Meeting Title: US x BF | Standup Date: 2025-09-12 Meeting participants: Emily Giant, Amber Lin, Uttam Kumaran


WEBVTT

1 00:00:11.390 00:00:16.470 Amber Lin: Hi, Emily! Sorry, I was stuck in another meeting. We ran a little bit over.

2 00:00:16.470 00:00:18.540 Emily Giant: Oh, no worries. How are you doing?

3 00:00:18.660 00:00:24.590 Amber Lin: Pretty good! I saw that you’re gonna be out of office starting next Monday, right? For 2 weeks?

4 00:00:24.760 00:00:31.980 Emily Giant: I’ll still be available, like, I’m not on vacation or anything like that, but, I am… yeah, I have this two weeks off.

5 00:00:31.980 00:00:47.700 Amber Lin: Okay. I edited the meeting invites, so I have Alex in our retros and groomings, and I have Zach in our planning, and I think…

6 00:00:48.000 00:00:49.630 Amber Lin: Stand-ups.

7 00:00:49.860 00:00:51.850 Emily Giant: Okay, perfect.

8 00:00:51.990 00:00:58.670 Emily Giant: Do you know if Utam’s able to join this meeting? I wanted to just think with him real quick on that, like, collision issue that was happening with Looker.

9 00:00:58.670 00:01:01.140 Amber Lin: Yeah, I just… I was just talking to him, let me…

10 00:01:01.140 00:01:01.700 Emily Giant: Okay.

11 00:01:01.700 00:01:03.779 Amber Lin: Make sure that he can come.

12 00:01:30.640 00:01:36.409 Emily Giant: If not, I can chat with him async, but I just want to make sure that that’s all, like, buttoned up before I’m out of office.

13 00:01:36.700 00:01:37.250 Amber Lin: Yeah.

14 00:01:45.250 00:02:00.030 Amber Lin: And then I am looking at the tickets. I think on your plate is the Looker stuff. Would someone else be taking over that, or would we ask Zach or ask Alex about those?

15 00:02:00.480 00:02:06.889 Emily Giant: No one else would be taking that over. I don’t even think anyone has the access, but, I have a lot of it

16 00:02:07.200 00:02:19.379 Emily Giant: in progress, but it’s, I need to be able to, like, sync with Demolade in order to know, like, what his idea is for those tables in Looker, and, like, how he wants them to be…

17 00:02:19.510 00:02:20.260 Emily Giant: like…

18 00:02:20.500 00:02:31.259 Emily Giant: joined together and, like, present to stakeholders, so I feel like with him being out the last two days and then me being out, it just was, like, difficult to get on the same page there.

19 00:02:31.550 00:02:33.669 Amber Lin: I see. Is it… do you think…

20 00:02:34.040 00:02:41.560 Amber Lin: I just know that when you’re out, we won’t be able to move much on those tasks. Do you think that’s okay for our velocity?

21 00:02:41.560 00:02:42.030 Emily Giant: Yeah.

22 00:02:42.030 00:02:42.500 Amber Lin: It’s totally…

23 00:02:42.500 00:02:56.720 Emily Giant: fine. Like, I think the work that we, are really looking for is the DBT, like, the new model build-outs, the subscription stuff, things of that nature, and then, like, the Looker stuff doesn’t really take that long.

24 00:02:56.960 00:03:02.360 Emily Giant: But, yeah, so I don’t think that’s really gonna be a big blocker.

25 00:03:03.410 00:03:06.189 Emily Giant: Okay, sounds good. It wasn’t, yeah, yeah.

26 00:03:06.640 00:03:07.900 Amber Lin: Yeah, Utom’s here.

27 00:03:08.270 00:03:09.340 Emily Giant: Hey.

28 00:03:09.340 00:03:10.330 Uttam Kumaran: Hello?

29 00:03:10.330 00:03:17.780 Emily Giant: Hi, I just wanted to check in on, if there is progress made with the, like, Looker refresh, the Looker user collision stuff?

30 00:03:18.480 00:03:30.210 Uttam Kumaran: I think it’s a lot better now. I am monitoring this dashboard. Yes, I do also use Looker.

31 00:03:30.750 00:03:33.749 Uttam Kumaran: I created this dashboard to monitor errors.

32 00:03:34.260 00:03:40.209 Uttam Kumaran: cooked this week, and this is because of a change I made around this date.

33 00:03:40.960 00:03:46.239 Uttam Kumaran: I… Reverted and modified, and things are looking better?

34 00:03:46.760 00:03:48.279 Emily Giant: Yeah, a lot better.

35 00:03:48.280 00:03:56.879 Uttam Kumaran: I have a bunch this morning. I… I’m happy to share, kind of, the changes we made, but a lot of them were, around…

36 00:03:57.060 00:04:01.169 Uttam Kumaran: the connection pooling, settings within Looker.

37 00:04:01.950 00:04:03.940 Uttam Kumaran: Made some modifications there.

38 00:04:04.220 00:04:07.490 Uttam Kumaran: Additionally, I’m setting up some…

39 00:04:07.670 00:04:11.029 Uttam Kumaran: Monitoring scripts that will alert us if

40 00:04:11.390 00:04:15.450 Uttam Kumaran: fees start to ramp up, like, that’ll run in dbt.

41 00:04:17.100 00:04:25.210 Uttam Kumaran: And, yeah, I feel like this is in a lot better place. We’ve also made a lot of progress, generally, on jobs.

42 00:04:26.670 00:04:34.329 Uttam Kumaran: I don’t know if you’ve been sort of poking at that, but… Yeah, I have. It does look like it’s going a lot better.

43 00:04:34.490 00:04:41.409 Uttam Kumaran: Yeah, so… There is some… this is… this fix for this is actually in a PR.

44 00:04:42.150 00:04:50.730 Uttam Kumaran: But, If we can just take a look at some of the recent, job runs,

45 00:04:53.390 00:04:57.370 Uttam Kumaran: We can see that inventory is taking, like, sub-10 minutes.

46 00:04:58.030 00:05:05.139 Uttam Kumaran: Which is great, and core model refresh is taking usually, like, 30 to 45 minutes.

47 00:05:05.480 00:05:07.599 Emily Giant: That’s pretty normal, though, so…

48 00:05:07.600 00:05:18.210 Uttam Kumaran: Yeah, I think that’s okay, like, if we look at, really, what is taking quite a while, it’s… there’s some things that take a minute, but these… a lot of these I’ve already… I’ve already pushed fixes for.

49 00:05:18.710 00:05:22.079 Uttam Kumaran: And then there’s…

50 00:05:22.450 00:05:27.600 Uttam Kumaran: couple things, like Tableau items. So my… one of the things that I may propose is that

51 00:05:27.900 00:05:31.100 Uttam Kumaran: If we do, like, an 80-20 on models, and some of these are just.

52 00:05:31.830 00:05:36.240 Uttam Kumaran: like, too big, I would propose moving these to run

53 00:05:36.670 00:05:39.849 Uttam Kumaran: Like, less frequently in another job.

54 00:05:40.160 00:05:42.469 Uttam Kumaran: That way it doesn’t, like.

55 00:05:43.030 00:05:49.489 Uttam Kumaran: The whole job, like, if the job fails in the middle, and like, it just makes testing this really, really difficult.

56 00:05:49.490 00:05:50.020 Emily Giant: Yeah.

57 00:05:50.020 00:05:56.430 Uttam Kumaran: this is an 8-minute… this is an 8-minute thing, or like, I don’t know what is, like, 9-minute or whatever, like…

58 00:05:57.060 00:06:08.620 Uttam Kumaran: Type thing, so I’m kind of thinking about, for the most, like, brutal jobs, moving them out, running them early frequency, and then I can start to isolate and chip away.

59 00:06:09.060 00:06:09.670 Emily Giant: Okay.

60 00:06:09.980 00:06:20.119 Emily Giant: And I’ll do the same, like, I’m looking through this list of models that run as core model, and, like, some of those, like, Zendesk anything does not need to run, ever. We don’t use Zendesk anymore. Like, things like…

61 00:06:20.120 00:06:26.770 Uttam Kumaran: I also have your… what you told me about last week or the week before, that is also PR waiting to go.

62 00:06:26.770 00:06:27.470 Emily Giant: Yeah.

63 00:06:27.470 00:06:29.560 Uttam Kumaran: fixes and stuff like that, so…

64 00:06:29.560 00:06:39.280 Emily Giant: Okay, yeah, and I think that, like, I’m waiting till close of business. Since I’m around for the next, like, 6 days, I don’t want to disrupt anyone’s flows today, so I’m just gonna deploy the,

65 00:06:39.390 00:06:41.029 Emily Giant: Part 1, 2, 3, 4.

66 00:06:41.160 00:06:44.660 Uttam Kumaran: Okay, tell me this week, I’ll be honest, just tell me if you want to do.

67 00:06:44.660 00:06:45.160 Emily Giant: Okay.

68 00:06:45.250 00:06:48.289 Uttam Kumaran: or if stuff starts failing, I’m around, so I don’t…

69 00:06:48.290 00:07:00.590 Emily Giant: Okay. And it should do the opposite. It should make it less likely to fail, and it should, like, lighten up Tableau Items XF with, like, what’s being ingested, yadda yadda yadda. So,

70 00:07:00.900 00:07:11.040 Emily Giant: I just didn’t want to, like, blow up anyone’s day. So I’m like, meh, I’ll just wait to close the business and do some, like, other deprecation work in Looker, etc. today, until…

71 00:07:11.040 00:07:13.630 Uttam Kumaran: Till 5 o’ Okay, great.

72 00:07:15.020 00:07:21.560 Emily Giant: Alright, cool. Okay, I think we’re all good, and, like, you know I’m around, so, like, feel free to reach out, and I might even, like.

73 00:07:22.030 00:07:26.720 Emily Giant: you know, just pop into the stand-up, but I’ll send you guys my WhatsApp information.

74 00:07:26.930 00:07:29.489 Uttam Kumaran: Yeah, I’d love to get these 3 out.

75 00:07:29.490 00:07:30.150 Emily Giant: Yeah.

76 00:07:30.150 00:07:37.410 Uttam Kumaran: And then, as soon as this one is just, like, some renames, then I still have the one where I’m moving folders around, so I want to, like.

77 00:07:37.580 00:07:39.570 Uttam Kumaran: Get through that.

78 00:07:40.030 00:07:41.190 Emily Giant: Yeah, yeah.

79 00:07:41.190 00:07:41.820 Uttam Kumaran: Who’s…

80 00:07:42.020 00:07:44.050 Emily Giant: That sounds good.

81 00:07:44.140 00:07:57.820 Emily Giant: Okay, yeah, I think all of these are, like, fine to go. In part 5, I ran, like, 5 different times last week, and it runs. There are just a couple, like, edge cases that are still, like, on my nerves that I’m trying to tweak.

82 00:07:57.820 00:08:06.600 Emily Giant: And it’s mostly around, like, forced upgrades and stuff, and how they’re appearing in the revenue… like, the historical revenue models. But, like, that’s…

83 00:08:06.600 00:08:14.980 Emily Giant: one teeny tiny edge case, whereas, like, the vast majority of, like, what’s distorting revenue will be fixed.

84 00:08:14.980 00:08:15.300 Uttam Kumaran: Okay.

85 00:08:15.300 00:08:18.340 Emily Giant: So, yeah, and that’s, like,

86 00:08:18.710 00:08:23.660 Emily Giant: Oh, I mean, I’ll tag you in the PR one. I can even do it now, just so you.

87 00:08:23.660 00:08:24.010 Uttam Kumaran: You can, like.

88 00:08:24.010 00:08:26.910 Emily Giant: See it, and, like, know what’s going on with it.

89 00:08:26.910 00:08:27.740 Uttam Kumaran: Fair. But…

90 00:08:27.770 00:08:33.489 Emily Giant: Yeah, okay, I’ll do that, and then, continue to chip away. What I’m doing is, like, just trying to, like.

91 00:08:33.809 00:08:43.769 Emily Giant: backfill some of the missing piece product SKUs in ComponentsXF, so that the quantity is correct. Because right now, like, every now and then.

92 00:08:44.090 00:09:03.770 Emily Giant: there’s, like, incorrect ingestion from the source. This happens mostly in 2022, so I, like, don’t really care that much, but, I’m a perfectionist, so it’s driving me crazy. But yeah, I’ll send that through right now as a PR, or in the next, like, 20, because I have to take away some of the changes that I was messing with, but I know that it runs.

93 00:09:03.770 00:09:06.370 Emily Giant: All the way through Tableau Items XF.

94 00:09:06.370 00:09:07.080 Uttam Kumaran: Okay.

95 00:09:07.080 00:09:11.619 Emily Giant: Yeah, okay, cool. Yeah, I’ll just, tag you in that PR.

96 00:09:11.790 00:09:13.750 Uttam Kumaran: And then how’s it going with cursor stuff?

97 00:09:13.860 00:09:17.570 Emily Giant: It’s going better, like, cursors are sometimes still, like…

98 00:09:17.780 00:09:25.929 Emily Giant: takes, takes liberties that I’m like, stop! But… but yeah, but it’s… it’s going better than it was. It’s just a learning curve.

99 00:09:25.930 00:09:29.780 Uttam Kumaran: And you’re able to just, like, test things… you’re testing things locally, and stuff is, like…

100 00:09:30.400 00:09:32.200 Emily Giant: Yeah. Yeah, totally.

101 00:09:32.620 00:09:36.509 Uttam Kumaran: Do you feel, like, pretty good? Like, are you still using the UI for stuff? Are you, like, all on.

102 00:09:36.510 00:09:37.069 Emily Giant: Huh?

103 00:09:37.070 00:09:38.200 Uttam Kumaran: Okay, great.

104 00:09:38.410 00:09:54.760 Emily Giant: But yeah, I use it for, like, specific things that I know how to control it, and other, like, larger things. Right. I won’t yet, but I’ll still try it, and then if it doesn’t, just because I’m under such a time crunch with leaving, I’ll go back to, like, what I know. But I’m really looking forward to, like, getting deeper in it once I’m back.

105 00:09:55.010 00:09:55.610 Uttam Kumaran: Great.

106 00:09:56.120 00:09:58.570 Emily Giant: Yeah, totally. Alright, well,

107 00:09:58.800 00:10:11.070 Emily Giant: I will tag you in that PR, please let me know if you need anything, and then, I would just send a message to Perry. I’ll, like, ping her with that dashboard, too, so she can see, but I know that she’s, like, the main stakeholder that’s affected.

108 00:10:11.320 00:10:16.190 Uttam Kumaran: Okay, so I’ll… Awesome. I will… I’ll just shoot you that on our channel, or I’ll shoot… Yeah.

109 00:10:16.190 00:10:32.540 Amber Lin: Yeah, for the last one minute, can I get an update on what’s been going on? Are the causes of local failure still unknown, and what are some other items that we’ve been working on? Because I know we’ve deviated from our roadmap because of all the ad hoc that came up.

110 00:10:32.970 00:10:47.549 Uttam Kumaran: Yeah, so I think a couple things. One is, like, we’ve cut the inventory job, like, almost 80% in terms of duration. We’ve cut the core model refresh down from about an hour and a half.

111 00:10:48.030 00:10:51.780 Uttam Kumaran: We are… we are also,

112 00:10:51.920 00:11:01.890 Uttam Kumaran: like, improving the… basically the, like, the query times in Looker, and so as part of that, like, we made some changes that caused some issues, and we…

113 00:11:02.000 00:11:07.309 Uttam Kumaran: Kind of reverted and went a different direction, and it looks like things are okay now.

114 00:11:07.990 00:11:13.830 Uttam Kumaran: you need, I can… I will send, sort of the dashboard that I’m looking at and monitoring.

115 00:11:15.700 00:11:24.999 Uttam Kumaran: And then, yeah, I think, really, it’s… one is also on our, like, development speed and sort of process, so we now have, like, code owners, the…

116 00:11:25.130 00:11:30.100 Uttam Kumaran: The staging, like, us actually staging and testing changes is working.

117 00:11:30.940 00:11:36.860 Uttam Kumaran: And then now, also, we’re all kind of using cursor, and we have, like, our cursor setup working, so…

118 00:11:36.860 00:11:37.580 Amber Lin: Yay!

119 00:11:37.580 00:11:41.929 Uttam Kumaran: Like, our development speed is definitely a lot higher.

120 00:11:42.460 00:11:49.230 Uttam Kumaran: So for me, I mean, what that means for me is, like, I’m just confident that when a job fails, like, we can actually resolve it within a meaningful amount of time.

121 00:11:49.500 00:11:49.870 Emily Giant: Yeah.

122 00:11:49.870 00:12:07.440 Uttam Kumaran: And then we’re… we’re continuing to have, like, alerts and things like that come in, so… yeah, I mean, I think we’re… we’re… we’re a lot better, and we’re… and then the next set of changes is, like, we’re gonna be… we’ve also basically deprecated, like, almost probably, like, 30, 40 models.

123 00:12:07.710 00:12:14.220 Uttam Kumaran: And that’s, again, just sped up jobs and led to less confusion.

124 00:12:14.530 00:12:16.529 Amber Lin: That’s in inventory, right?

125 00:12:17.370 00:12:20.110 Uttam Kumaran: Yeah, that’s related to the inventory work.

126 00:12:20.110 00:12:20.790 Amber Lin: Okay.

127 00:12:20.910 00:12:26.349 Amber Lin: And then, any of the Looker failure stuff, do we know what’s causing that there?

128 00:12:26.590 00:12:34.349 Uttam Kumaran: Yeah, it’s, it’s one that’s tweaks we made in order to optimize, like, basically query time, so…

129 00:12:34.910 00:12:41.110 Uttam Kumaran: There are a lot of issues in Looker right now around things running for a while, or PDTs, or broken looks.

130 00:12:41.110 00:12:42.840 Amber Lin: So we’ve made some tweaks.

131 00:12:42.940 00:13:01.280 Uttam Kumaran: it costs a couple things to go south. We… I moved away from those and, made some other changes. I’m happy to outline those, like, we did some connection pooling, we did some query killers and, like, some connection refresh things, but overall, like, it’s looking a lot better, and that graph will kind of show

132 00:13:01.580 00:13:02.770 Uttam Kumaran: that story.

133 00:13:05.660 00:13:09.700 Amber Lin: And then, we haven’t touched the redshift admin items yet, right?

134 00:13:10.390 00:13:14.499 Uttam Kumaran: I have not dropped anything… oh, no, we have, yeah, I don’t.

135 00:13:14.500 00:13:15.050 Amber Lin: Oh.

136 00:13:15.050 00:13:20.000 Uttam Kumaran: maybe, like, half of the Redshift performance improvements things that I talked about, which is, like.

137 00:13:20.190 00:13:30.990 Uttam Kumaran: user groups, which is a number of connections for Looker user, and generally just getting a wrangle on, like, monitoring for the instance.

138 00:13:32.920 00:13:33.770 Amber Lin: Okay.

139 00:13:35.040 00:13:44.900 Amber Lin: Okay, awesome. I’ll write an update, I’ll send it for you guys to review, might ask for some graphs for more visual ideas, and then I’ll send in the update today.

140 00:13:44.900 00:13:45.700 Uttam Kumaran: Okay, okay.

141 00:13:45.700 00:13:47.149 Amber Lin: Alright, awesome. Thanks a lot.

142 00:13:47.150 00:13:48.369 Emily Giant: Thanks all, bye.

143 00:13:48.370 00:13:49.100 Amber Lin: I…