Meeting Title: US x BF | Updates and Reviews Date: 2025-07-17 Meeting participants: Amber Lin, Zack Gibbs, Demilade Agboola, Uttam Kumaran


WEBVTT

1 00:00:09.500 00:00:10.579 Amber Lin: Hi Zach!

2 00:00:14.120 00:00:17.430 Zack Gibbs: Hello, sorry. Yeah.

3 00:00:17.720 00:00:22.509 Zack Gibbs: I my meeting my my calendar to

4 00:00:23.140 00:00:27.459 Zack Gibbs: this week in general has been all messed up. There’s been a bunch of interviews, and so I just

5 00:00:27.830 00:00:29.209 Zack Gibbs: got out of an interview.

6 00:00:29.710 00:00:35.360 Amber Lin: Oh, okay, no, worries I. I’ve heard from Emily that you had a lot of interviews recently.

7 00:00:43.150 00:00:45.979 Amber Lin: so demolan is here.

8 00:00:46.110 00:00:53.780 Amber Lin: I think we’re ready to go. I know, I added, Utam here. I think he will. He is going to join. Let me quickly ping him

9 00:01:00.070 00:01:01.220 Amber Lin: alright.

10 00:01:03.910 00:01:07.960 Amber Lin: So today we prepared a

11 00:01:08.070 00:01:34.479 Amber Lin: quick slide to show you our progress so far, and we want to align on the next steps, especially as revenue come up is also a big increment of work. And then I want to make sure that we’re doing it in the right direction, that you have expectations of what the outcome is going to be, and how long things are going to take, and also who are the resources we might need to pull or get time with.

12 00:01:34.800 00:01:41.490 Amber Lin: Utah said he’s joining. So I’ll give him another moment or 2, and then we can get started.

13 00:01:43.450 00:01:54.810 Amber Lin: Yeah, I guess before we do that. Is there anything that’s on your mind. That’s pretty pressing. Any pressing questions you want to ask. I just want to make sure that we account for that when we present our progress.

14 00:01:58.280 00:02:12.432 Zack Gibbs: I think, like high level where we? I know that the scoping for the revenue data mart is is happening. There’s some like technical design docs that I think last I had seen that you guys were going to present back to us for feedback

15 00:02:13.000 00:02:17.780 Zack Gibbs: But where are we in in just the inventory data mark.

16 00:02:18.840 00:02:22.694 Zack Gibbs: Complete like. Where? How close are we to completion? There?

17 00:02:24.520 00:02:29.498 Amber Lin: Yeah, that. Yeah, that’s that’s definitely part of what we want to show today. Hi,

18 00:02:29.830 00:02:30.500 Uttam Kumaran: A.

19 00:02:30.500 00:02:32.390 Amber Lin: And I mean

20 00:02:32.390 00:02:42.719 Amber Lin: we can get right into it, or we I can let give you like a quick, very quick overview of how much is left for inventory.

21 00:02:46.730 00:02:58.689 Zack Gibbs: Happy to help happy for you guys to, you know, go through whatever makes sense. I just that’s my high level question is, where where are we? There? And then, obviously, you know, revenue is revenues next which we’re gonna we’re gonna chat.

22 00:02:58.690 00:03:05.329 Amber Lin: Yeah, totally. So for inventory, I think we are at the final stages.

23 00:03:05.360 00:03:10.019 Amber Lin: Currently, it’s extended a bit longer than we expected, because since we

24 00:03:10.030 00:03:20.349 Amber Lin: started working with Felipe. He brought up a lot of great points and his pain points, and we want to make sure that before we wrap it up while we’re still working on inventory, that we help him out

25 00:03:20.350 00:03:43.210 Amber Lin: with that. So just to make sure we actually deliver the last mile, make sure that happens, and after that is documentation which we already have been doing really well throughout the progress. So wrapping that up, making sure that people say 5 years down the road can come back to the documentation and know what we did. So that would just be the last step wrapping up for inventory. After we finish.

26 00:03:43.210 00:03:45.100 Amber Lin: Philly Biz requests.

27 00:03:46.020 00:03:53.990 Zack Gibbs: And what are Felipe’s requests? Are they? These are more tactical requests. Of things that he’s missing doesn’t have or.

28 00:03:54.880 00:04:06.409 Demilade Agboola: Yeah, so they’re more tactical requests. And in some cases they’re queuing to the point of us, discovering certain like loopholes, so to speak.

29 00:04:07.116 00:04:11.880 Demilade Agboola: but for the most part the numbers are there. We have them available.

30 00:04:12.080 00:04:14.023 Demilade Agboola: But it’s things like

31 00:04:15.490 00:04:18.290 Demilade Agboola: So right now we are counting.

32 00:04:18.440 00:04:23.930 Demilade Agboola: Say, someone wanted the classic sub, and instead got the firecracker.

33 00:04:24.821 00:04:27.379 Demilade Agboola: We are counting that as.

34 00:04:27.670 00:04:30.970 Demilade Agboola: and uncommitted on the classic on the firecracker.

35 00:04:31.530 00:04:45.969 Demilade Agboola: but he wants it instead to be counted as an uncommitted on the fire sob. So he knows that on classic, so he knows that in a situation where things were needed, but they were not available. That counts, as you know, that sort of thing.

36 00:04:46.624 00:05:03.845 Demilade Agboola: So it’s things like that. We’re just trying to twist certain numbers, insurance numbers align. There were. There was also a case this past week where we had a slightly larger numbers, and we discovered that we’re also counting hard goods as well.

37 00:05:04.650 00:05:24.540 Demilade Agboola: for uncommitted values. And so just trying to remove those numbers, it’s kind of like a Qa. Process. The numbers they exist. The numbers are there. We have the shrinkage for every single lot. Id. We have the different inventory adjustments as well. The numbers are largely in a good spot. It’s just a final push.

38 00:05:28.090 00:05:28.710 Zack Gibbs: Okay?

39 00:05:29.860 00:05:39.060 Zack Gibbs: And then on the for the other analysts. Aside from Felipe, who’s been who’s been heavily involved? And

40 00:05:40.640 00:05:46.819 Zack Gibbs: in the the data mark, aside from Emily, who else other than Selipe, has been heavily involved to provide feedback.

41 00:05:47.100 00:05:54.230 Amber Lin: So Felipe has been heavily involved. We have yet to involve Perry. We discussed with Emily. I think

42 00:05:54.240 00:06:21.269 Amber Lin: she thinks that Felipe gives the most direct feedback. We are open to adding any analysts to our working sessions. We currently, I believe we have around 3 working sessions an hour each each week. So whoever you believe can also give very great feedback on the data marts, we will add them. And we want to have that standard practice as we start the revenue Mars, as well.

43 00:06:23.150 00:06:23.780 Zack Gibbs: Okay.

44 00:06:24.184 00:06:53.269 Zack Gibbs: yeah, I just wanted to get check. I mean, I’ve I’ve given Emily the feedback that you know, we need to make sure that the the and the analyst team is bought into the proposed structure, new structure and in general changes right? Because we want them to own it. Going forward. So I just wanna make sure that I’m telling. I’m telling Emily that, do we feel like that is happening? Or do we like that as a gap.

45 00:06:55.293 00:07:06.939 Demilade Agboola: I. Also, I feel like with part of also, just to add to the previous point about other analysts getting on board. It’s. I’ve also asked Emily to help roll out the numbers into Luca.

46 00:07:07.400 00:07:10.539 Demilade Agboola: because ultimately the best way to get

47 00:07:10.680 00:07:21.389 Demilade Agboola: the most out of the data and also curated data is when people are using it. And people who have used the data for multiple years get their spider senses tingling that like, you know, numbers look off.

48 00:07:21.520 00:07:39.199 Demilade Agboola: This doesn’t make sense, and you know they can also then validate against their previous methods. See that this is an easier method, and then, you know, lean into it. So that is one way to like, just get analysts on board, generally speaking, and I’m pretty sure once Perry gets her hands on the number. We’ll get a lot of feedback as well, and that would be very helpful.

49 00:07:39.722 00:07:49.869 Demilade Agboola: But yes, just generally in terms of like adopt adoption to the new system and the new ways. The new system is largely a Dbt concept.

50 00:07:49.990 00:07:53.700 Demilade Agboola: and we are rolling out the numbers in such a way that

51 00:07:54.690 00:08:04.799 Demilade Agboola: it can fix it. It doesn’t cause a huge issue into how the numbers are currently presented. We will have to create certain new dashboards or replace certain dashboards.

52 00:08:05.070 00:08:15.609 Demilade Agboola: but it wouldn’t be a huge showstopper. It shouldn’t be something that causes a huge disruption, because disruptions make people not adapt these changes illegally.

53 00:08:15.840 00:08:40.559 Demilade Agboola: So we’re we’re giving them the numbers. But in a in a slightly different format, getting into slightly different levels. So we’re either potentially going to enrich dashboards with more numbers than they maybe had before, or in some testing cases split out dashboards into smaller bits. So the idea now is we’re going to have a floral inventory space and a hard, good inventory space. So now you can look at the inventories for hard goods and floral goods like separately.

54 00:08:40.590 00:08:55.879 Demilade Agboola: and be able to make your decisions based off the different parts. You can also roll it up to Fc, you can also get things about the shrinkage. Redelivery subscription quantity. All of that. It’s all in one spot. So the idea is, we’re just gonna give people what they need

55 00:08:56.249 00:09:04.359 Demilade Agboola: but depending on who needs access to what and what decisions they might need on their dash they need from their dashboards. We might now end up like

56 00:09:05.080 00:09:12.159 Demilade Agboola: and kind of like, ensure the numbers that they get are more streamlined to the users of the dashboards.

57 00:09:16.040 00:09:31.260 Amber Lin: Okay, thank you. I think. We presented this to the inventory stakeholders last week, and then we’ll give you a quick overview. What of what’s new in the inventory. And then we can talk about

58 00:09:31.870 00:09:49.539 Amber Lin: how we want to roll out and the expectations on the timeline of what that might change about inventory, because we can wrap it up. The main part is is done, but as we incorporate more stakeholders, it might take more time. So I just want to make sure that we’re all aligned on that

59 00:09:50.750 00:10:12.530 Amber Lin: so quickly today. What we want to go through is 1st about the inventory mark. What we have done also to look at deprecations, and lastly, to make sure that we align on the timeline and expected outcomes, and also risks related to the revenue mark, and making sure that we document all your all your concerns and our goals.

60 00:10:13.960 00:10:32.310 Amber Lin: So after our audit, so we had the original roadmap, where we did a quick audit. I believe that’s late late May, and after we started the phase, 2 collaboration and we did a further audit. We found

61 00:10:32.720 00:10:55.099 Amber Lin: the inventory. Mart is the same findings as before, and for Looker we found that there was around 800 dashboards, 200 views, and around 100 explores, and we found around 4,000 registered tables, which is a lot, and after the work. So after our past 4 sprint, so about 2 months time, we

62 00:10:55.120 00:11:03.619 Amber Lin: moved most of the unused dashboards, most of the dashboards to archive. So currently only 93 dashboards are remaining.

63 00:11:03.940 00:11:14.509 Amber Lin: and for redshift we archived about half of the redshi tables, and also around 500 unused views.

64 00:11:15.440 00:11:25.050 Amber Lin: I want to concentrate a little bit on the inventory mark. So I really think we made a lot of progress from when we 1st started to where we are at now.

65 00:11:25.500 00:11:28.537 Amber Lin: So after we did

66 00:11:29.640 00:11:57.009 Amber Lin: clean up the inventory mart, we are able to find numbers that we’re not able to see before, so that, I think allows inventory stakeholders to make better decisions. We also have data in one central location which reduces the time and errors people have to spend between different reports. And also we can we enable near real time and up-to-date inventory tracking. And I think

67 00:11:57.030 00:12:06.519 Amber Lin: this is what we want to realize across different marts. And that’s that’s the same impact that we will have on the revenue market

68 00:12:06.700 00:12:07.659 Amber Lin: and I think

69 00:12:08.930 00:12:12.989 Demilade Agboola: I just want to quickly add to that point that the yesterday, the central location point.

70 00:12:13.452 00:12:29.500 Demilade Agboola: I know, Felipe, because Philip and I have been like working and working sessions. I know, he said. He had access to these numbers before, so it’s not like they’re necessarily all like brand new numbers, but he will. He mentioned having to go to different spots to get those numbers and have to run his numbers through, dash

71 00:12:29.883 00:12:36.119 Demilade Agboola: and then, having to like, go to different spots, look or dash, and combine everything, to be able to make sense of

72 00:12:36.440 00:12:39.300 Demilade Agboola: the numbers that were just presented to him, all in one line

73 00:12:39.646 00:13:07.249 Demilade Agboola: all in one space. He’s in fact, he’s even been asking when we can get when we can get numbers across to him because he’s looking forward to being able to use them. So yeah, I think that’s the key. Huge advantage. Because I mean, in some cases the numbers did exist. So they’re not like brand new numbers or anything, or brand new calculations. It was just not centralized, and obviously that the man hours of having to go through multiple spots just to get your daily reports or weekly reports, you know, that can be cut out.

74 00:13:13.130 00:13:19.580 Amber Lin: Quickly to give you a sense of what happened for inventory.

75 00:13:20.282 00:13:33.399 Amber Lin: This is was this was the timelines that we that we took on for inventory. So there’s parts before mother’s day, and then the initial parts post mother’s Day.

76 00:13:33.520 00:13:47.910 Amber Lin: And now we’re in the final steps. We’re adding observability tools. And we’re focusing on what exactly needs to go into different dashboards.

77 00:13:48.050 00:14:14.600 Amber Lin: And there was a lot that we learned from working from inventory. That will let us do better when we work with revenue, such as working directly with the stakeholders, will prevent what’s happening right now of having to extend the timeline because we’re receiving new requests. So we’ll show you in a moment what we roughly plan to do for revenue, and when to include different stakeholders.

78 00:14:14.600 00:14:21.059 Amber Lin: and I think that will mitigate the risk of having the timeline or scope increase indefinitely.

79 00:14:23.810 00:14:44.369 Amber Lin: and I’ll let I’ll let Damalade make a quick overview of the inventory, Mart just to just be conscious of your time. I think we have around 14 min left, and I do want to save some time to discuss over revenue. So, Damala, if you want to give like a quick 3 min overview of inventory, that will be great.

80 00:14:46.605 00:14:50.359 Demilade Agboola: So quick overview of inventory. We

81 00:14:50.520 00:14:56.950 Demilade Agboola: set up a new ingestions tool, but atomic dot new tables in adding through

82 00:14:57.510 00:15:03.070 Demilade Agboola: that new data as well as some of the old data that exists through some of the Oms data that exist.

83 00:15:03.740 00:15:08.590 Demilade Agboola: we’ve been able to create different applications for

84 00:15:08.700 00:15:17.860 Demilade Agboola: floral goods and hard goods separately, where we’re able to roll up every single, like every single transaction.

85 00:15:18.320 00:15:26.250 Demilade Agboola: into the inventory of every single like art, good and floral good, and that allows us to be able to quickly see things like

86 00:15:27.050 00:15:37.677 Demilade Agboola: how many, how many items went out the door, how many goods were sold, allows us to see how many of those sold were subscriptions, how many were deliveries, how many were

87 00:15:38.850 00:15:43.520 Demilade Agboola: all committed red delivery, so that means they were not meant to be by different item.

88 00:15:43.997 00:15:54.120 Demilade Agboola: So on committed subscriptions uncommitted with deliveries. We also see things like shrinkage right there as well. You can see things about the different adjustment types. So things like the

89 00:15:54.280 00:15:57.324 Demilade Agboola: how many were marketing? How many were?

90 00:15:59.250 00:16:16.770 Demilade Agboola: how many were markets, in how many were sales? And we can kind of quickly see those different adjustment types, and that allows us to be able to quickly make whatever decisions anybody needs to make writing on spot. Like, I said, we can always streamline based on who needs what and what they need to see.

91 00:16:17.537 00:16:20.219 Demilade Agboola: But that’s just basic high level

92 00:16:22.270 00:16:42.190 Demilade Agboola: view of everything. We’ve been able to create the floral goods. We’re able to create the hard goods and we can roll it up to different Fcs. We can roll it up to different locations, and so that allows us to partition it, and we can also see it by different like dates. So fulfillment dates and purchase dates, and that allows us to ultimately make decisions based off what we need to see in our data.

93 00:16:43.280 00:16:46.099 Demilade Agboola: And so that’s the that’s that’ll be the sum of everything.

94 00:16:46.400 00:17:06.299 Amber Lin: Thank you. I’ll leave these tables here I’ll share, and that we can share the slides with you if you want to close. Look at these tables more closely. So I think. Overall, Demo, I said. We added, non for hard goods and this is just what’s not captured in

95 00:17:06.380 00:17:17.580 Amber Lin: the tables before. So currently, we’re working around 3 working sessions per week with Felipe and identify his pain points. And what’s remaining here is to

96 00:17:17.700 00:17:30.760 Amber Lin: make sure that stakeholders feel comfortable, rolling these numbers out to their dashboards and finalizing documentation for the future. So I think I want to

97 00:17:30.880 00:17:36.039 Amber Lin: pause here and hear your feedback on what you think.

98 00:17:36.410 00:17:59.520 Amber Lin: This our collaboration has went so far. If there’s any ways of working that you think we can improve on any risk that you see that you want us to address? I just want to oh, like, feel feel free to give us any feedback. I’m going to document them here, and then we’ll take action on what you say.

99 00:18:00.445 00:18:13.800 Zack Gibbs: So I guess the 1st concern is that the it’s good that Felipe has been. There’s been collaboration there. What about Perry like Perry Perry is in my mind.

100 00:18:14.540 00:18:18.789 Zack Gibbs: missing from that deeper collaboration. Now

101 00:18:19.110 00:18:21.559 Zack Gibbs: she’s going to be involved in

102 00:18:21.940 00:18:27.210 Zack Gibbs: revenue side, since she’s, you know, heavily involved in the forecasting side, but

103 00:18:27.620 00:18:31.820 Zack Gibbs: she’s also they have a deep connection to inventory. So

104 00:18:33.190 00:18:36.560 Zack Gibbs: is. There been a reason that Perry hasn’t been included in those

105 00:18:38.160 00:18:40.079 Zack Gibbs: and those more deeper dive sessions.

106 00:18:41.286 00:18:41.959 Demilade Agboola: I get.

107 00:18:41.960 00:18:43.390 Amber Lin: Go ahead. Go ahead. Sorry.

108 00:18:43.720 00:18:48.350 Demilade Agboola: Yeah, I believe. At some point was, it was the

109 00:18:48.800 00:18:58.870 Demilade Agboola: the. It was an intersection of a bit of availability at some point. But also, I think emily’s suggestion was, Perry will lean more to like revenue

110 00:18:59.683 00:19:06.880 Demilade Agboola: then, and Felipe’s in person like inventory will be like really really important, because he would see.

111 00:19:07.540 00:19:10.819 Demilade Agboola: because he who’s inventory more often.

112 00:19:11.020 00:19:13.469 Demilade Agboola: but like he relies more on, like the inventory data.

113 00:19:13.900 00:19:27.850 Amber Lin: I see. But, Zack, I I see your point, and I think we still have time this week, and next week we’ll invite Perry to our working sessions and start onboarding her, and that will reduce the time that she needs to get on boarded.

114 00:19:27.850 00:19:32.219 Uttam Kumaran: Was she, I think. Was she out? I feel like she was. There was like some scheduling thing.

115 00:19:32.670 00:19:37.720 Amber Lin: It was slightly scheduled, but I think she’s in now, so we’ll be okay to include her moving forward.

116 00:19:38.430 00:19:39.110 Zack Gibbs: Okay?

117 00:19:40.163 00:19:59.019 Zack Gibbs: What? The remaining items of expand more dashboards. So there’s there’s not a hundred or so dashboards. That are that are out there now? In looker, what’s the to? What percentage of those 100

118 00:19:59.660 00:20:01.777 Zack Gibbs: need to still be rolled out?

119 00:20:04.190 00:20:06.140 Zack Gibbs: where? Where are we in the rollout

120 00:20:06.450 00:20:10.129 Zack Gibbs: phase, and what’s what’s left out of the dashboards that remain.

121 00:20:10.940 00:20:12.540 Amber Lin: I see, I think

122 00:20:13.820 00:20:26.400 Amber Lin: when we talked about rolling out dashboards, a percentage of that is going to be completely new dashboards, because the data allows for different ways of seeing the data.

123 00:20:26.727 00:20:46.710 Amber Lin: I believe for the existing dashboards. That’s what we’re doing when we’re talking to Felipe. And he understands what areas or what dashboard he uses. So for each mark, we’re going to make sure that the stakeholders know how to update their own dashboards, because I believe that’s what we decide before is going to take care of the looker. Dashboards.

124 00:20:48.310 00:21:05.459 Demilade Agboola: Yeah. And also just to add to that certain things will be net, new data net new dashboards in the sense of, hey? This is the data we have your dashboards that give you, that helps you track inventory for floral goods or inventory for hard goods, or maybe both

125 00:21:05.914 00:21:11.740 Demilade Agboola: and you can get to see like if they focus on adjustment types, or if they focus on just the pure sales.

126 00:21:12.222 00:21:14.160 Demilade Agboola: and we can create like those

127 00:21:14.510 00:21:19.409 Demilade Agboola: high level dashboard I think in terms of if you’re trying to incorporate

128 00:21:19.620 00:21:40.629 Demilade Agboola: that those numbers with like currently existing data. So you want to join, based off the lot. Id, to all that data, I think that would have to be more like, based on the analyst handling that. But obviously, in that regard. One of the things we did mention in our call last week was that we we would have a focus or have

129 00:21:40.890 00:21:43.850 Demilade Agboola: some priority on like data governance

130 00:21:44.030 00:22:07.700 Demilade Agboola: like, how do you build your dashboards? How do you ensure? You know good data quality? And how do we ensure that we don’t just get to the points where we were before, where people just build out looks or explores and build out dashboards that don’t serve any purpose after 5 months or 6 months, and it’s not nobody tends to it, just being able to ensure that we have, like good data, governance principles.

131 00:22:08.143 00:22:15.420 Demilade Agboola: And so that whatever dashboards are built are still actually useful and functional. One year, 2 years down the line.

132 00:22:18.989 00:22:22.319 Zack Gibbs: Do we have a sense for? What? What’s the

133 00:22:22.660 00:22:26.630 Zack Gibbs: quantity of net? New dashboards that need to be created? And who’s going to create those?

134 00:22:28.050 00:22:29.960 Zack Gibbs: Do we have that broken down anywhere?

135 00:22:30.540 00:22:37.342 Zack Gibbs: So I feel like that that goes alongside with the data governance and like checking of the work.

136 00:22:38.250 00:22:40.320 Zack Gibbs: that I think we need to.

137 00:22:40.320 00:22:46.219 Uttam Kumaran: Yeah, so this is something. This is something amber that I think should get into that design document that we’re working on.

138 00:22:46.700 00:22:54.550 Uttam Kumaran: For both data marts, which is what are the dashboards that the new data marts are displacing.

139 00:22:54.960 00:23:01.009 Uttam Kumaran: The new ones that need to get created, given the unlock of different views or cuts, and then, like

140 00:23:01.460 00:23:07.390 Uttam Kumaran: ideally, when those turn into tickets on urban stems, Europe or like who’s owning that right.

141 00:23:07.800 00:23:08.300 Amber Lin: Yeah.

142 00:23:08.300 00:23:10.269 Uttam Kumaran: That needs to make it into that document.

143 00:23:10.510 00:23:40.060 Amber Lin: I agree. So currently, we have all the current dashboards that we asked Felipe to present. So those are the one that we have all of those that will be adjusted. I don’t think we have a number of the number of new dashboards that will be created. I think that deserves a conversation directly with Felipe of? Does he want to create new ones? Or if we think we should create new ones? What what are the disagreements there. So I think that was a conversation. I can book that meeting.

144 00:23:44.010 00:23:46.390 Zack Gibbs: Yeah. And then to Ujam’s point.

145 00:23:46.390 00:23:50.910 Amber Lin: Yeah, I’ll include it in the design document, especially for revenue coming up.

146 00:23:52.780 00:23:58.587 Zack Gibbs: Yeah. And then wh, which of the 90 something that are existing today are being displaced?

147 00:23:58.950 00:23:59.400 Amber Lin: Hmm.

148 00:23:59.400 00:24:00.839 Zack Gibbs: Buy these net, new.

149 00:24:00.840 00:24:04.780 Amber Lin: Hmm, okay, which of them?

150 00:24:16.710 00:24:31.309 Amber Lin: Okay, do you think we should also have a data governance training session for the analysts? Or is it more of a documentation that we sent out to them, because that’s also a risk that we should think of how to mitigate.

151 00:24:35.030 00:24:40.159 Uttam Kumaran: yeah, I mean it. It depends. Like, if you’re talking to Felipe and Perry through the whole process, then what

152 00:24:41.160 00:24:45.290 Uttam Kumaran: I don’t think we need like a formal thing. I mean, they should be pretty aware of what the tables are.

153 00:24:46.310 00:24:47.150 Uttam Kumaran: So

154 00:24:47.310 00:24:52.529 Uttam Kumaran: ideally. The goal of involving them early is that we don’t need to do some sort of formal data governance.

155 00:24:53.080 00:24:58.640 Uttam Kumaran: They should be able to know that the tables, what tables are coming out and start to build.

156 00:25:03.170 00:25:07.900 Zack Gibbs: I don’t know. I think that I think that it’s a hybrid approach. In my opinion. One. The the

157 00:25:08.120 00:25:15.880 Zack Gibbs: at the early stages of net new net new based on new new data. Mart.

158 00:25:16.510 00:25:20.980 Zack Gibbs: Oh, there needs to be a like a a checkpoint approach where they

159 00:25:22.110 00:25:28.750 Zack Gibbs: they show their work, and we agree or disagree? Based on what they’re trying to see.

160 00:25:29.550 00:25:34.770 Zack Gibbs: Like I think we need, I think we need to have some type of of stage gate where we’re checking their work.

161 00:25:35.081 00:25:37.568 Amber Lin: Okay, that that makes a lot of sense.

162 00:25:37.880 00:25:40.065 Zack Gibbs: At least at least early on

163 00:25:41.880 00:25:42.720 Amber Lin: Yeah, I think.

164 00:25:42.720 00:25:49.979 Zack Gibbs: Because other, because otherwise the risk is that we’re back in the same place where we’re just. We have all we have a massive amount of junk created that.

165 00:25:51.356 00:26:10.113 Amber Lin: I see, I think both your your approach and approach doesn’t conflict. So we’ll try to onboard them as as early as possible to avoid additional training sessions. And when we 1st roll out, new things and analysts are taking actions, we’ll make sure that

166 00:26:11.270 00:26:37.520 Amber Lin: I think there will need to be a way where we can know what they did. Maybe that’s a reporting session. Maybe Emily helps us, or someone helps us look at what has been done internally during, say, a 1st month period of time. And then for all of those decisions, our team can go look in to say, Okay, that is in accordance to data governance that is not. And then whatever is not will help educate the analysts.

167 00:26:41.240 00:26:47.469 Zack Gibbs: Are we anticipating as some type of redshift savings by the reduction or no.

168 00:26:50.540 00:26:52.598 Uttam Kumaran: I’ll I’ll have to look.

169 00:26:53.760 00:26:56.070 Uttam Kumaran: Storage is pretty cheap on redshift.

170 00:26:56.400 00:27:01.870 Uttam Kumaran: so I don’t think that it’s gonna be the savings are gonna come necessarily from redshift.

171 00:27:02.398 00:27:08.560 Uttam Kumaran: I do think that where we’re gonna get the savings is, it’s it was previously basically like impossible to traverse

172 00:27:08.810 00:27:14.660 Uttam Kumaran: in there. So now it’s actually a lot easier. And so that’s you’ll kind of see that more in the speed up.

173 00:27:15.125 00:27:23.680 Uttam Kumaran: But I will. We can put an action item to measure that once we do that. But again, we’re not really. We haven’t really reduced the click concurrency

174 00:27:24.453 00:27:25.480 Uttam Kumaran: of like

175 00:27:26.170 00:27:32.709 Uttam Kumaran: removing stuff from there doesn’t mean things speed up necessarily. And the storage is really cheap. You know, we’re continuing to like

176 00:27:33.050 00:27:39.109 Uttam Kumaran: reduce the time it takes for models to run, and that’s where I think we’re gonna see most of the spend, but

177 00:27:39.440 00:27:41.370 Uttam Kumaran: can take that on to just double check

178 00:27:43.150 00:27:49.332 Uttam Kumaran: if anything I can. If I if I notice that like, Hey, maybe our cluster is like one size too large. I can have.

179 00:27:49.830 00:27:52.070 Uttam Kumaran: Alex downsize.

180 00:27:53.740 00:28:02.850 Uttam Kumaran: Okay, I’m just looking. I’m just looking at the last 2 months of of redshift costs, and they have remained pretty static. So I didn’t know with reduction that they were supposed to go down by some percentage.

181 00:28:04.340 00:28:11.200 Uttam Kumaran: Yeah, not necessarily. I mean, I feel like, if anything, because we turned off some Etl that may help.

182 00:28:11.690 00:28:12.370 Uttam Kumaran: But

183 00:28:13.320 00:28:23.739 Uttam Kumaran: it’s we’re not like doing a lot. We’re not frankly doing it too much in redshift. Looker is really the the biggest thing that’s running there. So our ability to optimize the looker queries

184 00:28:24.170 00:28:28.510 Uttam Kumaran: move pdts into materialized views.

185 00:28:28.780 00:28:32.209 Uttam Kumaran: That’s where I’m we’re gonna probably see some some bigger gains.

186 00:28:34.710 00:28:41.409 Uttam Kumaran: Because Looker is the primary user. I mean, we we land data in there. We’re doing. Dbt, but as a proportion of compute looker is like.

187 00:28:41.520 00:28:45.060 Uttam Kumaran: really the one that’s that that we’re supporting.

188 00:28:51.030 00:28:58.650 Zack Gibbs: cool. I I have. I do have a hard stop. I gotta run to another call. What’s what’s left that we haven’t covered.

189 00:28:58.830 00:29:15.049 Amber Lin: So quick, I think, just for revenue. We haven’t arrived at a definite timeline. This is what we foresee, based on what we had to do with inventory. We’ll take your feedback and make sure that we incorporate stakeholders early and put that in the design document.

190 00:29:15.050 00:29:37.200 Amber Lin: I think the team is still working on that. We got a few. We got a round of feedback from, so we’re going to improve on it, and probably in in a week’s time, or a little bit more than that, we’ll make sure that we get comments from you, and know that we align on all the risks and expectations for the revenue mart, and so

191 00:29:37.240 00:29:43.929 Amber Lin: just quickly. Right now, I do think it will take also around 2 months

192 00:29:44.100 00:30:01.480 Amber Lin: to get revenue up and running, and probably additional amounts of time for small testings and adjustments. But since we have experience front inventory, this could be a bit faster. But I know revenue is also very complex. So I just want to make sure that we are aligned on the timeline.

193 00:30:04.090 00:30:07.449 Zack Gibbs: When is the technical design, Doc, gonna be ready for review.

194 00:30:09.770 00:30:10.759 Amber Lin: Download it.

195 00:30:13.080 00:30:15.410 Zack Gibbs: And Kyle is responsible for that right?

196 00:30:15.410 00:30:24.800 Uttam Kumaran: Yeah, like, we’re, I think we’re discussing it. I mean, we should. I? I want to have it out next week. I thought we were. Gonna have it out this week. But I I gave some feedback that needs

197 00:30:25.423 00:30:27.789 Uttam Kumaran: to be addressed. So we didn’t have it today.

198 00:30:27.990 00:30:32.389 Uttam Kumaran: But I’d like to have it by end of next week. Like to review.

199 00:30:32.690 00:30:39.738 Zack Gibbs: I would like that to be a like a a sync meeting where? Where Alex is on? I’m on. Emily’s on, and we go through it together.

200 00:30:40.000 00:30:41.600 Amber Lin: Okay. I’ll reach out.

201 00:30:41.600 00:30:55.540 Zack Gibbs: Send it over. Send it over. Yeah. Send over as a the doc as a as a pre, you know. We’ll look at it before the call, but I would like to go through that together, and then go through the timeline together and make sure that we feel good about who we should involve, and general timing.

202 00:30:55.770 00:31:01.139 Amber Lin: Yeah, awesome. I’ll I’ll aim for Thursday or Friday. I’ll reach out in the Channel when you guys are free.

203 00:31:01.680 00:31:05.909 Zack Gibbs: Okay. Can you just a ask, Emily that she’ll be able to coordinate a little faster than I will? At least.

204 00:31:05.910 00:31:06.740 Zack Gibbs: Okay.

205 00:31:07.030 00:31:09.860 Amber Lin: Okay, thanks. I’ll reach out to Emily about that.

206 00:31:10.080 00:31:12.410 Zack Gibbs: Okay. Okay. Alright. Sounds good.

207 00:31:12.410 00:31:12.890 Uttam Kumaran: Thanks, Zach!

208 00:31:12.890 00:31:13.250 Amber Lin: Right.

209 00:31:13.250 00:31:14.570 Zack Gibbs: Alright. Thank you, Zack! See you.

210 00:31:14.570 00:31:15.300 Amber Lin: Bye, thanks.

211 00:31:15.540 00:31:16.300 Demilade Agboola: Nice.