Meeting Title: Eden Daily Standup Date: 2025-03-13 Meeting participants: Uttam Kumaran, Steven, Demilade Agboola, Robert Tseng, Awaish Kumar, James Freire
WEBVTT
1 00:00:24.180 ⇒ 00:00:25.270 Uttam Kumaran: Hey? Guys, good morning.
2 00:00:38.980 ⇒ 00:00:39.940 Demilade Agboola: Hello!
3 00:00:42.030 ⇒ 00:00:44.169 Demilade Agboola: It’s nice to see Steven.
4 00:00:44.800 ⇒ 00:00:46.080 Demilade Agboola: That’s been a minute.
5 00:00:47.110 ⇒ 00:00:48.520 steven: Hello! Can you hear me?
6 00:00:48.700 ⇒ 00:00:49.170 steven: Yes.
7 00:00:49.700 ⇒ 00:00:55.559 steven: Oh, hello! Sorry I I had to redo all of my
8 00:00:56.274 ⇒ 00:01:00.295 steven: headphones microphones cause I’m on my
9 00:01:01.270 ⇒ 00:01:05.569 steven: my desktop. Now I’m not on my laptop. I was on my laptop before, but Hello! Everyone.
10 00:01:07.620 ⇒ 00:01:08.670 James Freire: Hello!
11 00:01:10.370 ⇒ 00:01:11.340 Uttam Kumaran: Hey! James!
12 00:01:11.340 ⇒ 00:01:12.299 James Freire: How are you?
13 00:01:12.300 ⇒ 00:01:13.610 Uttam Kumaran: Hey? Good! How are you?
14 00:01:13.610 ⇒ 00:01:14.580 James Freire: Doing, good.
15 00:01:15.705 ⇒ 00:01:20.955 Uttam Kumaran: I just wanna maybe I’ll do a brief introduction. James is joining us.
16 00:01:21.490 ⇒ 00:01:35.630 Uttam Kumaran: just to help us on everything. Tableau for this client. But also is a really rock star analyst himself and you know, has a long career in the data business. Maybe, James, I’ll just have you give, like maybe, like a
17 00:01:35.730 ⇒ 00:01:40.400 Uttam Kumaran: brief intro. And then, as we wait for Robert to join, and then we can kind of kick stuff off.
18 00:01:41.030 ⇒ 00:01:42.958 James Freire: Yeah, Hi, I’m James.
19 00:01:43.670 ⇒ 00:01:57.150 James Freire: started. My, I’ve been in it since my teenage years. I started as a unix guy when I was during the.com days, and in my twenties I got into quantitative trading at a hedge fund. That’s kind of what kicked off my
20 00:01:58.132 ⇒ 00:01:59.848 James Freire: liking of data.
21 00:02:00.610 ⇒ 00:02:16.919 James Freire: I worked at aws for 6 years in oracle in the infrastructure space, doing data science. And and past couple of years I’ve been doing, you know, consulting slash contracting with the, you know, data engineering. And you know, data analytics.
22 00:02:17.110 ⇒ 00:02:19.490 James Freire: So that’s my short life story.
23 00:02:21.780 ⇒ 00:02:33.475 Uttam Kumaran: Yeah, and you have you have Steven, who is our Pm. On, on on Eden, and a few other projects. You also have a wish and demalade who are both analytics, engineers.
24 00:02:34.000 ⇒ 00:02:35.729 Uttam Kumaran: I’m of course, here as well.
25 00:02:36.161 ⇒ 00:02:41.649 Uttam Kumaran: I think we’re. I know Sahana. May not be able to join today. But let me just go ahead and
26 00:02:43.720 ⇒ 00:02:46.419 Uttam Kumaran: let me ping Robert and check
27 00:02:48.100 ⇒ 00:03:00.187 Uttam Kumaran: But yeah, I think today, really, I wanted to just talk through. Everything’s in everything that’s in flight. And I think Steven, the biggest help. And maybe we can talk about this today and sort of get everyone’s perspective.
28 00:03:00.900 ⇒ 00:03:12.110 Uttam Kumaran: is just about like how to best plan for this client. I think we we’ve had several things in flight around dashboards around data quality and I think the team
29 00:03:12.934 ⇒ 00:03:18.709 Uttam Kumaran: could really use some love on just getting things planned out. At least.
30 00:03:19.010 ⇒ 00:03:21.349 Uttam Kumaran: you know, on a day to day basis.
31 00:03:21.950 ⇒ 00:03:24.549 Uttam Kumaran: But of course, like trying to plan a little bit
32 00:03:25.020 ⇒ 00:03:28.430 Uttam Kumaran: towards the future. Robert said. He’s gonna be a few
33 00:03:28.580 ⇒ 00:03:32.799 Uttam Kumaran: minutes late. But maybe we can just start there.
34 00:03:33.230 ⇒ 00:03:39.040 Uttam Kumaran: I guess maybe Steve and I probably I don’t know if you had any chance to sort of look through
35 00:03:39.160 ⇒ 00:03:45.670 Uttam Kumaran: this client yet, or if not, we can, we can each sort of go through and talk about like what we’ve been handling.
36 00:03:45.980 ⇒ 00:03:52.260 steven: Yeah, I I haven’t really had the chance to get get too much into so.
37 00:03:52.860 ⇒ 00:03:54.240 Robert Tseng: Guys. Sorry I’m late.
38 00:03:54.600 ⇒ 00:03:55.260 Uttam Kumaran: Okay.
39 00:03:56.200 ⇒ 00:04:02.709 Uttam Kumaran: I think, Robert, I’m just trying to see what’s the best use of time here. I want to make sure that we both can talk about
40 00:04:02.860 ⇒ 00:04:14.050 Uttam Kumaran: things that are that need to happen, you know, soon, but also gives even an ability to start to move as much of this stuff to linear for tracking as possible.
41 00:04:14.050 ⇒ 00:04:23.550 Robert Tseng: I already met with Akash on this last night, and kind of get walked him through the notion. Timeline he already said that he was, gonna create some projects and move all the existing cards over.
42 00:04:23.950 ⇒ 00:04:29.740 Uttam Kumaran: Okay? And then, so I think, like, do we want to have Steven work with him on that? Or how do you see that happening.
43 00:04:30.223 ⇒ 00:04:39.956 Robert Tseng: Well, yeah. So right now, I mean, Akash has been pretty. I mean, he’s been plugging in. He’s been meeting with clients and stuff, so I don’t really know what the handoff is between Akash and Steven.
44 00:04:40.330 ⇒ 00:04:45.660 Uttam Kumaran: At at this point, Josh. Gonna handle the mix panel work, or do you see him also? Like
45 00:04:45.880 ⇒ 00:04:49.150 Uttam Kumaran: like helping on Pm, work like, how do you see that.
46 00:04:49.454 ⇒ 00:05:00.110 Robert Tseng: I, I kind of thought, was just gonna help on the mixed panel segment stuff. But I think he’s also just kind of kick got got the ball rolling on the on the Pm. Work.
47 00:05:00.280 ⇒ 00:05:01.069 Uttam Kumaran: Okay. Cool.
48 00:05:01.070 ⇒ 00:05:01.799 steven: Yeah,
49 00:05:02.630 ⇒ 00:05:07.239 steven: I’m also meeting with him later today. So it’s it’s something I can bring up.
50 00:05:07.490 ⇒ 00:05:21.539 Uttam Kumaran: Okay. Great. Then, yeah. I mean, I think, for for if I if I was to be the voice of folks on edge, I think definitely. It would be really, really helpful to start to see things end up there. So we can start to plan a little bit in the future.
51 00:05:22.295 ⇒ 00:05:26.500 Uttam Kumaran: Especially around work for data quality work. For
52 00:05:26.700 ⇒ 00:05:28.969 Uttam Kumaran: you know, actual data model changes.
53 00:05:29.378 ⇒ 00:05:45.249 Uttam Kumaran: And I know definitely for looking at structures around the dashboards that need to go out and what their statuses are. So yeah, if you can work with Akash, Steven, and sort of take that over. I know there also are a couple of work streams coming down the pipeline like the mix panel work.
54 00:05:45.420 ⇒ 00:05:48.130 Uttam Kumaran: So maybe I’ll let you guys handle that.
55 00:05:48.840 ⇒ 00:05:49.693 Uttam Kumaran: I guess.
56 00:05:50.460 ⇒ 00:05:56.159 Uttam Kumaran: Robert, do you think it’s best? We go around the Horn like? Do you want to kick this off somewhere in particular?
57 00:05:57.142 ⇒ 00:06:12.659 Robert Tseng: Well, if I can just like, say, I think I’m just, we’re spinning our wheels on tableau deployment issues, which I think has just been really frustrating. I mean, like I I don’t. I don’t really know what to say there, but like every dashboard we put out is just wrong because it doesn’t use refresh data.
58 00:06:12.770 ⇒ 00:06:19.980 Robert Tseng: I mean, I’m I already said that I’ve never administered tableau before I did what I could.
59 00:06:20.110 ⇒ 00:06:37.390 Robert Tseng: I mean. I don’t know why Sahana’s not able to help like she just I don’t know. I just like I’m just having to answer the same. I’m like stuck in this. I’m like stuck in the matrix, and like people are just asking me the same questions every day. Why, this is off. Why is that? Why, that’s off.
60 00:06:37.510 ⇒ 00:07:04.840 Robert Tseng: And I feel like that’s that’s taken up way. Too much time. I just I need somebody to come in and like Redeploy, or kind of take over tableau admin like I’m clearly not able to do it. And I don’t know. Sahana clearly is unable to do it as well, so that to me is like the immediate thing to. I think this. I think this team spent half their time on Eden yesterday kind of just chasing rabbit holes like
61 00:07:05.190 ⇒ 00:07:14.839 Robert Tseng: things that didn’t that were not actually broken. It was just the data was not, was not updated. It was. And like, I think we’re just stuck in that right now.
62 00:07:18.580 ⇒ 00:07:19.847 Uttam Kumaran: Though I think
63 00:07:20.620 ⇒ 00:07:32.580 Uttam Kumaran: I mean like, I guess. Stephen, do you want to run this meeting? Do you want me to go ahead and handle it. I mean, I would love for for us to just focus today on just how to tackle this issue, and who on the team is capable of.
64 00:07:32.680 ⇒ 00:07:38.990 Uttam Kumaran: you know, being able to own that and and tackle each of these. But if you want to run this, then I can just play backup.
65 00:07:39.570 ⇒ 00:07:44.880 steven: Yeah. So you’re basically want me to like, kind of dig in and see like, who on the team is gonna be able to like.
66 00:07:45.060 ⇒ 00:07:46.430 steven: run tableau.
67 00:07:49.072 ⇒ 00:07:53.030 steven: Okay, yeah, I can. I can dig in and try to find someone. So.
68 00:07:53.030 ⇒ 00:07:53.670 Uttam Kumaran: Okay?
69 00:07:53.890 ⇒ 00:08:12.233 Uttam Kumaran: And then I think, within, James, between you, Demo, like, what do you guys think here like, I sort of do want to take this piece off of Robert’s plate again. It’s totally right in that. We had data that was right. And it’s just stuff in tableau is sort of fucked up. What should we do?
70 00:08:14.560 ⇒ 00:08:19.719 James Freire: Yeah. So I mean the the scheduling thing. I mean, there’s all the
71 00:08:19.830 ⇒ 00:08:26.707 James Freire: the the scheduling around data refreshes and alerts and data quality checks to make sure it works properly.
72 00:08:27.779 ⇒ 00:08:38.169 James Freire: I don’t. You know. It’s I think it’s kind of like a workflow when something is brought in that it needs to have a schedule already put in place. I mean, once you connect. I don’t think there’s a default
73 00:08:38.890 ⇒ 00:08:50.731 James Freire: I don’t remember. There’s like a default setting for refreshes. When you when you connect and do an extract, it’s, I think initially, it’s just a 1 and done type of thing. I think that’s the default setting on it. So
74 00:08:51.230 ⇒ 00:08:55.789 James Freire: you know, that’s more of just like the logistics of consistency in terms of when
75 00:08:56.190 ⇒ 00:08:59.769 James Freire: you know somebody’s connected. And you’re creating a dashboard and making sure it runs.
76 00:09:01.500 ⇒ 00:09:04.640 Uttam Kumaran: So I guess, like, my, yeah, go ahead. Awash.
77 00:09:05.340 ⇒ 00:09:14.190 Awaish Kumar: Yeah, like, when we create a extract. So I’m not sure if tableau provides any functionality to.
78 00:09:14.310 ⇒ 00:09:21.270 Awaish Kumar: to to like schedule, this extract refresh. But in my previous job we had airflow.
79 00:09:21.470 ⇒ 00:09:36.639 Awaish Kumar: So we have been doing like whenever the data there is any data model changes, we just have an another task after that which refreshes the tableau extracts. So that way we were keeping the dashboards up to date.
80 00:09:36.860 ⇒ 00:09:39.199 Awaish Kumar: So like we have to like.
81 00:09:39.320 ⇒ 00:09:46.910 Awaish Kumar: But I I’m not sure if if tableau itself has built in feature to to refresh its extracts on on a given
82 00:09:46.910 ⇒ 00:09:49.260 Awaish Kumar: dynamically, yeah.
83 00:09:50.110 ⇒ 00:09:50.840 Robert Tseng: Yeah, I mean.
84 00:09:50.840 ⇒ 00:10:09.352 Robert Tseng: I mean, as far as I’m concerned, I downloaded and republished every live dashboard. I I updated all the data sources. They should all be using my credentials to schedule refreshes at 5 Am. Eastern every day like, that’s that’s the way that I set it up. And
85 00:10:10.370 ⇒ 00:10:11.510 Robert Tseng: yeah, I think
86 00:10:12.103 ⇒ 00:10:17.490 Robert Tseng: this morning there’s like one that’s broken. So I’m gonna go and like, check them all again. But like.
87 00:10:18.000 ⇒ 00:10:24.990 Uttam Kumaran: But let’s so. So, even for that issue, let’s hand it off. Robert. And I just want to know one person on this call that can go own this
88 00:10:25.978 ⇒ 00:10:29.469 Uttam Kumaran: and sort of one document existing refreshes.
89 00:10:31.070 ⇒ 00:10:34.710 Uttam Kumaran: What’s the schedules for each refresh which tables and like
90 00:10:34.830 ⇒ 00:10:39.040 Uttam Kumaran: basically owning the process of like, Hey, we’re confident that
91 00:10:39.140 ⇒ 00:10:45.659 Uttam Kumaran: these are gonna run. And if things are breaking, there is a triage process that happens ideally in slack.
92 00:10:46.810 ⇒ 00:10:58.750 James Freire: Also, I mean, there’s when I hear refreshes. I mean, there’s also all the data flows. You have going in the background to to bigquery. And those things, too, I mean, are there problems with.
93 00:10:58.750 ⇒ 00:11:07.079 Uttam Kumaran: There’s yeah. So that we’re we’re pretty good on. We are running those daily, and there haven’t haven’t been any failures
94 00:11:07.570 ⇒ 00:11:08.670 Uttam Kumaran: there.
95 00:11:08.800 ⇒ 00:11:12.840 Uttam Kumaran: So I think it’s a good assumption just for this exercise to just say
96 00:11:12.960 ⇒ 00:11:15.769 Uttam Kumaran: we need the tableau just to run every day.
97 00:11:16.080 ⇒ 00:11:29.390 Uttam Kumaran: 1 1 thing we can align on after just fixing, that is to say, Hey, we want to make sure that the data source refreshes like the actual data models get refreshed ideally before and complete before these extracts happen.
98 00:11:29.730 ⇒ 00:11:33.080 Uttam Kumaran: I think that is a very logical like next step, for sure.
99 00:11:34.700 ⇒ 00:11:38.510 James Freire: Okay, gotcha. Sorry I I got my dog barking. It’s kind of
100 00:11:38.820 ⇒ 00:11:41.650 James Freire: I’m beginning to distract you guys. You see me looking around.
101 00:11:42.570 ⇒ 00:11:51.400 Demilade Agboola: Yeah. So I I like what? I was just said the like my previous company as well. It was literally part of the airflow.
102 00:11:52.180 ⇒ 00:12:09.449 Demilade Agboola: environment such that like, once we had our runs, our Dbt runs. Then we had our tableau extracts refresh. Yeah. So I think we could try and see how we could schedule it into Github. It’s not something I have access nothing I’ve done before, but something I’m willing to look at if it’s something. But
103 00:12:09.800 ⇒ 00:12:17.869 Demilade Agboola: if it’s something I can just take it off the plate. Obviously, we need to figure out like a short term solution, or like a more reliable short term solution.
104 00:12:18.000 ⇒ 00:12:20.080 Demilade Agboola: But yeah, something to look at.
105 00:12:21.210 ⇒ 00:12:31.740 Robert Tseng: Yeah, I mean, we only have, like 4 or 5 dashboards floating around. I think the short term solution is, I’m not even. I’m just gonna go in and manually in cloud, in tableau cloud. Just
106 00:12:31.930 ⇒ 00:12:36.520 Robert Tseng: press refresh on the extract, and then just let it run like.
107 00:12:36.520 ⇒ 00:12:43.300 Uttam Kumaran: Yeah, the problem’s a lot smaller guys like it’s really just like it needs to run every day. And if it doesn’t run, we need an alert in slack.
108 00:12:43.740 ⇒ 00:12:44.406 James Freire: Yeah, the
109 00:12:44.740 ⇒ 00:12:45.060 Uttam Kumaran: Yeah.
110 00:12:45.060 ⇒ 00:12:54.169 James Freire: There is, I mean, tableau hasn’t has a an email alerting thing, for when the extracts run or don’t run.
111 00:12:54.170 ⇒ 00:13:03.669 Robert Tseng: Yeah, I mean, I I scheduled those based on what you said, James. And I’ve been looking at those. I mean, I I see that one of them failed this morning, so I haven’t looked at it yet, but like.
112 00:13:03.670 ⇒ 00:13:05.720 Uttam Kumaran: But can we have those come to slack.
113 00:13:06.570 ⇒ 00:13:07.230 James Freire: You know.
114 00:13:07.230 ⇒ 00:13:10.069 Robert Tseng: The tablet slack integration. Does anything.
115 00:13:10.070 ⇒ 00:13:15.709 Uttam Kumaran: No, I’ll just get you an email. I’ll get you the email of the slack channel, and it’ll just send directly there.
116 00:13:16.188 ⇒ 00:13:18.099 Robert Tseng: Okay, yeah, we should.
117 00:13:18.100 ⇒ 00:13:19.730 Uttam Kumaran: So I’m gonna do that. It’ll be.
118 00:13:19.920 ⇒ 00:13:28.570 Uttam Kumaran: It’ll be noisy. But I want to get this off of your plate and start to have someone on the team. That’s the 1st line of defense on going and checking this
119 00:13:30.520 ⇒ 00:13:39.860 Uttam Kumaran: and let’s do that. I’ll get. I’ll I’ll get you the. I’ll get you the email. We can swap this out, too. But like, I don’t know. James, do you think this is up your alley? You want to take this.
120 00:13:39.860 ⇒ 00:13:41.469 James Freire: Yeah, that’s not a problem at all. Cool.
121 00:13:41.470 ⇒ 00:13:42.819 Uttam Kumaran: Okay, perfect. I mean.
122 00:13:43.200 ⇒ 00:13:47.969 Uttam Kumaran: I mean, like, I, I did used to do a bunch of tableau refreshes. Yeah, I think
123 00:13:48.260 ⇒ 00:13:55.789 Uttam Kumaran: one thing is, we just need to make sure that all these run and that we get alerts when they’re not running. That’s the 1st thing, the second piece.
124 00:13:56.190 ⇒ 00:13:58.280 Uttam Kumaran: And for them. It’s it’s really not
125 00:13:58.430 ⇒ 00:14:09.210 Uttam Kumaran: 100% of concern right now, beyond just full data accuracy is that we want to have the previous on previous day close. We want to be able to measure that data today.
126 00:14:09.370 ⇒ 00:14:14.480 Uttam Kumaran: So the next piece, of course, is to refresh the Dbt data models.
127 00:14:15.014 ⇒ 00:14:28.860 Uttam Kumaran: Prior to running. This. We can do that manually, and then, of course, I think the the most elegant solution is to kick off the extracts dynamically, right after the jobs are done. I’m less concerned about that.
128 00:14:30.220 ⇒ 00:14:35.469 Uttam Kumaran: So that’s the that’s 1 piece. Yeah. So that’s so. That’s I think, one piece taken care of today.
129 00:14:35.966 ⇒ 00:14:41.989 Uttam Kumaran: I feel pretty good about that, Steven. If we can get that into linear that way, we can talk about it
130 00:14:42.481 ⇒ 00:14:45.609 Uttam Kumaran: and assigned to James. That would be perfect.
131 00:14:45.760 ⇒ 00:14:50.320 steven: Okay, so that’s just like the daily extract assignment to James.
132 00:14:50.320 ⇒ 00:14:56.189 Uttam Kumaran: Exactly. That’s just like daily extract and and the slack integration.
133 00:14:57.830 ⇒ 00:15:03.250 steven: Sounds good, so I don’t have to go chasing someone to find out to do this. James is gonna do it.
134 00:15:03.250 ⇒ 00:15:03.905 Uttam Kumaran: Yes.
135 00:15:04.610 ⇒ 00:15:05.700 steven: Yeah, it’s not.
136 00:15:05.700 ⇒ 00:15:06.160 James Freire: No problem.
137 00:15:06.780 ⇒ 00:15:07.669 Uttam Kumaran: Okay. Cool.
138 00:15:08.370 ⇒ 00:15:09.300 Uttam Kumaran: What else?
139 00:15:10.880 ⇒ 00:15:25.160 Demilade Agboola: Just something to mention, and Robert has said we should put it in the to do like in the back, in the backlog of things to do is rebecca mentioned something about once in a dashboard on the visibility of order counts by pharmacy.
140 00:15:25.764 ⇒ 00:15:28.679 Demilade Agboola: So that is something just like further ahead.
141 00:15:29.565 ⇒ 00:15:30.960 Demilade Agboola: But at least
142 00:15:31.360 ⇒ 00:15:37.480 Demilade Agboola: from my perspective of things, a lot of the issues we’re having seem to be more on the Bi side.
143 00:15:39.050 ⇒ 00:15:40.736 Demilade Agboola: So potentially
144 00:15:41.880 ⇒ 00:15:49.880 Demilade Agboola: I’m thinking, like, from the analytics engineering side, is there something we could get ahead of so that it doesn’t all just come crashing in like subsequent weeks.
145 00:15:53.040 ⇒ 00:15:54.790 Robert Tseng: Sorry. Was that a question for me?
146 00:15:55.380 ⇒ 00:16:11.070 Demilade Agboola: No, yeah. So sort of like, is there something that we need to start like looking at or like trying to put in place? So that if there are questions that we need to that need answers in the next couple of weeks. We’re not necessarily having to scramble to answer them like
147 00:16:12.400 ⇒ 00:16:13.410 Demilade Agboola: on the fly.
148 00:16:13.750 ⇒ 00:16:22.560 Robert Tseng: Yeah. I mean, this goes back to Utah. We haven’t really gone back to a roadmap with with this team. We’ve kind of just been playing catch up. So
149 00:16:22.690 ⇒ 00:16:35.720 Robert Tseng: yeah, I mean, there’s some stuff I don’t know. Maybe this is like a conversation we should kick for Tom. I don’t know if we’re meeting tomorrow, but a after like Steven and and Akash
150 00:16:36.030 ⇒ 00:16:37.280 Robert Tseng: do the
151 00:16:37.600 ⇒ 00:16:47.900 Robert Tseng: kind of clean up in in linear. Then maybe that’ll be a better starting point to kind of reintroduce new things. I can really just tell you like what we have for the rest of the week, but I think
152 00:16:48.510 ⇒ 00:16:50.089 Robert Tseng: I don’t know if I can really
153 00:16:50.740 ⇒ 00:16:53.890 Robert Tseng: prioritize like, what’s coming in the next few weeks. Yeah.
154 00:16:53.890 ⇒ 00:17:00.559 Uttam Kumaran: Yeah, I think one ticket, probably for for Steven, for you and Akash to work on is the roadmap here.
155 00:17:02.590 ⇒ 00:17:06.740 Uttam Kumaran: But I think demolati will wait for for them to get caught up, so
156 00:17:07.089 ⇒ 00:17:09.250 Uttam Kumaran: bear us another few days.
157 00:17:09.660 ⇒ 00:17:15.810 Awaish Kumar: And okay. And do you have, do you have any update on this membership plan? Field.
158 00:17:20.482 ⇒ 00:17:26.850 Robert Tseng: I’m sorry. I wish I don’t have context on that. I don’t know if you’ve if you sent me any message this morning. I haven’t looked at it yet, so.
159 00:17:28.191 ⇒ 00:17:33.829 Awaish Kumar: I just shared this sheet yesterday with you with the sample data protocols. Right? Yeah. Yeah. Sure
160 00:17:33.830 ⇒ 00:17:35.130 Awaish Kumar: with Adrian team. Yeah.
161 00:17:38.090 ⇒ 00:17:53.439 Robert Tseng: no, I didn’t get a response on that yet. I’m have, I will. But I did throw Akash and I were talking to them at 11 Am. Today. So Eastern. So in an hour I will. I will get an answer, because I had to ask him about a couple of other things too.
162 00:17:55.420 ⇒ 00:17:56.619 Awaish Kumar: Okay. No worries.
163 00:17:57.080 ⇒ 00:17:58.250 Uttam Kumaran: Okay, I think.
164 00:17:59.150 ⇒ 00:18:04.219 Uttam Kumaran: yeah. A wish. Can you? Can you work with Steven as well to make sure that that gets tracked in linear.
165 00:18:04.740 ⇒ 00:18:25.781 Robert Tseng: So I guess, like a wish 1 1 more. One more thought on that one. So you know, basically like the product mappings dam a lot couldn’t query from it. And then this, like, you know, seems like we fell out of sync with the membership plans because the product mapping sheet that we use is a Google sheet maintained by the pharmacy team.
166 00:18:26.600 ⇒ 00:18:47.950 Robert Tseng: they, you know. I guess it looks like they didn’t actually mean they haven’t been maintaining it, and for whatever reason they they booted our access, and, like that kind of threw some stuff off. So I I kind of sent that. I blasted that message out to the client channel to let them know. That’s why some of the dependencies on, like the downstream dependencies kind of broke.
167 00:18:48.060 ⇒ 00:18:52.459 Robert Tseng: But yeah, it is kind of weird that, or I don’t know what we can do to.
168 00:18:53.640 ⇒ 00:18:54.750 Awaish Kumar: Protect.
169 00:18:54.920 ⇒ 00:19:10.060 Robert Tseng: Or like, establish that process more like it. It is weird. It is not great that we’re all that we’re depend, all of our product variant and membership level. Kind of modeling is dependent on a Google sheet that
170 00:19:10.260 ⇒ 00:19:14.100 Robert Tseng: a non technical team maintains. Like I I
171 00:19:14.400 ⇒ 00:19:25.240 Robert Tseng: I I think that is kind of like a critical like vulnerability for us like who knows what? If somebody just boots us out of the sheet again, and then everything will break again, you know.
172 00:19:30.170 ⇒ 00:19:31.659 Uttam Kumaran: Yes, I I mean.
173 00:19:32.230 ⇒ 00:19:40.289 Uttam Kumaran: I feel like this is a I want to push this back on Demo A and a waste to answer. But can we add another ticket, Steven, on just
174 00:19:40.570 ⇒ 00:19:42.000 Uttam Kumaran: like, identify.
175 00:19:42.000 ⇒ 00:19:42.730 Awaish Kumar: Yeah, but.
176 00:19:42.730 ⇒ 00:19:45.850 Uttam Kumaran: How to lower the risk. Here. There’s a there’s options, of course.
177 00:19:47.460 ⇒ 00:20:02.239 Awaish Kumar: Like for for these membership plan, like, there are only 2 ways we can get it. Either we get it from order information we get. So we know that, okay, this order was placed and what was the membership plan?
178 00:20:03.027 ⇒ 00:20:07.842 Awaish Kumar: When that order was placed? Second, this is a business like.
179 00:20:09.961 ⇒ 00:20:20.530 Awaish Kumar: this is a information like which can which only the business guys can give us like on a variant level. What type of membership they are
180 00:20:21.370 ⇒ 00:20:24.810 Awaish Kumar: are supporting, or
181 00:20:26.310 ⇒ 00:20:36.099 Awaish Kumar: so like from there, like, even though, like I have seen the some of the product names have the names like monthly and
182 00:20:36.290 ⇒ 00:20:41.850 Awaish Kumar: and Quarterly or something like that. In their name, it’s product name itself.
183 00:20:42.040 ⇒ 00:20:47.850 Awaish Kumar: So even if we use that like, we, we are still dependent on the input from
184 00:20:48.450 ⇒ 00:20:54.279 Awaish Kumar: from the business. Like the client itself, like they, they share us with us like the variant name.
185 00:20:55.120 ⇒ 00:21:00.180 Awaish Kumar: And then we, using that, we figure out what membership plan. It is
186 00:21:00.860 ⇒ 00:21:06.240 Awaish Kumar: so. Either we get it in a separate field or we get it from a product name itself.
187 00:21:07.620 ⇒ 00:21:12.120 Awaish Kumar: It it’s going to do, remain dependent on the client
188 00:21:12.400 ⇒ 00:21:15.940 Awaish Kumar: who who, who are responsible for filling that sheet.
189 00:21:20.650 ⇒ 00:21:28.980 James Freire: Is this thing where the sheet like when we got our access cut to it, I mean, are we pulling data off of it and storing it somewhere else in the meantime, or is it we’re just using it. Live.
190 00:21:29.370 ⇒ 00:21:40.230 Uttam Kumaran: Yeah, it’s a lot. It’s a live view. We should. I mean, we should move. We don’t. We should. We could move this to just get pulled on a frequency. That’s 1 short term. Fix.
191 00:21:40.230 ⇒ 00:21:44.800 James Freire: I mean, you could just have like, I don’t know if you have access to the data fusion, but you just have it
192 00:21:44.930 ⇒ 00:21:45.640 James Freire: connect to a.
193 00:21:45.640 ⇒ 00:21:46.320 Awaish Kumar: Yeah.
194 00:21:46.320 ⇒ 00:21:48.820 James Freire: The sheet and then put into a Google Cloud bucket.
195 00:21:49.550 ⇒ 00:21:50.070 Uttam Kumaran: Yeah.
196 00:21:51.530 ⇒ 00:21:52.570 Awaish Kumar: I think.
197 00:21:52.570 ⇒ 00:22:04.160 Uttam Kumaran: That doesn’t solve oasis. Yeah, I I agree totally in terms of like, in terms of fixing this off issue 100. I think oasis problem. Is that sorry? I’ll just. I’m just gonna
198 00:22:04.380 ⇒ 00:22:06.190 Uttam Kumaran: cut to the chase is that
199 00:22:06.460 ⇒ 00:22:26.809 Uttam Kumaran: there, this data doesn’t exist. This knowledge doesn’t exist anywhere else. We could either get it from the product names which is not coming in a standardized way, or we need to get it from this from something like a relational database which also doesn’t exist. So this is our our current alternative, like
200 00:22:27.150 ⇒ 00:22:45.470 Uttam Kumaran: the there’s like, there’s not a better solution unless the business has this in a more fortified way for us to get this. I mean, we could we could do one. We could do one more thing. The the last option here is that we need to do a meeting every month, or on whatever cadence is to
201 00:22:45.470 ⇒ 00:22:59.059 Uttam Kumaran: save this as a static file, and then we own the process by which it gets updated on some cadence. Right? We run a meeting with whoever it is to say, tell us what’s updated. And we just say this is a seed guys like directly in Dbt.
202 00:23:01.510 ⇒ 00:23:05.700 James Freire: Somebody’s putting. But somebody’s typing that data into the spreadsheet somewhere. Right? I mean.
203 00:23:05.700 ⇒ 00:23:06.990 Uttam Kumaran: Yes. Yeah. Yeah.
204 00:23:06.990 ⇒ 00:23:09.969 James Freire: Where are they getting? It is? I always wondered. Like, where are they getting it from?
205 00:23:09.970 ⇒ 00:23:30.319 Uttam Kumaran: That’s also what I’m like, yeah, we’re. We’re just as good as this process is. This process is like someone. That’s what I’m saying is like we can either go to that source or we need to run some sort of meeting on a weekly or monthly basis to to triage what the new inputs are, Robert like this is how these sorts of things we typically go.
206 00:23:30.580 ⇒ 00:23:35.079 Uttam Kumaran: This is the next evolution. For, like some of one of these, some of these Google sheet processes basically.
207 00:23:35.080 ⇒ 00:23:40.463 Robert Tseng: Yeah, if I can just flash this for like 30 seconds. Can you guys see this?
208 00:23:42.190 ⇒ 00:23:46.930 Robert Tseng: Yeah, yeah, I don’t know. Can can everyone see this.
209 00:23:47.100 ⇒ 00:23:47.640 James Freire: Yes.
210 00:23:47.640 ⇒ 00:23:58.270 Robert Tseng: Yeah. Alright, yeah. So this is the Google sheet. So they have a bunch of tabs here. A lot of it is just kind of like noise this sheet. One is the main one they maintain.
211 00:23:58.670 ⇒ 00:24:14.419 Robert Tseng: but there’s just a lot going on here, and so we, and like not all of it is helpful. So we worked with them one time to I, we kind of came up with all these different fields, had them basically translate stuff from sheet one into this format that’s easier for us to consume.
212 00:24:14.740 ⇒ 00:24:25.789 Robert Tseng: But then they just don’t maintain it because they don’t understand like what this is really. And they just go back to just maintaining this. So either, we end up. I mean, I just, I, I think.
213 00:24:26.450 ⇒ 00:24:50.119 Robert Tseng: yeah, I guess I’m hearing you guys are saying, either go to the source directly and skip this sheet entirely, or we need to figure out like, is there a monthly or weekly cadence that we need to have with them where we’re translating, sheet one into this into this sheet. It’s quite manual, because, like, I don’t exactly know how this flows into to the sheet. But yeah, we are.
214 00:24:50.120 ⇒ 00:24:52.079 Uttam Kumaran: But they are updating sheet one.
215 00:24:52.510 ⇒ 00:24:54.130 Robert Tseng: That’s what they say, yeah.
216 00:24:54.490 ⇒ 00:24:59.529 Uttam Kumaran: Okay. But I mean so roughly this, this seems to me that this is a process we should own.
217 00:24:59.650 ⇒ 00:25:01.420 Uttam Kumaran: and that we should do this on.
218 00:25:01.610 ⇒ 00:25:05.669 Uttam Kumaran: We should bash these changes on a month, on a weekly or monthly basis.
219 00:25:06.270 ⇒ 00:25:14.850 Uttam Kumaran: like I’m I think, between James Demalade, a wish like it’s up to us to come up with a solution. Here, I think we see the problem
220 00:25:15.840 ⇒ 00:25:18.300 Uttam Kumaran: long term. We should get it from
221 00:25:21.300 ⇒ 00:25:24.230 Uttam Kumaran: like, I’m not gonna bank on that happening anytime soon.
222 00:25:24.780 ⇒ 00:25:29.519 Uttam Kumaran: Short term, we need to figure out like, if they are updating this here.
223 00:25:29.870 ⇒ 00:25:37.200 Uttam Kumaran: we could, of course, write something that moves from sheet one to the other thing. But you guys know, if they fuck this up, then that what that functions
224 00:25:37.780 ⇒ 00:25:40.480 Uttam Kumaran: busted anyways, I think
225 00:25:41.720 ⇒ 00:25:49.649 Uttam Kumaran: probably just run something weekly with them that can get the updates. We get the updates they need. And we make the updates in the product mapping
226 00:25:49.850 ⇒ 00:25:52.790 Uttam Kumaran: until find a better solution.
227 00:25:53.840 ⇒ 00:25:55.739 Uttam Kumaran: That’s that would be my feedback.
228 00:25:58.900 ⇒ 00:26:10.540 James Freire: Yeah, cause I mean that that has to like whoever’s putting the I mean like, this is like, I mean, even where you know other places I’ve been at. I mean, this is still a thing. If people are like entering stuff in spreadsheets and emailing around to each other.
229 00:26:10.840 ⇒ 00:26:11.540 Robert Tseng: Yeah.
230 00:26:13.120 ⇒ 00:26:20.410 Uttam Kumaran: So I I think, like the most, the high, the highest touch thing we could do here is to just run a meeting every week to do this. So
231 00:26:21.160 ⇒ 00:26:24.369 Uttam Kumaran: I mean, like, I think, Steven, that that’s what I would suggest.
232 00:26:28.890 ⇒ 00:26:31.930 steven: Yeah, alright. So yeah, we’ll still running a meeting. Is that like
233 00:26:32.080 ⇒ 00:26:33.979 steven: touching base with the client as well.
234 00:26:35.070 ⇒ 00:26:41.160 Uttam Kumaran: I guess I’ll let you. I think I’ll probably push, I think, between you and Robert, and then probably have one engineer there to
235 00:26:41.850 ⇒ 00:26:46.389 Uttam Kumaran: to make sure that this this happens, but I don’t see an alternative short term right now.
236 00:26:46.680 ⇒ 00:26:49.129 steven: But this, this will solve this problem.
237 00:26:50.390 ⇒ 00:26:54.258 Robert Tseng: Yeah. Steven, are you? Are you free? 11 am. Et
238 00:26:54.610 ⇒ 00:26:56.920 steven: Let’s see, that’d be.
239 00:26:57.720 ⇒ 00:26:58.060 Robert Tseng: Hour.
240 00:26:58.060 ⇒ 00:27:02.400 steven: Oh, an hour. Yeah, it should be 8 pm. Or 8 Am.
241 00:27:02.400 ⇒ 00:27:02.829 Robert Tseng: I hate it.
242 00:27:04.760 ⇒ 00:27:08.780 Robert Tseng: Yeah, I might add you to that call then with them. But okay.
243 00:27:09.670 ⇒ 00:27:16.279 Uttam Kumaran: Yeah, Robert, I think we run this weekly. We’ll just do it, brute force, for now, until we get to the bottom of like where this is coming from.
244 00:27:16.410 ⇒ 00:27:22.380 Uttam Kumaran: If it’s not, then this is it. Cause this, this is like something at every company like.
245 00:27:22.970 ⇒ 00:27:24.470 Uttam Kumaran: Yeah, I know.
246 00:27:25.186 ⇒ 00:27:25.680 Robert Tseng: It starts.
247 00:27:25.680 ⇒ 00:27:33.880 Uttam Kumaran: It starts off as a Google sheet. It starts off with someone, yeah, making some fat finger. And then it fucks everything up. And then we basically have to guard against the fat finger.
248 00:27:34.360 ⇒ 00:27:48.819 Uttam Kumaran: Ideally, we should. We should do 2 or 3 of these meetings, and then whoever’s on the other line should get trained to like, make updates and update us with what they are or or provide us with that Async like, here are the changes we need
249 00:27:49.060 ⇒ 00:27:50.830 Uttam Kumaran: go make them in the spreadsheet
250 00:27:51.150 ⇒ 00:27:56.018 Uttam Kumaran: cause. Then we can even move this out of like a spreadsheet and do something else. But like, that’s this is the
251 00:27:57.380 ⇒ 00:28:03.830 Uttam Kumaran: yeah, this is it, unless you guys, unless James or Demo, do you guys see another option.
252 00:28:04.450 ⇒ 00:28:13.933 James Freire: I mean, you know, whatever. I don’t know what their data sources on that like. Are they getting it from? the wholesale company, I mean, does that have
253 00:28:14.250 ⇒ 00:28:14.900 Uttam Kumaran: Yeah.
254 00:28:14.900 ⇒ 00:28:18.540 James Freire: Api. They have access to, or you know.
255 00:28:19.510 ⇒ 00:28:19.980 Awaish Kumar: No worries.
256 00:28:19.980 ⇒ 00:28:22.109 Demilade Agboola: I think it’s right now.
257 00:28:22.360 ⇒ 00:28:32.520 Demilade Agboola: like, even with our Dbt process, we actually use the data, and it comes in and forms part of what we use for our product sales summary
258 00:28:32.650 ⇒ 00:28:51.690 Demilade Agboola: solution. So we’re also dependent on our Dbt production account being in there and not being put it off, but at least with Dbt, if it’s removed, break. So we know that, like something went wrong with our process, we can easily track and say, hey? Someone has fucked it up with.
259 00:28:52.625 ⇒ 00:29:00.259 Demilade Agboola: the issue that we’re having yesterday is because it’s a view. So he always wants to query the live data. And you know the whole.
260 00:29:00.260 ⇒ 00:29:00.630 Uttam Kumaran: Secure.
261 00:29:00.830 ⇒ 00:29:13.739 Demilade Agboola: Thing is like, even though you have access to the drive. I’m sorry, even though you have access to bigquery. If you don’t have access to the drive. It will not give you access to the data because you don’t have access in sheets, so that that.
262 00:29:13.740 ⇒ 00:29:14.270 Awaish Kumar: Bye.
263 00:29:14.270 ⇒ 00:29:17.929 Demilade Agboola: We don’t have a way of identifying when it goes wrong.
264 00:29:17.930 ⇒ 00:29:28.709 James Freire: Yeah, that’s what I’m saying. Like having that temporary data flow where it’s a you know, we replicate it into a table or something. So at least you can go by the last.
265 00:29:28.930 ⇒ 00:29:34.100 James Freire: You know the last set that’s in the the that tape that table.
266 00:29:34.510 ⇒ 00:29:41.109 James Freire: So even if it does, even if it is broken. You still have, like, you know, the data from like last month to go off of instead of like a broken connection.
267 00:29:41.940 ⇒ 00:29:42.600 Uttam Kumaran: Yeah.
268 00:29:42.600 ⇒ 00:29:43.160 Demilade Agboola: Yeah.
269 00:29:43.160 ⇒ 00:29:43.850 Uttam Kumaran: Definitely.
270 00:29:44.560 ⇒ 00:29:49.058 Demilade Agboola: Yeah, so we could just use the table. We could just materialize as a table instead.
271 00:29:50.900 ⇒ 00:29:55.400 Uttam Kumaran: I think we probably several steps here. Maybe we should just start a thread on this in slack
272 00:29:55.570 ⇒ 00:30:00.729 Uttam Kumaran: and hit with all of our ideas. I do think that very short term
273 00:30:01.040 ⇒ 00:30:04.370 Uttam Kumaran: just getting a meeting where we we get in the
274 00:30:04.710 ⇒ 00:30:09.049 Uttam Kumaran: the changes, and then we make the changes is best.
275 00:30:09.320 ⇒ 00:30:11.729 Uttam Kumaran: Certainly we should materialize this table.
276 00:30:13.800 ⇒ 00:30:21.760 Uttam Kumaran: and certainly we should have some downstream data quality checks because someone will put in an a string in a number field soon enough.
277 00:30:22.340 ⇒ 00:30:25.420 Uttam Kumaran: So we want to catch that.
278 00:30:34.970 ⇒ 00:30:39.219 Uttam Kumaran: Okay, cool. What else we got.
279 00:30:43.200 ⇒ 00:30:56.890 Robert Tseng: I think. Yeah. Seems like we’re wrapped. We’ve wrapped up some of the most of the modeling work that for the dashboards that we were playing catch up on. I know, James, I haven’t sent you kind of like a list of the new the other stuff that we wanted to add
280 00:30:58.310 ⇒ 00:31:00.919 Robert Tseng: But yeah, I mean, I think.
281 00:31:04.090 ⇒ 00:31:05.570 Robert Tseng: yeah, I bet I don’t.
282 00:31:05.930 ⇒ 00:31:13.139 Robert Tseng: I think the mixed panel stuff is like on my mind right now, and I have some follow ups with Akash to kinda keep. Keep that going. And
283 00:31:13.310 ⇒ 00:31:19.660 Robert Tseng: so yeah, I I think I I just, I’m not. I’m not that we’re not very organized on what else needs to be.
284 00:31:19.660 ⇒ 00:31:20.380 Uttam Kumaran: That’s fine!
285 00:31:20.380 ⇒ 00:31:20.840 Robert Tseng: Then that’s
286 00:31:20.840 ⇒ 00:31:26.690 Robert Tseng: that’s fine. So let’s get you out of this tableau stuff today. Let’s get you out of this spreadsheet stuff today.
287 00:31:27.380 ⇒ 00:31:37.640 Uttam Kumaran: I think between between you guys, Robert Steven and Akash would be amazing to have somewhat of a roadmap, even if it doesn’t make it in linear
288 00:31:37.790 ⇒ 00:31:42.720 Uttam Kumaran: for next week for this team to start crushing through.
289 00:31:43.171 ⇒ 00:31:48.509 Uttam Kumaran: I I definitely. I mean, we didn’t hear from Sahana, and I assure her updates are in slack. So
290 00:31:48.994 ⇒ 00:31:52.539 Uttam Kumaran: let’s just keep talking there, I think, for the team.
291 00:31:52.640 ⇒ 00:31:55.110 Uttam Kumaran: It would be amazing if we can
292 00:31:55.350 ⇒ 00:32:05.730 Uttam Kumaran: start if we could put in some things today and tomorrow to harden that Google Sheet and the process around it as well as James. Whatever you can do to give us some assurance. Going into the weekend on
293 00:32:06.200 ⇒ 00:32:07.790 Uttam Kumaran: tableau extracts.
294 00:32:07.790 ⇒ 00:32:09.780 James Freire: Yeah, that’s not a big. It’s not.
295 00:32:09.780 ⇒ 00:32:10.370 Uttam Kumaran: Yeah, I don’t.
296 00:32:10.370 ⇒ 00:32:10.900 Uttam Kumaran: But actually.
297 00:32:10.900 ⇒ 00:32:15.280 Uttam Kumaran: I feel like it’ll be pretty chilled. But as long as it gets alerted into there that way.
298 00:32:15.550 ⇒ 00:32:18.860 Uttam Kumaran: even if I’m online and I could go resolve, I can go do that.
299 00:32:18.980 ⇒ 00:32:23.070 Uttam Kumaran: And then, yeah, I think, Steven, the biggest thing coming into next week, would be just
300 00:32:23.380 ⇒ 00:32:28.761 Uttam Kumaran: being able to make sure that this team can can, can pick off things and and keep working.
301 00:32:29.770 ⇒ 00:32:51.200 Uttam Kumaran: I’m happy to also, once once we have that I do have several data modeling and and documentation asks that I can layer onto that roadmap as well. And some other some other data ingestion pieces so definitely in your conversation. If you guys can get to the marketing sources as well that we need to bring in, that would be very, very helpful.
302 00:32:53.610 ⇒ 00:32:54.330 Uttam Kumaran: great.
303 00:32:54.670 ⇒ 00:32:59.859 Uttam Kumaran: Okay, cool. And then, yeah, just before before you log off today, just please just send a
304 00:32:59.980 ⇒ 00:33:02.949 Uttam Kumaran: short update, or even as you’re going out through through the day.
305 00:33:05.880 ⇒ 00:33:06.959 steven: Sounds good. Yeah.
306 00:33:08.270 ⇒ 00:33:14.949 Uttam Kumaran: Alright. Thanks everyone. Thanks, James. It’s great to have you all pumped. I know you’re kind of in the fire right now. Things will get a little bit easier.
307 00:33:14.950 ⇒ 00:33:21.080 James Freire: No, I’m like I was like digging through all this stuff I could yesterday. And yeah, I’m starting to. I’m starting to get a layout of of the flow.
308 00:33:21.600 ⇒ 00:33:25.310 Uttam Kumaran: This is low hanging fruit. There’s some stickier items that we’ll get to, so.
309 00:33:25.310 ⇒ 00:33:26.810 James Freire: Yup, Nope, not a problem at all.
310 00:33:26.810 ⇒ 00:33:29.649 Uttam Kumaran: Cool. Okay. Alright. Thanks. Everyone.
311 00:33:30.580 ⇒ 00:33:31.130 steven: Bye.