346 00:41:41.740 ⇒ 00:41:48.090 Caitlyn Vaughn: We did talk about last week that we’re not going to use Polytomic for reverse ETL, because it’s too expensive.

347 00:41:48.090 ⇒ 00:42:04.559 Uttam Kumaran: So that’s where, also, I was like, look, if it ends up being a big thing, then another option is, like, I’ll just put… I could put pressure on them to bring the cost down. But again, I want to just… like, I really just want to define the scope. So, again, like, maybe we’ll kick it. If you’re talking about it, and you want me to hop in a thing on… and just, like…

348 00:42:04.630 ⇒ 00:42:08.289 Uttam Kumaran: drive towards, like, okay, it would be great if I could just see…

349 00:42:08.420 ⇒ 00:42:19.260 Uttam Kumaran: like, really, we also have Deanna’s ask, which were, like, I just want to see the number of seats they have, and, like, when the last time they added an admin, and, like, if it’s, like, one or two of those.

350 00:42:19.400 ⇒ 00:42:23.569 Uttam Kumaran: then we can support it. Yeah, so maybe… yeah.

351 00:42:23.570 ⇒ 00:42:38.219 Caitlyn Vaughn: She’ll be back in, like, 2 weeks, and I think we’re gonna get rid of Catalyst, which maybe she’ll fight it when she comes back. But thinking about it more, I don’t think that there is any reason that she needs reverse ETL into Catalyst, if it’s purely for data reasons, like to…

352 00:42:38.220 ⇒ 00:42:38.940 Uttam Kumaran: That’s the same, yeah.

353 00:42:39.490 ⇒ 00:42:47.470 Uttam Kumaran: I think it’s just the same thing. She should… I think she could use Salesforce, but I think her ask is gonna be the same as, like, I need to see the number of users.

354 00:42:47.470 ⇒ 00:42:48.140 Caitlyn Vaughn: You know, she doesn’t need.

355 00:42:48.140 ⇒ 00:42:50.669 Uttam Kumaran: Yeah, yeah, also, yeah, she could use Omni, yeah.

356 00:42:50.670 ⇒ 00:43:08.269 Caitlyn Vaughn: But for Salesforce, I understand the idea behind it. Like, they’re wanting to use updated Salesforce fields to trigger certain flows for sales, which… it makes sense to me for that. But yeah, let’s kick this, let’s use the hours that we’re gonna spend on this on, like, getting the core project done.

357 00:43:08.610 ⇒ 00:43:09.470 Uttam Kumaran: Okay, okay.

358 00:43:14.530 ⇒ 00:43:15.190 Uttam Kumaran: Cool.

359 00:43:15.500 ⇒ 00:43:23.050 Demilade Agboola: Okay, so in terms of ID matching, I know this has come up in terms of conversations as to how we’re going to tie the different systems together.

360 00:43:23.220 ⇒ 00:43:25.629 Demilade Agboola: After, like, further exploration.

361 00:43:25.840 ⇒ 00:43:35.070 Demilade Agboola: I’ve seen that, like, Salesforce ID appears to be a common denominator across multiple different systems, so you can see the account ID in Salesforce.

362 00:43:35.250 ⇒ 00:43:40.880 Demilade Agboola: also presenting, like, Hyperline data and QuickBooks data, so that allows us to be able to tie things together.

363 00:43:41.040 ⇒ 00:43:47.789 Demilade Agboola: That being said, if there are any systems that do not integrate that as well, potentially what we could just do as a fallback option is

364 00:43:48.380 ⇒ 00:43:59.530 Demilade Agboola: match the account names based off the domains together. That’d be more messy, but, like, that would just be a fallback option. But for now, we’re fine with using the account IDs to match things together.

365 00:44:00.440 ⇒ 00:44:06.799 Uttam Kumaran: Does this make sense on ID stitching? And, like, Caitlin, did the thing I sent you earlier, like, kind of help a little bit?

366 00:44:07.340 ⇒ 00:44:17.800 Caitlyn Vaughn: Yeah, I’ve been talking to our, one of our, like, data engineers that built out Core Data Model about this, and he says…

367 00:44:22.340 ⇒ 00:44:28.229 Caitlyn Vaughn: Sorry, there’s, like, a bunch of stuff. Give me, like, 10 seconds to fill out the important part here.

368 00:44:30.390 ⇒ 00:44:33.390 Caitlyn Vaughn: He said, I mean, using the unique ID to help

369 00:44:33.660 ⇒ 00:44:36.610 Caitlyn Vaughn: The, like, separate entities in the system makes sense.

370 00:44:37.670 ⇒ 00:44:47.529 Caitlyn Vaughn: it’s probably better, but, like, once you have questions like, how do I decide when this entity from Y.com is different than X.com?

371 00:44:47.660 ⇒ 00:44:53.739 Caitlyn Vaughn: then… Like, unique ideas probably better than, like, a domain?

372 00:44:54.240 ⇒ 00:45:00.170 Caitlyn Vaughn: I don’t think we need a ha- okay, he’s also saying, like.

373 00:45:01.310 ⇒ 00:45:07.980 Caitlyn Vaughn: we need to have some kind of answer for those questions outside of the dbt system that inform, like, what we want to use.

374 00:45:07.980 ⇒ 00:45:13.020 Uttam Kumaran: We have to decide, not us, but we’re probably… we’re probably the first people to, like, be like.

375 00:45:13.490 ⇒ 00:45:16.590 Uttam Kumaran: joining things and being like, yo, this is not right, so…

376 00:45:16.680 ⇒ 00:45:29.460 Caitlyn Vaughn: Exactly. So, I don’t know, I think it… maybe it makes sense to just stick with domain for now, but, like, what is the lift on us transitioning later to, like, a unique ID versus starting with it now?

377 00:45:29.460 ⇒ 00:45:31.070 Uttam Kumaran: Just changing the joins.

378 00:45:31.630 ⇒ 00:45:32.200 Caitlyn Vaughn: Yeah.

379 00:45:32.200 ⇒ 00:45:39.229 Uttam Kumaran: So as long… if, like, if every system gives us that same ID, then we just move to that. We will find, though, the problem that I told you, which is, like.

380 00:45:39.440 ⇒ 00:45:50.870 Uttam Kumaran: like, Samsung, for example, there’s, like, 50 domains that all are, like, Samsung, so you’re gonna get into this… it’s probably what’s happening in Salesforce, like, you have, like, orgs, you have, like, parent orgs and childs, right? So…

381 00:45:50.870 ⇒ 00:45:51.200 Caitlyn Vaughn: Yeah.

382 00:45:51.200 ⇒ 00:45:52.620 Uttam Kumaran: That is something that I think…

383 00:45:53.540 ⇒ 00:45:57.080 Uttam Kumaran: The sales team needs to define.

384 00:45:57.380 ⇒ 00:46:13.380 Uttam Kumaran: like, what is the definition of an org versus, like, I don’t know, I don’t know if it’s org and then, like, team or whatever, and then the individual, and, like, when they sell an enterprise contract to where they’re like, yeah, like, 5 subsidiaries are using it, what happens? If that’s not the case now, and it’s…

385 00:46:13.380 ⇒ 00:46:20.030 Uttam Kumaran: and I know we’re just, like, it’s just flat, like, there’s companies and then there’s people, then domain is totally fine, because

386 00:46:20.180 ⇒ 00:46:27.290 Uttam Kumaran: Whatever they are putting into their domain in Salesforce is what we’re gonna use, and as long as the user emails all match.

387 00:46:27.980 ⇒ 00:46:31.500 Uttam Kumaran: Then we’re kinda like, We’re kind of chill, until…

388 00:46:31.720 ⇒ 00:46:34.410 Uttam Kumaran: Until the kind of complexity happens, right?

389 00:46:34.410 ⇒ 00:46:34.970 Caitlyn Vaughn: Hmm…

390 00:46:34.970 ⇒ 00:46:45.519 Nandika Jhunjhunwala: So, I, like, foresee some complexity, like… I see, like, there’s a lot of complexity, like, in Salesforce already, like, even though we don’t have, like, parent-child company objects.

391 00:46:45.530 ⇒ 00:47:08.380 Nandika Jhunjhunwala: Sometimes, like, company domains are different from, like, email domains, and, like, things get acquired, like, we might, like, in the future, be selling to, like, multiple subsidiaries of a company, and then they may all have one parent domain, and they may want to exist as separate entities in our CRM. So, like, those sort of edge cases, I think we would expect to encounter more and more as our growth team, like, expands.

392 00:47:08.380 ⇒ 00:47:09.050 Uttam Kumaran: Yeah.

393 00:47:09.450 ⇒ 00:47:19.490 Nandika Jhunjhunwala: start going more upmarket. So, like, I think, like, personally, like, for the sake of future-proofing and not having to come back and then generate unique IDs on, like, objects.

394 00:47:19.730 ⇒ 00:47:25.169 Nandika Jhunjhunwala: In the future, I, like, also am of the opinion that we need, like, unique IDs,

395 00:47:25.170 ⇒ 00:47:30.680 Uttam Kumaran: Yeah, so, like, we will form kind of an ID on our side, and, like, identity stitching is, like, this whole…

396 00:47:31.220 ⇒ 00:47:32.090 Uttam Kumaran: Thing. Yeah.

397 00:47:32.310 ⇒ 00:47:39.149 Uttam Kumaran: is something we do for a bunch of people when, like, there’s not a primary key. So we’re always trying to identify what that is.

398 00:47:39.510 ⇒ 00:47:49.690 Uttam Kumaran: Like, eventually, in our system, someone… there’ll be two companies that are listed with the same domain, because you guys sold a contract to two teams within one company, and then…

399 00:47:49.800 ⇒ 00:47:52.950 Uttam Kumaran: we’ll get a WTBT alert, or, like, something will break.

400 00:47:53.090 ⇒ 00:48:02.139 Uttam Kumaran: And then… and then… but again, I think, Kalen, whoever in the product team is thinking about it, that’s perfect. Like, they should be making the decision for us.

401 00:48:02.240 ⇒ 00:48:06.409 Caitlyn Vaughn: We will… we have to do whatever it takes to get accurate data.

402 00:48:06.530 ⇒ 00:48:09.960 Uttam Kumaran: reported on, but… Yeah.

403 00:48:11.120 ⇒ 00:48:11.960 Caitlyn Vaughn: Okay, cool.

404 00:48:13.590 ⇒ 00:48:15.139 Caitlyn Vaughn: Maybe a TBD, then.

405 00:48:15.140 ⇒ 00:48:16.190 Uttam Kumaran: Yeah, TBD.

406 00:48:18.670 ⇒ 00:48:19.430 Demilade Agboola: Okay…

407 00:48:22.090 ⇒ 00:48:27.350 Demilade Agboola: So yes, this is just about metric definition and schema, so we currently have a work in progress doc.

408 00:48:27.560 ⇒ 00:48:34.719 Demilade Agboola: That has every single… Dashboard we’re looking to build out right now.

409 00:48:35.250 ⇒ 00:48:38.049 Demilade Agboola: So we can share that with the team.

410 00:48:38.580 ⇒ 00:48:43.870 Demilade Agboola: So yeah, different people can hop in here and then eventually just let us know, like, hey, this works, this doesn’t work.

411 00:48:44.170 ⇒ 00:48:46.899 Demilade Agboola: Or, like, can we add more to this?

412 00:48:47.410 ⇒ 00:48:50.539 Uttam Kumaran: So I think we’ll probably try to get that in, like, we… in our spreadsheet, too.

413 00:48:50.650 ⇒ 00:48:53.549 Uttam Kumaran: And we’re working on that right now, so… yeah.

414 00:48:54.690 ⇒ 00:48:57.129 Nandika Jhunjhunwala: Yeah, really appreciate that, thank you so much.

415 00:48:57.500 ⇒ 00:49:05.850 Demilade Agboola: Okay, and then also in terms of, like, the list of all tables that currently exist, and the columns, and all of that, with, like, the different schemas that…

416 00:49:06.010 ⇒ 00:49:08.879 Demilade Agboola: are currently being ingested. I’ll send that over.

417 00:49:09.120 ⇒ 00:49:15.139 Demilade Agboola: As well, and then… Basically, the definition of…

418 00:49:16.150 ⇒ 00:49:21.470 Demilade Agboola: The metrics will be in… are in the document, like, that document that we’re sharing,

419 00:49:22.000 ⇒ 00:49:28.370 Demilade Agboola: And so, that’s that for that. In terms of data governance, so, you know, setting

420 00:49:28.670 ⇒ 00:49:33.090 Demilade Agboola: data needs to be only seen by certain people. I know Laura made a request.

421 00:49:33.350 ⇒ 00:49:39.779 Demilade Agboola: for certain, dashboards that I’m gonna make to be only be seen by her and, say, the CEO.

422 00:49:40.120 ⇒ 00:49:46.870 Demilade Agboola: So Val will be… done, or go over that with her, and just to ensure that, like.

423 00:49:47.340 ⇒ 00:49:59.729 Demilade Agboola: the access is restricted to that. Also, just going forward, if there are any, like, other sort of data governance principles you would like us to ensure that, you know, certain dashboards don’t, you know, get seen by everyone.

424 00:49:59.910 ⇒ 00:50:02.360 Demilade Agboola: Or certain dashboards are only seen by certain teams.

425 00:50:02.570 ⇒ 00:50:05.730 Demilade Agboola: We’ll definitely have to work together on that, and just ensure that

426 00:50:06.020 ⇒ 00:50:08.269 Demilade Agboola: Across the board, everyone’s on the same page with that.

427 00:50:10.170 ⇒ 00:50:11.379 Caitlyn Vaughn: Cool, it looks good.

428 00:50:11.610 ⇒ 00:50:12.490 Demilade Agboola: Okay.

429 00:50:13.240 ⇒ 00:50:17.609 Demilade Agboola: Yeah, in terms of, like, acceleration and things we want to ramp up on, going forward.

430 00:50:17.800 ⇒ 00:50:29.960 Demilade Agboola: We want to start, like, building out the dashboards, and in case… in the situations where we don’t have all the data, in terms of, like, an ex… like, in terms of, like, a pipeline, we’ll just use one-time exports as best as possible.

431 00:50:30.210 ⇒ 00:50:36.830 Demilade Agboola: So, things like the customer product activity dashboard, we’ll just try and use manual export.

432 00:50:37.130 ⇒ 00:50:44.640 Demilade Agboola: But ideally, we will want to, you know, just push on Victor as well, so we can have that in parallel as well.

433 00:50:44.860 ⇒ 00:50:50.840 Demilade Agboola: The aim is to have that final week next week, but potentially, you know, we will rip that out early next week.

434 00:50:53.020 ⇒ 00:51:02.169 Demilade Agboola: path forward for, like, the university, I’ve kind of talked about that, and potentially Utem would reach out to the team once we get clearer ideas of the scope of that.

435 00:51:02.650 ⇒ 00:51:06.780 Demilade Agboola: And then, in terms of, like, our attribution dashboard for, like, LinkedIn.

436 00:51:06.900 ⇒ 00:51:14.370 Demilade Agboola: We don’t have a connector yet. We are currently talking to the Polyatomic team, about playing

437 00:51:14.660 ⇒ 00:51:16.799 Demilade Agboola: Fructose AI and Email Bison.

438 00:51:17.040 ⇒ 00:51:20.130 Demilade Agboola: So you can have those 3 connectors built out.

439 00:51:20.330 ⇒ 00:51:33.579 Demilade Agboola: And then, but for now, we could… we’ll try and get, like, a one-time export from these different sources, so we have an idea of what sort of data we’re working with, build out the dashboard based off of that, and then once the connectors are built and we have

440 00:51:33.740 ⇒ 00:51:37.340 Demilade Agboola: Connect everything together, we can start to…

441 00:51:37.960 ⇒ 00:51:43.309 Demilade Agboola: We can start to change the source so that it uses the connected data instead.

442 00:51:44.650 ⇒ 00:51:55.139 Uttam Kumaran: Yeah, so I think, like, the kind of theme of this month is just to get as many V1 dashboards as possible, and get as many people, like, in Omni. I think one thing

443 00:51:55.330 ⇒ 00:51:59.870 Uttam Kumaran: That would be great to start sharing weekly is also just, like, the Omni usage.

444 00:51:59.980 ⇒ 00:52:16.420 Uttam Kumaran: It’s one thing that I looked at this week, and so really, it’s like, I just want to start seeing, you know, like, a bunch of names pop up on there, and people are using the AI pieces in particular. We had two other clients that are using… that we set up the AI, like, semantic layer in Omni, and it’s answering, like.

445 00:52:16.520 ⇒ 00:52:28.889 Uttam Kumaran: It’s just, like, answering the questions perfectly. So I think, like, I want to go one step beyond just, like, the dashboards ready, but, like, as many people as we can get using their… the thing is called Blobby.

446 00:52:29.280 ⇒ 00:52:42.209 Uttam Kumaran: As much as we can get them using that, it’s able to answer, like, level 1 and level 2 questions, like, how much money are we making, and things like, who, like, churned this week.

447 00:52:42.210 ⇒ 00:53:00.809 Uttam Kumaran: And then it’s also, like, if you structure the question really good, you could start building, like, cohorts and things like that. It’s oftentimes, though, as we know about AI, just as good as, like, what you ask it. So if you ask it kind of, like, if you don’t say the right words, it’s not gonna be able to do it. But for things that most of the people are asking, like, how much money do we make? Tell me about this customer.

448 00:53:00.940 ⇒ 00:53:03.330 Uttam Kumaran: Like, it nails it, so…

449 00:53:03.560 ⇒ 00:53:09.430 Caitlyn Vaughn: Yeah. Okay, cool, yay! Our team will definitely use that over SQL queries.

450 00:53:09.700 ⇒ 00:53:27.599 Nandika Jhunjhunwala: Yeah, I played around with it a little bit. Found some, like… I think you have to be really specific about what you’re asking it, otherwise it will sort of point towards the wrong data, is, like, what I found. So you’re right, like, it needs to… like, the query doesn’t need to be SQL, but it needs to be sort of exact.

451 00:53:28.340 ⇒ 00:53:29.880 Uttam Kumaran: Yeah, yeah, I agree.

452 00:53:32.720 ⇒ 00:53:42.440 Uttam Kumaran: Cool. So, I mean, I think, like, that’s… that’s what I’m hoping for next week, is, like, finance is on a good track, we move towards stuff for Lev, which is all Salesforce.

453 00:53:42.710 ⇒ 00:53:59.949 Uttam Kumaran: and customer activity. The nice thing is, like, I think Nandica, as soon as those topics are modeled, go build whatever dashboard on those topics. So I think that’s where we wanted to sort of get to the first foundational layer of, like, core topics are built, because then adding in metrics and tweaking is… we’re going to be doing

454 00:54:00.120 ⇒ 00:54:19.729 Uttam Kumaran: For the history. That’s just, like, what data teams are gonna do forever, so that is actually, like, we’re expecting there to be tweaks. It’s more of, like, can we get the topics ready so that you can start creating dashboards? And we have other places that we haven’t gone yet, right? Like, Google attribution, LinkedIn attribution, where those are net new sources, net new topics.

455 00:54:19.910 ⇒ 00:54:25.049 Uttam Kumaran: But, like, it doesn’t mean we can’t continue to add metrics or create more dashboards on

456 00:54:25.510 ⇒ 00:54:29.039 Uttam Kumaran: the Salesforce data, on the finance data, on the usage data, so…

457 00:54:29.660 ⇒ 00:54:48.110 Uttam Kumaran: If we want to indicate some of that, like, and one thing we could do, maybe as we get those core topics out, is think about, like, what other dashboards or what other views need to get answered. But the best bang for the buck is to use our team to set the foundation, and then making sure that, like, Blobby can answer any of those questions, and, like.

458 00:54:48.440 ⇒ 00:54:50.410 Uttam Kumaran: Set that layer so that, like.

459 00:54:51.060 ⇒ 00:54:54.850 Uttam Kumaran: Anyone in default can start to build, as long as they’re able to, you know?

460 00:54:55.280 ⇒ 00:55:10.940 Caitlyn Vaughn: Yeah, I also feel like, Nandika, if you’re getting, like, wrong answers from Lobby, amazing name, then potentially there are gaps on our side that we need to fix in order to get the right answers. Because I think we’re definitely at a place just, like, within

461 00:55:10.940 ⇒ 00:55:15.289 Caitlyn Vaughn: the scope of AI that, like, it should be able to answer the right questions.

462 00:55:16.020 ⇒ 00:55:24.450 Caitlyn Vaughn: No, I was definitely using Blobby for more, like, SQL queries to create, like, views. Oh, okay. So that’s not, like, its main use case. Okay.

463 00:55:24.450 ⇒ 00:55:39.310 Nandika Jhunjhunwala: I think I was, like, before, like, I met with Demi, I was, like, looking at, like, you know, what’s the dev, like, what’s the raw export sort of thing, and, like, trying to decide where I should pull data from to create those views for those dashboards, so… Okay, cool. That’s not its use case.

464 00:55:39.310 ⇒ 00:55:44.570 Nandika Jhunjhunwala: Or maybe it is, I’m not sure, but I think it requires a lot more experimentation and playing around on my end.

465 00:55:44.830 ⇒ 00:55:45.520 Uttam Kumaran: Yeah.

466 00:55:45.690 ⇒ 00:56:01.000 Uttam Kumaran: And that’s why, like, one thing we’re gonna do is, like, we’re taking as much context as default as we know and putting it into table-level definitions, metric-level definitions, topics, and it’s using all that context to then be like, okay, I need to do this join, or I need to query this metric.

467 00:56:01.070 ⇒ 00:56:19.249 Uttam Kumaran: two metrics are named churn, but, like, in this situation, you need to use that. But again, it’s only as good as how much context we give it. So I think also, like, you’re probably gonna start asking the questions that people at default are gonna ask once they’re all in. So if it’s not working for you, like, we should just fix it. So, like, even if you’re seeing that, like, it messed something up.

468 00:56:19.400 ⇒ 00:56:31.220 Uttam Kumaran: don’t blame yourself first, just yet. It’s for sure a system thing, and, like, just say, like, hey, I asked this to Blobby, didn’t work, and we’re not gonna take offense, because there’s probably not context in there about, like.

469 00:56:31.300 ⇒ 00:56:34.339 Caitlyn Vaughn: How it should be used, and so we’ll just continue to layer that in.

470 00:56:34.340 ⇒ 00:56:46.120 Nandika Jhunjhunwala: Yeah, no, maybe we can grab some time. I… I had some issues with the Omni UI as well, like, I created some, like, query views, and I, like, couldn’t access it, and I think maybe, like.

471 00:56:46.120 ⇒ 00:57:08.639 Nandika Jhunjhunwala: that’s, like, enablement on my end, that I definitely, definitely need to spend more time in Omni to, like, get more up to speed with the platform, and that’s, like, been on my docket for a second. I’ve just been, like, really, like, hands-on with growth, so I haven’t had so much time in analytics, dashboarding world, but that’s, like, definitely something that I’m planning on doing more, because I feel like I need to be enabled to enable the rest of the team here. Yeah.

472 00:57:08.640 ⇒ 00:57:11.150 Nandika Jhunjhunwala: is… I’m gonna be the point of contact, so that’s definitely.

473 00:57:11.150 ⇒ 00:57:18.849 Uttam Kumaran: And luckily, Greg, does a lot of Omni work, so I’m actually gonna… he’s gonna probably help out a little bit more with, like, Omni setup.

474 00:57:18.990 ⇒ 00:57:23.799 Uttam Kumaran: he just… they just finished for another client, like, a big end-to-end Omni implementation, so…

475 00:57:23.980 ⇒ 00:57:26.649 Caitlyn Vaughn: Like, I think he’s gonna be really helpful, and…

476 00:57:26.800 ⇒ 00:57:45.169 Uttam Kumaran: again, like, we’ve been saying the word blobby, like, every day, like, for, like, months now, so I feel like we’re trying our best to make sure, because that is actually… a lot of people you’re gonna find are not comfortable with dashboards, but they’re comfortable… everyone at DeFi knows comfortable using a chat interface, so it’s gonna be much simpler for them

477 00:57:45.230 ⇒ 00:57:53.690 Uttam Kumaran: Versus finding the right dashboard, or in the dashboard being like, what does this mean? To just start via chat? And that’s, like, one of the big reasons we, like.

478 00:57:54.120 ⇒ 00:58:02.240 Uttam Kumaran: we, like, recommend Omni’s because it does everything and that, and I think that’s gonna be more and more of the primary mode of accessing data, you know, so…

479 00:58:04.380 ⇒ 00:58:05.350 Caitlyn Vaughn: Agreed.

480 00:58:05.660 ⇒ 00:58:06.749 Caitlyn Vaughn: We’re building it.

481 00:58:06.750 ⇒ 00:58:14.410 Uttam Kumaran: Yes, yes, and then Defo, I heard DeFo will have something like that, too, so our team will be using that, because we… every… you know, everybody in our company…

482 00:58:14.580 ⇒ 00:58:18.160 Uttam Kumaran: Uses default now for all their calendar bookings, and

483 00:58:18.660 ⇒ 00:58:22.849 Uttam Kumaran: I also want to chat with our default metrics someday, so…

484 00:58:22.850 ⇒ 00:58:31.959 Caitlyn Vaughn: Yay! That’s so exciting. I know, and we’re doing, like, the whole context layer. Honestly, I need to write that PRD, like, probably this week, so maybe I’ll send it to you before you can get.

485 00:58:31.960 ⇒ 00:58:33.200 Uttam Kumaran: I would love to read it, yeah.

486 00:58:33.200 ⇒ 00:58:36.330 Caitlyn Vaughn: Your data engine feedback, too.

487 00:58:36.330 ⇒ 00:58:38.829 Uttam Kumaran: Yeah, yeah, yeah, that’d be great.

488 00:58:38.830 ⇒ 00:58:40.840 Caitlyn Vaughn: Dead. Okay, cool.

489 00:58:41.990 ⇒ 00:58:48.680 Demilade Agboola: Okay, also, yeah, in terms of, like, the query view, like, Madika, like.

490 00:58:49.110 ⇒ 00:58:52.440 Demilade Agboola: do you… have you been able to try out the GitHub, like, pull,

491 00:58:52.440 ⇒ 00:58:53.360 Nandika Jhunjhunwala: Yeah, yeah, yeah.

492 00:58:53.660 ⇒ 00:58:54.410 Demilade Agboola: Where?

493 00:58:54.410 ⇒ 00:59:03.100 Nandika Jhunjhunwala: I wasn’t sure, like… like, I, like, Demi told me that I need to push GitHub PRs to have those views available to me.

494 00:59:03.460 ⇒ 00:59:08.640 Nandika Jhunjhunwala: Which I’m happy to do, it’s just… then maybe I should be experimenting elsewhere, it’s like…

495 00:59:08.810 ⇒ 00:59:13.799 Nandika Jhunjhunwala: what I think, because I don’t want to be pushing PRs for, like, test queries and stuff.

496 00:59:14.280 ⇒ 00:59:15.739 Demilade Agboola: Yeah, so… I mean.

497 00:59:15.740 ⇒ 00:59:18.250 Nandika Jhunjhunwala: Like, what’s the best way to navigate the platform?

498 00:59:19.110 ⇒ 00:59:21.460 Demilade Agboola: Yeah, I mean, we could definitely have, like, a session, like, you know.

499 00:59:21.460 ⇒ 00:59:22.210 Nandika Jhunjhunwala: Fair enough.

500 00:59:22.210 ⇒ 00:59:30.010 Demilade Agboola: Where we can just walk through that and just try and figure out, like, how you want to do test queries, and how, like, a schema for you to be able to put that in there.

501 00:59:33.470 ⇒ 00:59:34.690 Demilade Agboola: Okay…

502 00:59:34.690 ⇒ 00:59:45.909 Caitlyn Vaughn: Oh wait, go back one. The, giving Stan the LinkedIn attribution dashboard, is the email that you need an invite to admin at brainforge.ai?

503 00:59:46.830 ⇒ 00:59:49.220 Demilade Agboola: No, it’s brainforge at default.com?

504 00:59:49.650 ⇒ 00:59:51.899 Caitlyn Vaughn: Oh, you have a default email now?

505 00:59:51.900 ⇒ 00:59:53.690 Uttam Kumaran: We have a default email. We’ve always had one.

506 00:59:53.690 ⇒ 00:59:54.750 Nandika Jhunjhunwala: Yeah, yeah.

507 00:59:55.390 ⇒ 00:59:56.140 Uttam Kumaran: That way, I mean.

508 00:59:56.550 ⇒ 01:00:01.309 Uttam Kumaran: I always give clients, like, two options, so that… that way we’re all under your domain, so that it’s not like…

509 01:00:02.150 ⇒ 01:00:03.590 Nandika Jhunjhunwala: You don’t have an external…

510 01:00:03.790 ⇒ 01:00:06.059 Caitlyn Vaughn: But yeah, that’s what we’ve been using for everything. Okay.

511 01:00:06.430 ⇒ 01:00:08.269 Caitlyn Vaughn: Brainforge, I default.

512 01:00:08.590 ⇒ 01:00:09.149 Uttam Kumaran: dot com.

513 01:00:09.150 ⇒ 01:00:24.059 Caitlyn Vaughn: Why? Because you don’t… Like, panicking. Okay, alright, I’m gonna have… I’ll have, Stan invite you to Factors, and I think I can invite you to Clay? Yeah, I can.

514 01:00:24.060 ⇒ 01:00:24.510 Nandika Jhunjhunwala: Yeah.

515 01:00:25.250 ⇒ 01:00:25.800 Nandika Jhunjhunwala: I think we’re.

516 01:00:25.800 ⇒ 01:00:28.249 Demilade Agboola: But we’re in factors already.

517 01:00:28.250 ⇒ 01:00:31.019 Nandika Jhunjhunwala: Yeah, Matt told me that he gave you access.

518 01:00:31.490 ⇒ 01:00:32.430 Nandika Jhunjhunwala: And he said…

519 01:00:32.430 ⇒ 01:00:33.229 Demilade Agboola: So this is a…

520 01:00:33.230 ⇒ 01:00:36.050 Nandika Jhunjhunwala: access on that email across things.

521 01:00:36.050 ⇒ 01:00:37.940 Caitlyn Vaughn: Okay, so you do have access.

522 01:00:38.020 ⇒ 01:00:38.720 Demilade Agboola: Yeah.

523 01:00:38.970 ⇒ 01:00:40.010 Caitlyn Vaughn: And to Clay.

524 01:00:40.810 ⇒ 01:00:41.190 Demilade Agboola: Yes.

525 01:00:41.190 ⇒ 01:00:44.820 Nandika Jhunjhunwala: I can give you access to Clay, but I thought we decided on, like.

526 01:00:44.820 ⇒ 01:00:46.019 Mustafa Raja: I already have access to clean.

527 01:00:46.020 ⇒ 01:00:46.530 Nandika Jhunjhunwala: it out.

528 01:00:47.660 ⇒ 01:00:50.230 Caitlyn Vaughn: Do we need… do we really need clay as a source?

529 01:00:50.570 ⇒ 01:00:53.409 Nandika Jhunjhunwala: I’m not sure, there’s, like, so much noise in play. There’s, like, a lot of…

530 01:00:53.410 ⇒ 01:00:54.580 Caitlyn Vaughn: That maybe we don’t need to.

531 01:00:54.580 ⇒ 01:00:56.390 Uttam Kumaran: So I was like, what are you… what are you guys…

532 01:00:56.690 ⇒ 01:00:59.530 Uttam Kumaran: What are you guys bringing from Clay? Like, is it just lists?

533 01:00:59.970 ⇒ 01:01:01.050 Caitlyn Vaughn: Yeah, just lists.

534 01:01:01.050 ⇒ 01:01:03.950 Uttam Kumaran: Why don’t they just send it to Salesforce, and we’ll take it from there?

535 01:01:03.950 ⇒ 01:01:07.659 Nandika Jhunjhunwala: No, no, no, you’re right, like, all of that does live in Salesforce already.

536 01:01:08.630 ⇒ 01:01:09.660 Caitlyn Vaughn: What’s up?

537 01:01:09.660 ⇒ 01:01:15.939 Nandika Jhunjhunwala: Like, things we’re not syncing to Salesforce from Clay is just things we actually don’t need to sync.

538 01:01:15.940 ⇒ 01:01:16.580 Caitlyn Vaughn: Hello?

539 01:01:16.580 ⇒ 01:01:22.909 Nandika Jhunjhunwala: the data warehouse. So I would intentionally err on the side of not syncing play it on a dock, because.

540 01:01:23.730 ⇒ 01:01:26.299 Nandika Jhunjhunwala: A lot of trashy data, yeah.

541 01:01:28.470 ⇒ 01:01:29.870 Caitlyn Vaughn: Nice, sorted.

542 01:01:30.110 ⇒ 01:01:30.870 Uttam Kumaran: Okay, cool.

543 01:01:33.880 ⇒ 01:01:39.340 Demilade Agboola: Okay, so in terms of, like, risks and mitigation, I know I just saw your message, Caitlin.

544 01:01:40.990 ⇒ 01:01:49.719 Demilade Agboola: But yeah, so the major thing is still, like, just Postgres, because a lot of data does live in Postgres, and we will need that to be able to…

545 01:01:52.200 ⇒ 01:01:55.430 Demilade Agboola: Model a lot of, like, customer behavior.

546 01:01:55.860 ⇒ 01:01:56.620 Demilade Agboola: he’s out.

547 01:01:56.820 ⇒ 01:01:59.330 Demilade Agboola: So that is a huge blocker in that regard.

548 01:02:00.100 ⇒ 01:02:01.500 Demilade Agboola: So, yeah.

549 01:02:02.160 ⇒ 01:02:03.619 Caitlyn Vaughn: Cool, working on it.

550 01:02:05.120 ⇒ 01:02:09.180 Demilade Agboola: And then… Yeah, I think that’s it from us for this week.

551 01:02:11.240 ⇒ 01:02:12.480 Caitlyn Vaughn: Awesome.

552 01:02:12.480 ⇒ 01:02:14.850 Uttam Kumaran: So I think we have some follow-ups on our end, I think.

553 01:02:15.040 ⇒ 01:02:17.810 Uttam Kumaran: B, if you want to drive some of the Gantt.

554 01:02:17.820 ⇒ 01:02:35.260 Uttam Kumaran: changes, and we can try to, like, lock all that down by tomorrow, and then, Nandika, we’ll send that doc over. You could just comment if you want us to change anything about the formatting, we can do a re-pull, but by, like, early next week, we’ll end up putting that all into, like, a spreadsheet format, because you could just see the list of everything.

555 01:02:36.570 ⇒ 01:02:45.090 Uttam Kumaran: And then, yeah, I mean, I think we’re driving towards the two dashboards, like, basically as soon as possible, so… we’ll just kind of keep updated daily on, like, how those are coming out.

556 01:02:45.470 ⇒ 01:02:47.359 Nandika Jhunjhunwala: Thank you so much, that sounds great.

557 01:02:47.360 ⇒ 01:02:48.300 Uttam Kumaran: Yeah, perfect.

558 01:02:49.330 ⇒ 01:02:49.870 Caitlyn Vaughn: Okay.

559 01:02:50.090 ⇒ 01:02:51.559 Demilade Agboola: Thanks, guys!

560 01:02:51.560 ⇒ 01:02:52.020 Uttam Kumaran: Thank you.

561 01:02:52.390 ⇒ 01:02:52.880 Caitlyn Vaughn: Hey guys.

562 01:02:53.440 ⇒ 01:02:54.330 Uttam Kumaran: Bye.

563 01:02:54.330 ⇒ 01:02:54.880 Demilade Agboola: Right.

Brainforge Knowledge

Explorer

2026-03-05_default_project_dashboard_sync_8de69dac

Graph View