Meeting Title: MatterMore x Brainforge | Standup Date: 2025-06-05 Meeting participants: Awaish Kumar, Trevor’s Notetaker (Otter.ai), Mathew’s Notetaker (Otter.ai), Fireflies.ai Notetaker Awaish, Mathew, Uttam Kumaran, Amber Lin
WEBVTT
1 00:03:33.980 ⇒ 00:03:35.110 Uttam Kumaran: Hello!
2 00:03:54.850 ⇒ 00:03:55.960 Mathew: What’s up? Team.
3 00:03:56.300 ⇒ 00:03:57.080 Uttam Kumaran: Hey!
4 00:03:57.920 ⇒ 00:03:58.820 Amber Lin: Hi! There!
5 00:04:08.700 ⇒ 00:04:11.390 Mathew: You want to take it from here to the table, Bob?
6 00:04:11.500 ⇒ 00:04:12.160 Mathew: Alright!
7 00:04:29.670 ⇒ 00:04:31.509 Mathew: Let’s dance. How do you see him?
8 00:04:32.370 ⇒ 00:04:33.419 Mathew: We waiting for anybody.
9 00:04:33.420 ⇒ 00:04:33.970 Uttam Kumaran: Cool.
10 00:04:34.430 ⇒ 00:04:35.180 Mathew: But hot.
11 00:04:35.550 ⇒ 00:04:36.260 Uttam Kumaran: Us.
12 00:04:37.390 ⇒ 00:04:40.379 Amber Lin: No, it’s gonna be us. We’re all here.
13 00:04:41.720 ⇒ 00:04:42.420 Mathew: Wonderful.
14 00:04:44.350 ⇒ 00:04:47.119 Mathew: Well, how do you guys want to do this? Do you want me to lead?
15 00:04:47.430 ⇒ 00:04:52.039 Mathew: Well, stop last. Guy’s ear pierced. Okay, great.
16 00:04:52.440 ⇒ 00:04:53.500 Uttam Kumaran: Oh, nice.
17 00:04:55.110 ⇒ 00:04:57.270 Uttam Kumaran: Yeah, amber. How do you want to run this.
18 00:04:58.640 ⇒ 00:05:03.384 Amber Lin: So I know, Matthew, you sent a
19 00:05:04.240 ⇒ 00:05:19.420 Amber Lin: new document that is really awesome. It outlines the metrics. I prepared a document of based on previous requirements what we have done, and I consolidated the graphs a little bit. So I say, we could
20 00:05:20.570 ⇒ 00:05:23.320 Amber Lin: start with what we’ve done so far.
21 00:05:25.210 ⇒ 00:05:33.910 Amber Lin: we can talk about what we identified, we need or is missing. And then we can also look at this new document, and then let’s talk about
22 00:05:34.190 ⇒ 00:05:41.290 Amber Lin: how we want to move forward any retros or things we should improve on how to work together.
23 00:05:41.440 ⇒ 00:05:45.119 Amber Lin: and then we can talk about next steps. How’s that.
24 00:05:46.800 ⇒ 00:05:47.889 Mathew: Well, let’s do it.
25 00:05:48.250 ⇒ 00:05:49.080 Mathew: Awesome.
26 00:05:49.320 ⇒ 00:05:56.000 Amber Lin: Okay, so I’m gonna share my screen. I’ll run you guys really quick through what we have.
27 00:05:59.030 ⇒ 00:06:01.020 Amber Lin: So right here.
28 00:06:04.630 ⇒ 00:06:17.820 Amber Lin: right here. So I made a quick checklist just to make sure that we’re hitting everything. So this is the checklist I made from the deck that you guys shared. And we have, because there’s 2 documents floating around. I also made one, for
29 00:06:17.970 ⇒ 00:06:21.760 Amber Lin: you know this document that you also shared.
30 00:06:22.270 ⇒ 00:06:37.410 Amber Lin: And then I was just going through making sure that we have each one, and mostly I’m going to base it off of this this slide deck, because that’s the most comprehensive one, and then we’ll make sure that we don’t miss anything over there over there. And so, looking at that.
31 00:06:37.720 ⇒ 00:06:41.300 Mathew: Franklin, it’s it’s referencing the slide deck. So they’re the same.
32 00:06:41.300 ⇒ 00:06:42.460 Amber Lin: Yeah.
33 00:06:43.400 ⇒ 00:06:48.979 Amber Lin: sounds good. They’re just organized a little bit differently, I think this one’s a better one to base off of
34 00:06:49.480 ⇒ 00:07:04.470 Amber Lin: and so I listed each one that had a visualization. So focusing on 2 main segments here. So we have work patterns, hybrid productivity. And that’s essentially slide 8 through 24.
35 00:07:05.060 ⇒ 00:07:29.470 Amber Lin: I checked up all of these, the ones with the approximation mark is that we have all the modeling in place. It’s just we need to adjust the visualizations a little bit. And I think, overall, it’s just these 2 that needs I want to check in with Annie about the visualizations for that. Otherwise, I think we’re pretty good. So just to run you guys through about what we have.
36 00:07:29.840 ⇒ 00:07:43.299 Amber Lin: And I also, I took a look at the document that Matthew you sent, and I believe we pretty much have everything that we need for phase one in python. But we need. We should talk about how we’re gonna do that and see one do that in power. Bi.
37 00:07:43.660 ⇒ 00:07:52.690 Amber Lin: And so 1st of all, let’s go look at. So I have the graph from
38 00:07:57.870 ⇒ 00:08:06.130 Amber Lin: the graph from the slide that you shared under here. So 1st of all, we have meetings
39 00:08:06.760 ⇒ 00:08:10.550 Amber Lin: by day a week. We have emails by day a week.
40 00:08:11.300 ⇒ 00:08:27.289 Amber Lin: and I know here what we really wanted to do is to look at the different departments. I know the visualizations looks a little bit different, but overall. We can present it in this format. But ultimately what we did is how we enabled us to look at the different departments
41 00:08:27.410 ⇒ 00:08:30.129 Amber Lin: over the different days. A week day of week.
42 00:08:30.990 ⇒ 00:08:39.589 Amber Lin: And this is just a slightly different format. So we can see all the departments in one graph. Similarly for emails
43 00:08:40.130 ⇒ 00:08:41.299 Amber Lin: right over here.
44 00:08:42.010 ⇒ 00:08:45.930 Amber Lin: So that’s the first, st the 1st few ones
45 00:08:46.120 ⇒ 00:08:48.540 Amber Lin: that we weren’t able to look at.
46 00:08:48.820 ⇒ 00:08:55.779 Mathew: Yeah, I think I added, feedback on this one is that it was that I’m not sure about how valuable a heat map would be as a visualization for different departments, because.
47 00:08:58.870 ⇒ 00:08:59.810 Mathew: Like.
48 00:09:00.150 ⇒ 00:09:14.459 Mathew: if we’re gonna compare things to each other. I feel like line charts that show like the relative like that. Yeah, would be more valuable for like a team view, just because then, like with this, the height doesn’t really mean anything. It’s just like.
49 00:09:15.576 ⇒ 00:09:19.580 Mathew: it’s just the color that we’re basing it off of, and I think, for comparison.
50 00:09:19.730 ⇒ 00:09:22.650 Mathew: speak like that. That was one that was one piece of feedback.
51 00:09:22.650 ⇒ 00:09:26.719 Amber Lin: Yeah, awesome. Thank you for that. And so
52 00:09:27.180 ⇒ 00:09:33.680 Amber Lin: then we can just apply what we did here. This is by hour of day, pretty similar by department.
53 00:09:34.225 ⇒ 00:09:50.009 Amber Lin: Which is kind of, I believe, what your goal here is as well. That’s this one is singularly overall. And this is how we broke it down by department and based on your feedback. We’ll also do the by day, week in a similar fashion.
54 00:09:51.000 ⇒ 00:09:59.710 Amber Lin: So that’s number 12 and let’s see.
55 00:09:59.840 ⇒ 00:10:02.829 Amber Lin: So that’s email sent by R and J,
56 00:10:03.270 ⇒ 00:10:10.820 Amber Lin: I, that’s pretty much, very similar to this one. So I wanted to show how we did a in office versus remote comparison.
57 00:10:11.800 ⇒ 00:10:14.399 Amber Lin: and that’s something else that we could do as well.
58 00:10:14.820 ⇒ 00:10:15.510 Mathew: Okay.
59 00:10:16.200 ⇒ 00:10:17.040 Amber Lin: Yeah, cool.
60 00:10:18.532 ⇒ 00:10:24.010 Amber Lin: Next one is, I put a approximate
61 00:10:24.170 ⇒ 00:10:32.380 Amber Lin: here. So right now we have. We don’t have a stacked bar chart, but that’s definitely something we can do. We have all the
62 00:10:32.540 ⇒ 00:10:40.079 Amber Lin: data in place, and we did it by each tool. So that’s the overall. And then this is based off of each tool.
63 00:10:40.600 ⇒ 00:10:42.030 Amber Lin: I would note.
64 00:10:42.030 ⇒ 00:10:45.739 Mathew: And so what is what is this one showing? I’m not sure what this screen is.
65 00:10:46.410 ⇒ 00:10:47.840 Mathew: what this graph is for
66 00:11:00.380 ⇒ 00:11:01.230 Mathew: amber.
67 00:11:01.980 ⇒ 00:11:06.629 Amber Lin: Huh? Oh, I thought you were looking at the screen. This is from the slide
68 00:11:08.240 ⇒ 00:11:11.680 Amber Lin: This is from the slide deck. Can you share my screen. Am I sharing the right screen.
69 00:11:11.680 ⇒ 00:11:13.900 Mathew: Yeah, I’m looking. I’m looking. I’m seeing the bottom of it.
70 00:11:14.070 ⇒ 00:11:16.630 Amber Lin: Yeah, so this is
71 00:11:17.050 ⇒ 00:11:25.820 Amber Lin: hybrid, typical workday productivity. I. So so I think this is comparing remote versus in office.
72 00:11:26.120 ⇒ 00:11:37.950 Amber Lin: How much people are spending on these tools, and then it’s broken down by a stack bar chart of how much each specific tool takes up.
73 00:11:38.320 ⇒ 00:11:46.370 Amber Lin: So right now, we have remote versus in office, and then I believe what you want is to slice this
74 00:11:46.550 ⇒ 00:11:53.999 Amber Lin: and break it down by okay, this is how much people spend in outlook. This is how much they spend
75 00:11:54.300 ⇒ 00:12:08.069 Amber Lin: in messages, and then how much they spent in other activities. And just here is how is per tool. But I think we need to convert it into a stack bar chart.
76 00:12:08.360 ⇒ 00:12:09.390 Mathew: Got it. Okay.
77 00:12:09.750 ⇒ 00:12:10.340 Amber Lin: Yeah.
78 00:12:11.780 ⇒ 00:12:15.034 Mathew: Okay? And then I think we should be really explicit about
79 00:12:15.720 ⇒ 00:12:21.040 Mathew: yeah, calling it not just productivity. I would call it just like typical workday.
80 00:12:21.850 ⇒ 00:12:26.249 Mathew: like this, the overalls for this like 15 would be like.
81 00:12:26.490 ⇒ 00:12:34.860 Mathew: Don’t yeah. I know in ours we called a typical workday productivity. But just as we’ve gone deeper into all of this, we have to be very mindful of of how we name something.
82 00:12:36.370 ⇒ 00:12:39.900 Mathew: Productivity is a loaded term. So we should just call this like tool usage.
83 00:12:40.180 ⇒ 00:12:40.510 Amber Lin: Go ahead!
84 00:12:41.040 ⇒ 00:12:47.559 Mathew: Yeah, like, time spent in in those tools. Before we start like calling things like, productive or not. But yeah, yeah.
85 00:12:47.560 ⇒ 00:12:49.869 Amber Lin: Yeah, awesome. So I’ll note.
86 00:12:49.870 ⇒ 00:12:57.753 Uttam Kumaran: I think we just need the like, if we could just have a list of those or something, because we’re just gonna cop, we’re just basically copying it right off the slides.
87 00:12:58.460 ⇒ 00:13:00.029 Uttam Kumaran: So I don’t know.
88 00:13:00.030 ⇒ 00:13:10.689 Mathew: Okay, yeah, that’s the what we shared in in that other. Doc is really like the root of what we want to base. That that’s like the most refined thinking from our side.
89 00:13:11.990 ⇒ 00:13:12.870 Amber Lin: Awesome.
90 00:13:12.870 ⇒ 00:13:14.329 Mathew: That we want to base everything off.
91 00:13:15.980 ⇒ 00:13:16.950 Amber Lin: Sounds good.
92 00:13:17.980 ⇒ 00:13:21.940 Amber Lin: And then so right here.
93 00:13:22.310 ⇒ 00:13:29.840 Amber Lin: what we wanted. So the next one is productivity tasks over the week.
94 00:13:30.580 ⇒ 00:13:40.779 Amber Lin: And you can see here we have the different overall productivity over the week, and we also went ahead and broke down by tools.
95 00:13:40.780 ⇒ 00:13:43.109 Mathew: This one’s just showing. Yeah, this is just showing like.
96 00:13:43.920 ⇒ 00:13:51.389 Mathew: basically aggregated tool usage. So not messages. Not that, not email. It’s just time spent in Microsoft Onedrive working on.
97 00:13:51.390 ⇒ 00:13:51.980 Amber Lin: Yeah.
98 00:13:52.440 ⇒ 00:13:54.179 Mathew: Tools. That’s that’s all. This is.
99 00:13:54.540 ⇒ 00:13:56.740 Amber Lin: Hmm, yeah.
100 00:13:57.260 ⇒ 00:14:03.329 Amber Lin: And then I remember here, we will also want office versus remote. That’s something that we’ll we need to.
101 00:14:03.330 ⇒ 00:14:17.820 Mathew: Yeah, office versus remote is gonna end up. Just I know on even in our doc, we call it like an overlay, or it’s just gonna be a filter, or like a different different form of segmenting. All of all of this, like like each of these fundamental points.
102 00:14:18.480 ⇒ 00:14:20.370 Amber Lin: Okay, sounds good.
103 00:14:21.730 ⇒ 00:14:22.630 Amber Lin: Yeah, yeah.
104 00:14:23.790 ⇒ 00:14:33.770 Amber Lin: I think your doc makes it very clear, and it’s a lot easier to understand now. So I would love to just speak through this. And I want to talk about the doc that you brought up, because I think it’s it’s going to be very helpful.
105 00:14:35.730 ⇒ 00:14:41.810 Amber Lin: So next one, this is something that we
106 00:14:41.920 ⇒ 00:14:49.470 Amber Lin: need to visualize. We don’t have it by, because right here, this graph shows it as a typical day.
107 00:14:50.138 ⇒ 00:14:53.900 Amber Lin: Up here in the title. It wants it by
108 00:14:54.360 ⇒ 00:14:58.740 Amber Lin: day a week, which we were able to do down here. But
109 00:14:59.296 ⇒ 00:15:06.829 Amber Lin: here is where we’re trying to figure out specifically in the day for a single person
110 00:15:07.060 ⇒ 00:15:08.949 Amber Lin: how it’s distributed.
111 00:15:09.510 ⇒ 00:15:18.600 Amber Lin: and we have the tables and the modeling we haven’t been able. We haven’t done the by day breakdowns yet.
112 00:15:19.760 ⇒ 00:15:20.120 Mathew: Yes.
113 00:15:20.120 ⇒ 00:15:21.339 Amber Lin: Something to know.
114 00:15:21.340 ⇒ 00:15:26.769 Mathew: I don’t even, I think, until we have all the other ones hammered out that we shouldn’t worry about focus time just yet.
115 00:15:27.260 ⇒ 00:15:34.730 Amber Lin: Okay, okay, sounds good. We were focusing on trying to figure out
116 00:15:35.020 ⇒ 00:15:46.720 Amber Lin: everything on the slides. And I think focus time took up a few few of these. So I would just show you what we have. And if you want us to refocus on certain stuff, we can totally do that.
117 00:15:47.020 ⇒ 00:15:51.509 Amber Lin: And so the next one is also about focus time. This is about day of week.
118 00:15:51.630 ⇒ 00:15:55.560 Amber Lin: So we have day a week. We have a distribution
119 00:15:55.850 ⇒ 00:16:01.360 Amber Lin: over the day of week as well. If you want a further breakdown, and then
120 00:16:02.211 ⇒ 00:16:06.630 Amber Lin: this one is also focused time. And that’s by department.
121 00:16:06.730 ⇒ 00:16:14.259 Amber Lin: So right now, we also need to make it into a stack bar chart. But we have activities individually.
122 00:16:16.780 ⇒ 00:16:23.651 Mathew: Yeah. Same same comment. I think if we could get the fundamental foundational
123 00:16:24.590 ⇒ 00:16:32.860 Mathew: stuff down for just activity before going another layer down and then dissecting and saying, Well, this is how interrupted they are. This is how much like.
124 00:16:33.570 ⇒ 00:16:40.260 Mathew: Space they have, I think. Let’s let’s like lock in one. Those foundational
125 00:16:40.510 ⇒ 00:16:43.690 Mathew: charts, graphs, etc. Make sure we have them.
126 00:16:44.121 ⇒ 00:16:50.560 Amber Lin: Let’s make sure that the way that you’ve built it out is as modular and flexible and reusable as possible.
127 00:16:50.690 ⇒ 00:16:55.830 Mathew: Because there’s gonna even if even just on those, I foresee
128 00:16:56.330 ⇒ 00:17:04.119 Mathew: a ton of different filtering segmenting questions that the client is going to be asking. They’re gonna say, well, what about this team
129 00:17:04.230 ⇒ 00:17:14.700 Mathew: that works this way in this location compared to that team this way, that location before introducing focus time which can make things. It’s I, wanna I’m.
130 00:17:14.700 ⇒ 00:17:15.746 Amber Lin: Second, layer,
131 00:17:16.270 ⇒ 00:17:23.650 Mathew: Yeah, yeah, yeah, exactly. I want to start with baby steps with them. And so I think then our internal baby steps would be like.
132 00:17:24.750 ⇒ 00:17:29.020 Mathew: how far? Where are the graphs of that? You’ve you know that? How are the graphs.
133 00:17:29.210 ⇒ 00:17:32.910 Mathew: and how are they corresponding to what we put together? How is it built?
134 00:17:33.100 ⇒ 00:17:44.377 Mathew: And if it’s so that it’s it’s built in a way that’s easy and fast for Trevor. And you guys to work to work with. So that when they ask for things and they ask for changes, we’re we’re we’re not like.
135 00:17:45.060 ⇒ 00:17:46.440 Mathew: we’re not handicapped.
136 00:17:47.380 ⇒ 00:17:59.610 Amber Lin: Awesome. Then I think we’ll skip because a lot of these are further metrics, because it’s further down this list. I think all of the ones. The most basic ones are here already addressed.
137 00:17:59.960 ⇒ 00:18:01.206 Mathew: Yeah. And and we all
138 00:18:01.970 ⇒ 00:18:12.619 Mathew: like, right after it’s like, I call it, phase one plus like let’s get. We’ll get phase one down, and then I’ll just continue to stay ahead of like, what’s the next most important thing on for these for slides like these.
139 00:18:13.610 ⇒ 00:18:32.860 Amber Lin: Okay, awesome. And I know that through our communication we kind of had other requests here and there. So I just wanted to show that Annie was able to do those so sync versus Async, and then some correlations here and there. But I know that’s not the focus for now. So this is just the.
140 00:18:32.860 ⇒ 00:18:34.920 Mathew: First, st what are you searching?
141 00:18:35.370 ⇒ 00:18:37.644 Mathew: Okay? Well, I’m talking. Okay.
142 00:18:38.626 ⇒ 00:18:48.099 Mathew: Sync, I get excited. Sync versus Async is. That’s actually, that’s that’s actually, gonna be important to relative relevant to what we’re talking about, and you’ll see.
143 00:18:48.945 ⇒ 00:18:49.790 Amber Lin: Awesome.
144 00:18:50.990 ⇒ 00:18:51.770 Amber Lin: Okay?
145 00:18:51.970 ⇒ 00:18:53.140 Amber Lin: So that’s
146 00:18:53.830 ⇒ 00:19:02.820 Amber Lin: so just want to show you guys, what we did based on our last conversation and then also some bit of correlations based on what we have already.
147 00:19:02.950 ⇒ 00:19:09.310 Amber Lin: And I think at this point I can hand it over. Matthew, if you want to talk about
148 00:19:09.460 ⇒ 00:19:11.140 Amber Lin: this talk.
149 00:19:11.730 ⇒ 00:19:17.560 Amber Lin: And we can talk about what we need, what we want to focus on.
150 00:19:18.190 ⇒ 00:19:21.640 Amber Lin: And then after this, we can talk about how we want to move forward.
151 00:19:22.670 ⇒ 00:19:23.330 Amber Lin: Cool.
152 00:19:26.720 ⇒ 00:19:34.069 Mathew: Yeah, I mean, I shared this Doc with the intention of of grounding you guys in our thinking, with the hope that you could then take it.
153 00:19:34.580 ⇒ 00:19:40.510 Mathew: digest it and then ultimately confirm where we’re at with
154 00:19:41.570 ⇒ 00:19:44.860 Mathew: like how we feel. We’re at with
155 00:19:45.090 ⇒ 00:19:50.115 Mathew: phase, you know the the 6 core charts. But then how we’re building it. So
156 00:19:50.570 ⇒ 00:19:54.069 Mathew: I think the implementation and I could share my screen again.
157 00:19:56.230 ⇒ 00:20:02.690 Mathew: I think like confirming like, how you guys are implementing it, and if it’s some, if it’s along these lines or not, would be
158 00:20:03.250 ⇒ 00:20:05.950 Mathew: useful in understanding what that delta is.
159 00:20:07.700 ⇒ 00:20:19.350 Mathew: But yeah, these, this is. And then, yeah, like, that’s, that’s that’s where I would start is understanding like, how have you guys been implementing it? Like is all just like stuff on Annie’s workbook, or is it built out in a way that’s like, you know.
160 00:20:20.240 ⇒ 00:20:21.760 Mathew: like this, essentially.
161 00:20:22.520 ⇒ 00:20:23.470 Uttam Kumaran: Yeah. So let maybe let’s.
162 00:20:24.510 ⇒ 00:20:48.670 Uttam Kumaran: Yeah, maybe let’s just go through. Let yeah, let’s just go through this list. So we sort of did a big review yesterday. And maybe I’ll sort of just walk through like what the architecture is right now. So one is, we’ve we don’t have live data, of course. So we’ve generated synthetic data based on like what we’re sort of like, fairly confident we’ll get from the clients on top of that. We generate Dbt models.
163 00:20:48.960 ⇒ 00:21:04.309 Uttam Kumaran: These are this is the sequel that sits and converts raw data into a set of mart tables. Annie is then using those mart tables, doing like light combinations and logic and then creating those charts.
164 00:21:05.460 ⇒ 00:21:08.759 Uttam Kumaran: So as of now, the reason we’ve sort of
165 00:21:09.490 ⇒ 00:21:14.160 Uttam Kumaran: right now we do have sequel, for like core metric processing.
166 00:21:15.147 ⇒ 00:21:36.302 Uttam Kumaran: and so we’ll be. It’s a constant thing between analysts and analytics engineering as the analysts have to do, custom work to produce charts. That business logic then moves back into equal and then into the Dbt model. So that process will always happen. But like, that’s where we are right now in power. Bi
167 00:21:36.840 ⇒ 00:21:37.620 Uttam Kumaran: the
168 00:21:37.950 ⇒ 00:21:50.540 Uttam Kumaran: that’s that’s basically power bi is just like tableau or like looker. Basically, you’re gonna have a set of dimensions and metrics that you can cut and filter in any way you want. Those will sit on top of the Dbt mart models.
169 00:21:50.942 ⇒ 00:21:56.029 Uttam Kumaran: So we’re kind of constrained right now by the fact that we just don’t have power bi to
170 00:21:56.130 ⇒ 00:22:10.570 Uttam Kumaran: to sort of set that up. So right now, it’s really custom like it’s it’s not self service we since we’re doing this all in python. So the 1st sort of blocker that I called out to Amber is that we really just need to set that up or have a path forward on
171 00:22:10.800 ⇒ 00:22:14.859 Uttam Kumaran: being able to set those Dbt marts up in power. Bi.
172 00:22:15.434 ⇒ 00:22:20.779 Uttam Kumaran: because then we can, you can basically pick dimensions and metrics, and filter pretty pretty easily.
173 00:22:21.290 ⇒ 00:22:25.039 Mathew: Cool. Thank you that. So it sounds like you guys are doing this.
174 00:22:25.510 ⇒ 00:22:26.220 Uttam Kumaran: Yes.
175 00:22:26.730 ⇒ 00:22:32.031 Mathew: Okay, to what degree do you think that we have them? Against these?
176 00:22:32.780 ⇒ 00:22:36.749 Mathew: like, obviously like this? This is, these are the atomic metrics
177 00:22:37.170 ⇒ 00:22:44.350 Mathew: that Annie and he’s working from these are just roll ups of these. That’s that was like the purpose of what this was trying to demonstrate.
178 00:22:45.052 ⇒ 00:22:58.359 Mathew: And then, obviously, these are just these are the different filters, and these are the different time grains and filters that are coming in. So I guess that’s that’s my, that’s my, you know, 1 million dollar question, which is like.
179 00:22:58.690 ⇒ 00:23:02.330 Mathew: to what degree is this set up.
180 00:23:02.820 ⇒ 00:23:05.100 Amber Lin: Yeah, like, yeah, I mean, do you trained?
181 00:23:05.100 ⇒ 00:23:10.360 Uttam Kumaran: Well, yeah, I’ll I’ll keep going. So do you have the do you have like the Dbt
182 00:23:12.230 ⇒ 00:23:16.299 Uttam Kumaran: like re, like repo? Because we’ve been making sort of the
183 00:23:17.170 ⇒ 00:23:21.149 Uttam Kumaran: we’ve been making some of the changes there, like we have
184 00:23:21.550 ⇒ 00:23:27.689 Uttam Kumaran: both some of the Dbt models there, and some of the stuff is in python. So
185 00:23:28.030 ⇒ 00:23:32.469 Uttam Kumaran: like right now, because we don’t have power, bi, we sort of. And
186 00:23:32.760 ⇒ 00:23:46.149 Uttam Kumaran: the other kind of thing is because we just didn’t have like really clear metrics, we’re sort of developing it and then moving that business logic back into Dbt, so now that we sort of have these metrics, we should be able to develop all this
187 00:23:47.160 ⇒ 00:23:52.639 Uttam Kumaran: in Dbt, but this is sort of what Dbt allows is we’ll create tables with these metrics and dimensions.
188 00:23:53.010 ⇒ 00:23:54.019 Mathew: Yeah, that’s great.
189 00:23:55.750 ⇒ 00:24:03.869 Mathew: What do you think the time like, how far? What’s the Delta in terms of like being able to get to that point like is that something we could have done ready by next week, for instance.
190 00:24:04.880 ⇒ 00:24:11.659 Uttam Kumaran: But it depends like what what it is we need like. Like, if if we have a spreadsheet of
191 00:24:11.920 ⇒ 00:24:12.939 Uttam Kumaran: sorry, go ahead.
192 00:24:13.590 ⇒ 00:24:18.389 Mathew: He, where? Where are we? Let me go here.
193 00:24:18.500 ⇒ 00:24:21.860 Mathew: because that’s this is just that basically like
194 00:24:22.290 ⇒ 00:24:25.749 Mathew: up to here for phase one, we’ll call it.
195 00:24:28.640 ⇒ 00:24:33.149 Uttam Kumaran: Yeah, so we can’t. So we we certainly need power bi. So we can’t do anything until.
196 00:24:33.870 ⇒ 00:24:40.200 Mathew: Sorry, not even in power. Bi, it’s from Dbt and like sequel. And Dbt
197 00:24:40.960 ⇒ 00:24:48.019 Mathew: understood that power bi is a blocker, and I’m I’m saying, like to be able to have the like. The the model set up.
198 00:24:49.130 ⇒ 00:24:53.730 Uttam Kumaran: Yeah. Wish. Do you know? Like, if if this is already ready, because I feel like this is already ready.
199 00:24:54.707 ⇒ 00:25:01.850 Awaish Kumar: Yeah, like, we already write SQL, to get all of these things. So you can, you can say, like
200 00:25:02.020 ⇒ 00:25:07.909 Awaish Kumar: most of it is already there in a, in a table. We just need to connect it with power. Bi.
201 00:25:08.150 ⇒ 00:25:15.220 Uttam Kumaran: Yeah. So right now, it’s all set up as views in bigquery. So all of these, everything we’ve used to produce
202 00:25:15.440 ⇒ 00:25:16.850 Uttam Kumaran: those graphs.
203 00:25:17.030 ⇒ 00:25:20.479 Uttam Kumaran: If it’s a metric, it’s already ready as a SQL. Query.
204 00:25:21.990 ⇒ 00:25:24.259 Mathew: Well, wait. So Utam, are you?
205 00:25:24.560 ⇒ 00:25:30.230 Mathew: I’m looking at the latest Dbt repo, have you? Have you guys added anything to it yet? Because I haven’t seen anything. I don’t see anything.
206 00:25:30.230 ⇒ 00:25:34.170 Uttam Kumaran: Well, right. Well, right now it’s all like sitting as views in.
207 00:25:34.770 ⇒ 00:25:35.310 Mathew: Cool.
208 00:25:35.310 ⇒ 00:25:45.319 Uttam Kumaran: It’s all sitting as views in bigquery. So if you, if in bigquery, if you go to those tables and I don’t have it up in front of me. The logic is sitting there, so all we would do is convert those to
209 00:25:45.540 ⇒ 00:25:52.040 Uttam Kumaran: Dbt models. The reason we haven’t yet, I believe, is just because we were. We’ve been making changes so frequently.
210 00:25:52.713 ⇒ 00:25:55.189 Uttam Kumaran: That we’ve just been recommitting it as
211 00:25:55.440 ⇒ 00:25:59.609 Uttam Kumaran: views. But I will have them. I’ll have Luke today.
212 00:25:59.890 ⇒ 00:26:02.809 Uttam Kumaran: Submit A. Pr. I’ll send them a note right now.
213 00:26:02.940 ⇒ 00:26:15.960 Mathew: No rush on that at all, and I mean it makes sense. If you’re iterating quickly to to not have to like, make a pull request every time. But yeah, if you’re ready for that, that’s great, and I just want to confirm. I don’t think you need anything else for me like I’ve set up the Dvt project. I think it’s ready for things to be added to it right.
214 00:26:16.200 ⇒ 00:26:17.180 Uttam Kumaran: That’s correct.
215 00:26:17.790 ⇒ 00:26:22.420 Mathew: Okay, cool. And then, like, the biggest thing is just to keep it like, keep everything namespaced by
216 00:26:22.815 ⇒ 00:26:30.800 Mathew: by data set right? So that, like, you know, all these, all these are like just applying to the synthetic, or whatever the data sets called in bigquery.
217 00:26:31.230 ⇒ 00:26:33.843 Uttam Kumaran: Exactly way so ideally.
218 00:26:35.245 ⇒ 00:26:40.610 Uttam Kumaran: Matthew, what you’ll get in power Bi is you’ll get each of these as dimensions and metrics.
219 00:26:40.770 ⇒ 00:26:45.119 Uttam Kumaran: and that, though, like the gap between Dbt. And that is like
220 00:26:45.390 ⇒ 00:26:57.039 Uttam Kumaran: dbt, I would say, dbt and SQL. And bigquery is considered just like tables. Right? So you just have. There’ll be final tables. The final tables will have columns like timeframe
221 00:26:57.469 ⇒ 00:27:12.639 Uttam Kumaran: team function role right in power Bi, the the timestamp column will turn into a day of week hour day. But it’s all basically pulling from that one column, and then each of those other columns will get mapped, and then you’ll have metrics which are
222 00:27:12.810 ⇒ 00:27:17.730 Uttam Kumaran: whatever any sort of aggregates, like hours or 3 swipe sort of things like.
223 00:27:18.240 ⇒ 00:27:29.570 Mathew: And yeah, I think that like the one thing that we wanted to emphasize, which I think Matthew has, is that the more we want to try to do as much as possible and get it as far down the line as possible in sequel, so that we can keep those.
224 00:27:29.930 ⇒ 00:27:59.460 Mathew: you know, like the idea of being that, like you said that we have all of these unit metrics that we can then reuse from those whether they’re says views, or like, I think if we’re running a ton of queries based on them, it makes sense to actually make them tables, so that, you know it’s not too expensive. So then run all our queries off of them. But yeah, that that like python basically should just be like the pulling everything into like a 1 like into a single table, but that anything that we think we’re going to reuse, which I think is a lot of stuff doing in SQL. Just allows us to.
225 00:27:59.570 ⇒ 00:28:03.310 Mathew: you know you put in Dbt. And and have it.
226 00:28:03.310 ⇒ 00:28:11.960 Uttam Kumaran: Yeah, like, the only reason it’s living in python is because we’ve been changing it so frequently and but like the the handoff between the analyst crew. And
227 00:28:12.210 ⇒ 00:28:22.270 Uttam Kumaran: Dbt, this is like just something that happens in data teams where the analysts will move quicker to produce. And then that logic gets shoved back into like
228 00:28:23.150 ⇒ 00:28:42.140 Uttam Kumaran: through through everything, back into things. So that that process I’m I’m sort of not so worried about. I think right now. There’s just a lot in python, because we’ve been moving sort of like we’ve been iterating. So one feedback I gave to the team was to 1st that we should just move all as much of what we know is cemented into the repo.
229 00:28:43.960 ⇒ 00:28:46.160 Uttam Kumaran: And then, yeah, I think the more like
230 00:28:46.460 ⇒ 00:28:53.030 Uttam Kumaran: drag and drop stuff will just be enabled a lot easier, and and y’all will get the chance to explore it. Once we have
231 00:28:54.070 ⇒ 00:28:55.400 Uttam Kumaran: the bi tool set up.
232 00:28:55.670 ⇒ 00:29:10.560 Mathew: Well, okay. So then, how far do you think we are from? What? What’s needed from your side? Obviously, we have to. We have to clear the power bi blocker. Oh, yeah, and sorry it was on just one last thing. Can you just make sure that I’m tagged in that pr to ask.
233 00:29:10.690 ⇒ 00:29:11.090 Uttam Kumaran: Yeah.
234 00:29:11.090 ⇒ 00:29:11.630 Mathew: That.
235 00:29:11.630 ⇒ 00:29:14.190 Uttam Kumaran: I’ll tag you as a code owner on all that stuff. Yeah.
236 00:29:14.190 ⇒ 00:29:20.720 Mathew: Yeah, just to just to make sure that the way that is set up like structurally matches, the way that I’m having everything like triggered, and all that logic.
237 00:29:20.720 ⇒ 00:29:21.250 Uttam Kumaran: Okay.
238 00:29:21.570 ⇒ 00:29:22.140 Uttam Kumaran: Sure.
239 00:29:22.400 ⇒ 00:29:23.360 Mathew: What how far.
240 00:29:23.360 ⇒ 00:29:23.880 Uttam Kumaran: Yeah.
241 00:29:25.011 ⇒ 00:29:31.330 Uttam Kumaran: yeah. So so one is, I think, on our side. And a wish we should take this. And
242 00:29:31.500 ⇒ 00:29:33.379 Uttam Kumaran: we should basically turn use our
243 00:29:33.540 ⇒ 00:29:37.290 Uttam Kumaran: typical documentation to track these metrics and the definitions.
244 00:29:37.743 ⇒ 00:29:42.150 Uttam Kumaran: That way, we can make sure we align this to what we’re writing in sequel.
245 00:29:42.682 ⇒ 00:29:50.530 Uttam Kumaran: For example, like team like, what? What source is that from like? What is the definition? Is their business logic?
246 00:29:50.670 ⇒ 00:29:53.510 Uttam Kumaran: Like we, we actually just need like
247 00:29:54.210 ⇒ 00:29:57.709 Uttam Kumaran: metric by metric by metric, and like what the definition is, there’s some.
248 00:29:57.910 ⇒ 00:30:01.219 Uttam Kumaran: There’s something that exists now, so we’ll make sure that’s mapped.
249 00:30:01.510 ⇒ 00:30:06.250 Uttam Kumaran: But like that’s the sort of back and forth, which is like, Okay, we need to change the definition of this metric
250 00:30:06.500 ⇒ 00:30:07.750 Uttam Kumaran: to XYZ.
251 00:30:07.960 ⇒ 00:30:18.439 Uttam Kumaran: I don’t think we’re very far like I mean, I don’t think I think this is our. This is basically like a few days away. But I think, what what’s been challenging, I think is just like, okay, we want to.
252 00:30:19.159 ⇒ 00:30:26.940 Uttam Kumaran: Add a new visualization or change it. And like, there’s there’s just met like metrics or business logic that we have to implement to support that.
253 00:30:27.260 ⇒ 00:30:29.020 Uttam Kumaran: That’s been taking some time.
254 00:30:29.370 ⇒ 00:30:37.889 Mathew: I think the challenge from our side is, we didn’t have this kind of clarity. And we were just like, here are the graphs, build these graphs and then. Now, this is more systematic.
255 00:30:38.250 ⇒ 00:30:41.700 Mathew: Bottoms up approach to actually, the graphs are.
256 00:30:41.950 ⇒ 00:30:48.489 Mathew: if we do our job right, should be able to be called on demand. If we have this infrastructure set up appropriately.
257 00:30:50.080 ⇒ 00:30:55.180 Uttam Kumaran: Yeah, I mean, but this is because this, like our goal, was to support the graphs. Right? That was the directive.
258 00:30:55.745 ⇒ 00:30:57.440 Mathew: That’s on me.
259 00:30:57.440 ⇒ 00:31:01.305 Uttam Kumaran: Oh, okay, I was like, I was like, yeah, because we were just what we were going for.
260 00:31:01.520 ⇒ 00:31:05.670 Uttam Kumaran: I don’t know. Having seen it and gone through totally. Yes, yes.
261 00:31:05.670 ⇒ 00:31:06.190 Mathew: Instead of.
262 00:31:06.190 ⇒ 00:31:06.610 Uttam Kumaran: Correct.
263 00:31:06.950 ⇒ 00:31:10.820 Uttam Kumaran: The real risk of this project is when the stakeholder says, Yeah.
264 00:31:10.820 ⇒ 00:31:21.509 Uttam Kumaran: because when so dwelled, it’s so narrowly scoped. There’s no flexibility. But again, this is why, like I think, what you’ll see is that we didn’t just like go from Csv to graph
265 00:31:21.660 ⇒ 00:31:26.199 Uttam Kumaran: this. There are a bunch of steps in the middle. Because, yes, like the.
266 00:31:26.990 ⇒ 00:31:49.710 Uttam Kumaran: this is data. It’s like you immediately want to cut it 10 different ways. And, in fact, that’s what you want them to be doing. That is the that is the product like, you want to be like, Oh, yeah, we can totally support that I think, for your like to quell. I think the your nervous system, that is, we’ll make sure that all gets into Dbt. And then, once once you see in the Bi tool, you’ll see that you can quickly chop it 10 different ways.
267 00:31:50.182 ⇒ 00:31:53.019 Mathew: I have a working product like.
268 00:31:53.020 ⇒ 00:31:55.650 Uttam Kumaran: Yeah. Oh, yeah, yeah, yeah. I mean you literally. Yes, yes.
269 00:31:55.650 ⇒ 00:32:01.340 Mathew: I have an end to end product. Until that point I’m I’m like as the as the least technical person here being.
270 00:32:01.340 ⇒ 00:32:02.580 Uttam Kumaran: Okay. Okay.
271 00:32:02.760 ⇒ 00:32:04.235 Mathew: Where where are we at?
272 00:32:04.530 ⇒ 00:32:12.119 Uttam Kumaran: No, we did it. We did it like, you’re, I think, coming. Yeah, it’s actually nice that you see it this way, because this is how we see it.
273 00:32:12.320 ⇒ 00:32:15.739 Uttam Kumaran: And we’ve developed. And this is actually what the team needs.
274 00:32:15.980 ⇒ 00:32:29.790 Uttam Kumaran: But you know, we we filled out of. We filled in the gaps as we went. And so this is actually the way we need it. In fact, I think the what we’ll produce as part of the data platform documentation that we work on. It’s very similar to this.
275 00:32:30.620 ⇒ 00:32:45.840 Mathew: Yeah, this was, yeah. This is my intuitive way, working with Gpt to be like, it’s on the left side. It’s like, How do atomic metrics roll up because, like email and chat are communication, right? Tool usage. It’s basically one of my, these 3 are, when am I clicking on a computer.
276 00:32:46.010 ⇒ 00:32:46.950 Uttam Kumaran: Yes.
277 00:32:46.950 ⇒ 00:32:47.790 Mathew: Activity, this.
278 00:32:47.790 ⇒ 00:33:02.670 Uttam Kumaran: No, actually, the left one is actually really so. We, for all of our clients, we do time we do the time grains, we do metrics, and then dimensions right? And then metrics roll up to aggregate, you aggregate them up, and then you filter by dimensions, and then time grains are just
279 00:33:02.860 ⇒ 00:33:05.190 Uttam Kumaran: based on whatever your your timestamp is.
280 00:33:05.190 ⇒ 00:33:05.660 Mathew: Look at that! But.
281 00:33:05.660 ⇒ 00:33:08.029 Uttam Kumaran: The last thing is actually, really, really helpful.
282 00:33:09.020 ⇒ 00:33:29.679 Mathew: Yeah, I mean, I was like beating the crap out of this. I was like, how does this work? How did like, what are we going to show? Because they might say, Show us all the you know. Show us all the communication people have, and let’s crush that cross check. How like tool usage works against meetings, and that you have to work against all that. And that’s more important than like, show me a graph about XY, and.
283 00:33:29.680 ⇒ 00:33:52.980 Uttam Kumaran: Yes, but that, you know, the thing is like the way we structure the tables is not the way we structure the activity categories because of. That’s like, that’s where there needs to be. For example, there we may have email and chat, but, like communication as a category, we may, we may join those 2 together to build a communication table or a view.
284 00:33:52.980 ⇒ 00:33:55.669 Mathew: Yeah, that’s what I’m asking for. Here.
285 00:33:55.670 ⇒ 00:33:56.350 Uttam Kumaran: Yeah.
286 00:33:56.500 ⇒ 00:34:09.370 Uttam Kumaran: So we see it as like the the smaller atomic, like activity tables, which is like all meetings, all chats, all emails. And then we build like a company communication table which, like, joins those together, and then has like
287 00:34:09.480 ⇒ 00:34:14.749 Uttam Kumaran: 5 metrics like pace, whatever. I don’t know, whatever you know. Yeah.
288 00:34:14.750 ⇒ 00:34:20.480 Mathew: And then we can go back. We can go literally the CEO, and say, This is activity per person across.
289 00:34:20.480 ⇒ 00:34:21.250 Uttam Kumaran: Yes.
290 00:34:21.650 ⇒ 00:34:27.139 Mathew: This is meetings per person across day of the week. Your team here’s how tool usage
291 00:34:27.300 ⇒ 00:34:33.060 Mathew: differs or cross checks against meetings, and then you can against all of this. Obviously.
292 00:34:33.219 ⇒ 00:34:33.969 Uttam Kumaran: Yes.
293 00:34:35.159 ⇒ 00:34:42.239 Mathew: So. So it sounds like this. Logic is not even. I don’t know what you would call is this like.
294 00:34:42.239 ⇒ 00:34:46.049 Uttam Kumaran: No, this is semantic. This is semantic model. This is what’s called semantic model. Yeah.
295 00:34:46.050 ⇒ 00:34:52.804 Mathew: This is probably like, is this also just like, yeah, there, this is. Also, you’d say, like,
296 00:34:53.750 ⇒ 00:34:54.969 Mathew: days away.
297 00:34:56.086 ⇒ 00:35:02.479 Uttam Kumaran: Well, but the the left side is there’s not much on our end. Meaning
298 00:35:02.950 ⇒ 00:35:17.560 Uttam Kumaran: semantic model is like just a semantic understanding of like groups of metrics and dimensions. The right side is what we’re looking to make available. So you can produce anything you want. Right? We’ll we’ll have a set of 5 to 10, to 20 tables
299 00:35:17.730 ⇒ 00:35:22.493 Uttam Kumaran: that can be joined, that that have all of these available to them.
300 00:35:23.050 ⇒ 00:35:32.669 Uttam Kumaran: And then you can form like, Hey, I want a chat table that pulls from this column, this column, this column, this column and these 4 metrics. And then.
301 00:35:33.360 ⇒ 00:35:42.250 Uttam Kumaran: yeah. And then that’s in power. Bi, right now, because we because it’s in python, it’s like very bespoke meaning. We just produce everything to get to the chart
302 00:35:42.720 ⇒ 00:35:46.039 Uttam Kumaran: behind the scenes. We are doing these joins and things like that
303 00:35:46.270 ⇒ 00:35:51.060 Uttam Kumaran: in power. Bi, you’ll literally just be given like you’ll be given 30 metrics and dimensions.
304 00:35:51.620 ⇒ 00:35:54.369 Uttam Kumaran: and you’ll be able to just drag and say cool. I have.
305 00:35:54.790 ⇒ 00:35:55.519 Mathew: So, so.
306 00:35:55.520 ⇒ 00:35:56.580 Uttam Kumaran: Produce this chart.
307 00:35:56.800 ⇒ 00:35:59.939 Mathew: These definitions, semantic, built into what? Go ahead.
308 00:36:00.400 ⇒ 00:36:18.349 Uttam Kumaran: Yeah, so, or or the other thing is, you’re like, Hey, we have 4 tables relate to chat 3 related to email. You could go into like an email related, basically set of tables. Or if you’re like, Hey, I want it. I just want to have a communication module that has all the email tables and chat tables joined together.
309 00:36:18.640 ⇒ 00:36:19.590 Mathew: That’s what we want.
310 00:36:19.910 ⇒ 00:36:24.319 Uttam Kumaran: So this. So so the way we would structure this is in power. It would be in power. Bi.
311 00:36:24.560 ⇒ 00:36:34.920 Uttam Kumaran: because we don’t. You don’t want the user to care. This is the reason why you have this like divide here is because the user doesn’t care what table. This is all coming from
312 00:36:35.340 ⇒ 00:36:44.549 Uttam Kumaran: right, they shouldn’t. Instead, they care about like, I want to ask a communication related question, and I want to have all the metrics and dimensions related to communication.
313 00:36:44.790 ⇒ 00:36:50.110 Mathew: Having the metrics and dimensions related to communication is some is a semantic understanding. Again, I’m just using like.
314 00:36:50.370 ⇒ 00:36:54.110 Uttam Kumaran: Big jargon, but like that is what the technical term is is like.
315 00:36:54.280 ⇒ 00:36:57.139 Uttam Kumaran: This metric is related to communication
316 00:36:57.810 ⇒ 00:37:04.369 Uttam Kumaran: may also be even that level. I, my question for you is like, Will you guys allow us that capability quote?
317 00:37:04.370 ⇒ 00:37:08.609 Uttam Kumaran: Yes, yeah, yeah. Okay, yeah.
318 00:37:08.920 ⇒ 00:37:09.670 Mathew: Great.
319 00:37:10.900 ⇒ 00:37:13.470 Uttam Kumaran: So I think the current, the current gap really is one.
320 00:37:13.700 ⇒ 00:37:15.799 Uttam Kumaran: Oh, I should just get this in Dbt.
321 00:37:17.158 ⇒ 00:37:19.870 Uttam Kumaran: that way. We can see all the logic.
322 00:37:20.170 ⇒ 00:37:21.550 Mathew: Second is.
323 00:37:22.080 ⇒ 00:37:25.470 Uttam Kumaran: Yeah, we as soon as you give us the prbi, we’ll we’ll make it happen.
324 00:37:26.040 ⇒ 00:37:28.349 Mathew: You got him in, I think.
325 00:37:28.350 ⇒ 00:37:35.220 Uttam Kumaran: The other thing is like we could do this in another bi tool, if you want like short term, that’s like free or but like whatever.
326 00:37:35.650 ⇒ 00:37:40.070 Mathew: No, let me let me look into doing the power Bi, and I’ll I’ll plan to get to you tomorrow.
327 00:37:40.070 ⇒ 00:37:43.260 Uttam Kumaran: Or call me if you’re setting it up and like you want to walk through it together.
328 00:37:45.980 ⇒ 00:37:47.430 Mathew: Say wait, sorry. Say it again.
329 00:37:47.430 ⇒ 00:37:51.289 Uttam Kumaran: I said, call me if you want, if you’re setting it up, and you like. Want to walk through it together.
330 00:37:51.980 ⇒ 00:37:55.899 Mathew: Yeah, I mean, I’ll probably just provision it and then give you access and let you guys take it from there.
331 00:37:55.900 ⇒ 00:37:56.660 Uttam Kumaran: Alright, cool.
332 00:37:56.660 ⇒ 00:38:00.710 Mathew: We need? Do we need another power bi person? Or is it? Or you guys got that.
333 00:38:01.120 ⇒ 00:38:03.289 Uttam Kumaran: No, it’s just. It’s just the same like.
334 00:38:03.290 ⇒ 00:38:03.680 Mathew: Right.
335 00:38:03.680 ⇒ 00:38:07.520 Uttam Kumaran: It’s the same stuff as like tableau. It’s just all clear.
336 00:38:08.060 ⇒ 00:38:09.260 Mathew: Okay, okay. We’ll take it.
337 00:38:09.260 ⇒ 00:38:15.539 Uttam Kumaran: Yeah, Trevor, if you could just connect it and then connect it to to Bq and see if just issue a test query.
338 00:38:15.650 ⇒ 00:38:18.299 Uttam Kumaran: and then we’ll go from there.
339 00:38:18.570 ⇒ 00:38:20.819 Mathew: They cause they they do.
340 00:38:22.030 ⇒ 00:38:30.159 Mathew: they use SQL. Server. And so I was thinking it might make sense to also provision a SQL. Server, and then, like we.
341 00:38:30.160 ⇒ 00:38:31.530 Uttam Kumaran: Oh, send it there!
342 00:38:31.890 ⇒ 00:38:33.738 Mathew: Yeah, we need it. We needed to. Well.
343 00:38:34.520 ⇒ 00:38:50.309 Mathew: we actually aren’t going to be on in production. We’re not going to be putting it directly in their SQL. Server. We’re going to be doing it pushing to a blob so it would be extra work to push it there. But I mean, I don’t know the query language isn’t that different? Right? Like it’s still like, not anymore. Bigquery just uses regular. SQL, so let’s just connect to bigquery directly.
344 00:38:50.600 ⇒ 00:38:54.389 Uttam Kumaran: Yeah, we’re not using any like crazy bigquery specific stuff. So.
345 00:38:55.380 ⇒ 00:39:02.380 Mathew: Is there anything infrastructure to make sure it’s easy to go from like whatever the staging into our each customer or whatever. What do you mean.
346 00:39:02.620 ⇒ 00:39:13.510 Mathew: meaning like, you’re obviously gonna be working with them. They’re gonna hook it up. Do you have to do anything infrastructurally to then like, make sure like we provision this customer their specific power bi stuff. Oh, I see like segment. It
347 00:39:13.770 ⇒ 00:39:19.600 Uttam Kumaran: I think, like I think it’s the question is probably gonna be, are you guys gonna put the power bi dashboard behind an interface
348 00:39:20.450 ⇒ 00:39:23.059 Uttam Kumaran: or not. Really right. They’re gonna they’re not. They’re gonna.
349 00:39:23.060 ⇒ 00:39:25.430 Mathew: They work in power, they work in power, bi.
350 00:39:26.390 ⇒ 00:39:30.779 Uttam Kumaran: Like they’re not accessing it through a matter more like Portal or anything. Okay.
351 00:39:30.780 ⇒ 00:39:39.100 Mathew: No, we’re gonna actually probably pipe it directly into their power bi. So we wanna make sure that whatever we give them is like a copy paste for Trevor, and he feels.
352 00:39:39.100 ⇒ 00:39:43.169 Uttam Kumaran: Yeah, then you can. Yeah, Trevor, you can either send it to them through blob.
353 00:39:43.910 ⇒ 00:39:44.510 Mathew: At the plan.
354 00:39:44.510 ⇒ 00:39:46.879 Uttam Kumaran: And then that’s probably that’s probably easy enough.
355 00:39:47.270 ⇒ 00:39:56.430 Mathew: Yeah. Oh, yeah, no. That’s definitely what we’re doing. We’re we’re we’re posting stuff in their blob stores. And then they’re using their system to import into their their stuff. So.
356 00:39:56.430 ⇒ 00:40:21.840 Uttam Kumaran: Yeah. So we so what? What you’ll let leave them with, as you mentioned, is just the final marks. Tables. So what cause? So kind of at the end. What we’ll give them is like the documentation which is like, Hey, we have these like 10 or 20. Here’s like the 3 email channels, chat, chat tables. Blah! Blah! Here’s how to join. Here are the primary keys. Here’s how to join them. And then also we can give them the PBX file right away like the actual power bi config file.
357 00:40:22.020 ⇒ 00:40:37.180 Mathew: Yeah, yeah, the plan is to give them so right to to do all like the model model table stuff. And then, like, figure out which of those tables we actually need to do the metrics that we want, and then, additionally, like what they want to do. Their own metrics.
358 00:40:37.450 ⇒ 00:40:38.140 Uttam Kumaran: Okay.
359 00:40:38.480 ⇒ 00:40:43.860 Mathew: Data team, and then we’ll say, Well, great! We’ll give them whatever either the
360 00:40:44.340 ⇒ 00:40:51.690 Mathew: we’ll give them whatever they need to like, create the visualizations in bigquery or sorry in power. Bi.
361 00:40:52.390 ⇒ 00:40:53.990 Uttam Kumaran: You want like.
362 00:40:54.140 ⇒ 00:40:59.521 Uttam Kumaran: I guess at that point, like, I’m happy to join you guys on a call, or whatever like.
363 00:40:59.820 ⇒ 00:41:03.060 Mathew: Like subcontractor. So, unless.
364 00:41:03.060 ⇒ 00:41:04.799 Uttam Kumaran: Okay. Okay. Okay. Okay. Okay. Whatever.
365 00:41:04.800 ⇒ 00:41:06.909 Mathew: Full time employee. We could figure that out, but, like.
366 00:41:06.910 ⇒ 00:41:14.839 Uttam Kumaran: Okay, I don’t. Whatever like, or just hit me. If you guys are on the meeting and you want me to just be like online answering questions. Just let me know. But.
367 00:41:15.060 ⇒ 00:41:16.200 Mathew: Get that cool.
368 00:41:17.940 ⇒ 00:41:21.020 Uttam Kumaran: Okay. So on our side, we’ll get everything in dbt.
369 00:41:21.290 ⇒ 00:41:23.040 Uttam Kumaran: like, I would like to see that
370 00:41:23.490 ⇒ 00:41:27.439 Uttam Kumaran: I don’t know if Luke’s cause. Tomorrow’s eid, though a lot of people are off
371 00:41:29.290 ⇒ 00:41:38.919 Uttam Kumaran: and Monday, I think also. So I’ll let me send a message. If not, I’ll try to. I’ll try to get this stuff from Annie into Dbt, second thing is, we’re look. I’m gonna go ahead and
372 00:41:39.380 ⇒ 00:41:43.160 Uttam Kumaran: put in our data platform documentation here.
373 00:41:43.540 ⇒ 00:41:51.419 Uttam Kumaran: which will it’ll you’ll see. It’ll just break this out a little bit further, and then we’ll categorize it by the metrics, categories, sub cats, cats, and then family
374 00:41:52.414 ⇒ 00:42:00.339 Uttam Kumaran: and then we’ll actually have the definitions. You can. Actually, that would be great, because you’ll just hand you could just hand that to them, basically as like the documentation.
375 00:42:06.220 ⇒ 00:42:09.659 Uttam Kumaran: What else we? And then we have the feedback on the charts right.
376 00:42:09.880 ⇒ 00:42:16.670 Mathew: Yeah, I mean, yeah, I care less about the specific charts even for like even here than than just this.
377 00:42:17.700 ⇒ 00:42:25.629 Uttam Kumaran: So we’re just gonna work on those phase one charts. And like, let’s just finish that. And like, continue just to push push away on the stuff that’s edge. But yes.
378 00:42:25.780 ⇒ 00:42:30.509 Uttam Kumaran: I I agree with you. I want to just make sure we can enable any basically like.
379 00:42:30.790 ⇒ 00:42:31.220 Mathew: That’s it.
380 00:42:31.220 ⇒ 00:42:32.119 Uttam Kumaran: Any of those charts.
381 00:42:32.120 ⇒ 00:42:36.959 Mathew: Forget about the charts. The charts was almost way of boundary is is not forget it, but like that was a way.
382 00:42:36.960 ⇒ 00:42:37.670 Uttam Kumaran: Yeah, yeah.
383 00:42:38.120 ⇒ 00:42:46.879 Mathew: What they think that they care about. But if I, if we even I, have the capability of being able to say all of this we’ve over. We’ve already like over delivered.
384 00:42:46.880 ⇒ 00:42:51.890 Uttam Kumaran: Okay, okay, okay, yeah, I. I also think the charts are gonna help. Because a lot of people see this like
385 00:42:52.050 ⇒ 00:42:59.649 Uttam Kumaran: this is where even in our world people like are like, Oh, I have these. But like, what kind of what can I even do like? What can I ask? Then you can be like yo. Here’s like.
386 00:42:59.790 ⇒ 00:43:01.120 Uttam Kumaran: here’s something you like.
387 00:43:01.640 ⇒ 00:43:02.770 Mathew: Well, that’s what I that’s what I saw.
388 00:43:02.770 ⇒ 00:43:03.180 Uttam Kumaran: Brain.
389 00:43:03.814 ⇒ 00:43:04.450 Mathew: Effectively.
390 00:43:04.450 ⇒ 00:43:04.890 Uttam Kumaran: Yeah.
391 00:43:05.260 ⇒ 00:43:12.080 Mathew: Be able to see charts to answer these questions so obviously like you guys handing that to me and us when I do our.
392 00:43:12.080 ⇒ 00:43:12.410 Uttam Kumaran: Yeah.
393 00:43:12.850 ⇒ 00:43:16.129 Mathew: Next week, because we said we just signed the contract today.
394 00:43:16.500 ⇒ 00:43:18.659 Uttam Kumaran: Oh, let’s go, guys! Congrats!
395 00:43:18.850 ⇒ 00:43:20.259 Mathew: Let’s go. We’re locked.
396 00:43:20.680 ⇒ 00:43:25.870 Uttam Kumaran: Nice? Oh, great. Okay. Who? Who? Who technically are you talking to? Do you have like a role?
397 00:43:26.370 ⇒ 00:43:28.270 Uttam Kumaran: Is it someone on the data side.
398 00:43:28.810 ⇒ 00:43:31.309 Mathew: Your people. It’s like, Vp, people analytics.
399 00:43:32.470 ⇒ 00:43:38.619 Uttam Kumaran: Okay? Well, okay, the the documentation we give you, whoever
400 00:43:38.770 ⇒ 00:43:42.700 Uttam Kumaran: was on the data side will be very happy. So use it.
401 00:43:42.980 ⇒ 00:43:46.759 Mathew: Yeah, he he’s talking to them tomorrow so like anything that you can do to support.
402 00:43:46.760 ⇒ 00:43:57.190 Uttam Kumaran: Okay. Okay, so let me try to. I’ll try to get something. Yeah, if you show them that they’ll be very happy because it’s rare. The documentation will put in. It’s rare that teams have this, that level of granularity. So.
403 00:43:58.020 ⇒ 00:44:02.099 Mathew: Would it be helpful to have that tomorrow for your conversation with them? He was.
404 00:44:02.100 ⇒ 00:44:06.019 Uttam Kumaran: Like a pretty good data data documentation on the models.
405 00:44:06.320 ⇒ 00:44:08.229 Mathew: That’d be sick that would be sick.
406 00:44:08.390 ⇒ 00:44:09.940 Uttam Kumaran: Okay. Okay. Okay.
407 00:44:10.050 ⇒ 00:44:11.050 Uttam Kumaran: Okay. Good.
408 00:44:11.780 ⇒ 00:44:15.149 Mathew: Yeah, like that. Delivering that is worth the time like for.
409 00:44:15.150 ⇒ 00:44:18.330 Uttam Kumaran: It’s very cause cause I I want to tell you it’s very rare
410 00:44:18.500 ⇒ 00:44:28.429 Uttam Kumaran: to have like metric level documentation at any company that’s like what we try to do for people, because it comes in handy. And that’s the 1st thing that that person will ask.
411 00:44:28.660 ⇒ 00:44:31.160 Uttam Kumaran: and you will be probably one of the 1st people
412 00:44:31.410 ⇒ 00:44:33.480 Uttam Kumaran: in their career to actually like have it so.
413 00:44:33.480 ⇒ 00:44:37.640 Mathew: When when could you cause I’ll format if you send that to me. If you could get that to me.
414 00:44:37.640 ⇒ 00:44:39.690 Uttam Kumaran: I will try to get something.
415 00:44:39.800 ⇒ 00:44:42.500 Uttam Kumaran: I’ll try to do something today, and then.
416 00:44:42.500 ⇒ 00:44:42.930 Mathew: You guys.
417 00:44:42.930 ⇒ 00:44:44.440 Uttam Kumaran: I’ll ask you for some formatting up.
418 00:44:44.980 ⇒ 00:44:52.470 Mathew: Yeah, if you don’t format anything, if you can get it to like like whatever by like, even midnight, I’ll work till, however long, to turn.
419 00:44:53.145 ⇒ 00:44:53.820 Uttam Kumaran: Okay.
420 00:44:53.820 ⇒ 00:44:54.580 Mathew: We’re glad to hear it.
421 00:44:55.180 ⇒ 00:44:58.490 Uttam Kumaran: I wish maybe we’ll talk after this and see what I can get.
422 00:44:59.650 ⇒ 00:45:00.940 Awaish Kumar: Yeah, sure.
423 00:45:01.190 ⇒ 00:45:03.960 Uttam Kumaran: Sure you guys have something you’ve done before. And then you just say.
424 00:45:03.960 ⇒ 00:45:08.999 Uttam Kumaran: No, no, we do. We do. Yeah, we do. We just have to make sure it maps to all of your
425 00:45:09.560 ⇒ 00:45:10.790 Uttam Kumaran: your stuff.
426 00:45:12.946 ⇒ 00:45:15.170 Awaish Kumar: Alright, so that so.
427 00:45:15.760 ⇒ 00:45:19.340 Uttam Kumaran: I have to. I have to jump for another thing.
428 00:45:20.060 ⇒ 00:45:31.009 Mathew: My only my only final question was like, How confident are we if he gets you power Bi by tomorrow that we could have something like, what data we want to shoot for. That’s all. That’s I guess. My only remaining question.
429 00:45:32.350 ⇒ 00:45:35.650 Uttam Kumaran: The power Bi is, gonna take some time like to set it up in power. Bi.
430 00:45:37.580 ⇒ 00:45:39.949 Mathew: What like? What are we talking? Days, weeks.
431 00:45:40.638 ⇒ 00:45:46.300 Uttam Kumaran: Like, probably 2 weeks, like, I have to buffer cause.
432 00:45:47.060 ⇒ 00:45:50.237 Uttam Kumaran: Yeah, but it’s like, probably closer to a week.
433 00:45:53.100 ⇒ 00:45:55.280 Uttam Kumaran: but I have to say 2 weeks.
434 00:45:56.370 ⇒ 00:46:01.919 Uttam Kumaran: because it’s a with a setup like metrics and dimensions and set up the power bi environment.
435 00:46:03.440 ⇒ 00:46:10.150 Mathew: And if we wanted the output, I mean, should we just basically skip python graphs and just go straight to power Bi now with.
436 00:46:12.350 ⇒ 00:46:17.710 Mathew: we just get python graph, like, Oh, I see, like and leading with just like analyst reports, we just go straight to. Just like.
437 00:46:18.200 ⇒ 00:46:21.539 Mathew: yeah, I mean, that’s fine we could do. We could work with that.
438 00:46:22.360 ⇒ 00:46:27.079 Uttam Kumaran: Okay. I mean, it’ll be. You’ll see stuff. It’s not like we go silent, but you’ll see stuff.
439 00:46:27.510 ⇒ 00:46:28.640 Mathew: Yeah, cause I’m
440 00:46:28.890 ⇒ 00:46:37.840 Mathew: I guess. When could you get us the for the for, I guess. Is it like one of those things where you have to set up. I know you have to go. You set up all of the thing, and then you press a button, and now.
441 00:46:37.840 ⇒ 00:46:42.369 Uttam Kumaran: Yeah, we have to like, connect all the core tables into power, bi
442 00:46:42.520 ⇒ 00:46:47.760 Uttam Kumaran: map them to like dimensions and metrics. And then we’ll be able to start producing the 1st graphs.
443 00:46:48.500 ⇒ 00:46:51.099 Awaish Kumar: And like we, we might have to adjust some
444 00:46:51.450 ⇒ 00:46:53.709 Awaish Kumar: SQL. Queries we have wrote, because
445 00:46:53.930 ⇒ 00:47:03.310 Awaish Kumar: now, because they are in, we are using python, so some are. Some of the modeling is an SQL. And that analyst might have some did some tuning on their part.
446 00:47:03.440 ⇒ 00:47:08.229 Awaish Kumar: So we have to put add that all into the Dbt. Project, as well.
447 00:47:11.710 ⇒ 00:47:12.589 Mathew: Are you good with that?
448 00:47:13.070 ⇒ 00:47:13.780 Mathew: Yeah.
449 00:47:14.060 ⇒ 00:47:24.689 Mathew: yeah. I mean the risk there is. It’s like we lose our analyst edge or whatever. And we’re now like working within power Bi, but like everything that we’re trying to do here should be done doable in power. Bi, right?
450 00:47:25.670 ⇒ 00:47:33.300 Uttam Kumaran: Yeah, I’m not worried at all power bi, but like I don’t, there’s not. There’s not. An I don’t think anybody’s gonna tell you this is, gonna take a day like.
451 00:47:33.300 ⇒ 00:47:35.510 Mathew: Alright! No, no, no, that’s fine, that’s fine. We could go. We.
452 00:47:35.510 ⇒ 00:47:39.959 Uttam Kumaran: Yeah, it’s good. I have. It’s gonna I think it’s, I think, 2 weeks is.
453 00:47:39.960 ⇒ 00:47:44.130 Mathew: That’s fine. And he and you’re gonna need time to integrate with their systems anyway and adjust all the data. Right?
454 00:47:44.480 ⇒ 00:47:47.010 Mathew: Yeah, so okay, all right. But.
455 00:47:47.010 ⇒ 00:47:53.740 Uttam Kumaran: And also you’ll as soon as we’re in, you’ll have access to this, Matthew. So you can start to play around like that’s gonna be nice.
456 00:47:53.950 ⇒ 00:47:56.750 Mathew: Okay, cool. Start, do that.
457 00:47:57.200 ⇒ 00:47:57.590 Mathew: Okay?
458 00:47:57.880 ⇒ 00:48:01.470 Mathew: Like the data dictionary, the metrics dictionary, or whatever that you’re sharing
459 00:48:01.610 ⇒ 00:48:06.609 Mathew: huge for tonight. We’ll get you power. Bi, you’ll start on that.
460 00:48:07.010 ⇒ 00:48:07.895 Mathew: And
461 00:48:09.000 ⇒ 00:48:16.520 Mathew: I mean that was that. Was it coming out of it from our side? Like, if we have this capability the charts are. It’s like charting is moot. Because that’s actually just.
462 00:48:16.520 ⇒ 00:48:19.280 Uttam Kumaran: Sure. Yeah, yeah, yeah, I agree. No, that’s what I agree.
463 00:48:20.160 ⇒ 00:48:22.949 Mathew: Then we could talk about all this stuff and get crazy over here.
464 00:48:24.800 ⇒ 00:48:27.179 Uttam Kumaran: Yes, this is what I want to help you guys on.
465 00:48:27.490 ⇒ 00:48:30.460 Mathew: Yeah, this is, this is like some next level, like.
466 00:48:30.460 ⇒ 00:48:31.120 Uttam Kumaran: Yeah.
467 00:48:31.260 ⇒ 00:48:31.930 Mathew: Yeah.
468 00:48:33.750 ⇒ 00:48:35.149 Uttam Kumaran: Okay, I gotta run.
469 00:48:35.380 ⇒ 00:48:37.019 Mathew: Alright. Thank you. Team.
470 00:48:37.020 ⇒ 00:48:38.110 Uttam Kumaran: Thank you. Thank you.
471 00:48:39.320 ⇒ 00:48:40.350 Mathew: See you awesome.
472 00:48:40.570 ⇒ 00:48:41.170 Awaish Kumar: You know.
473 00:48:41.930 ⇒ 00:48:42.930 Mathew: Take care!