2024-05-30_uttam_robert--bi-tool-discussion_68de1b3a

Meeting Title: Uttam <> Robert—BI-Tool-Discussion Date: 2024-05-30 Meeting participants: Robert Tseng, Nicolas Sucari, Uttam Kumaran

---

WEBVTT

1 00:01:19.710 ⇒ 00:01:20.600 Nicolas Sucari: Hi! You! Tom!

2 00:01:21.190 ⇒ 00:01:21.680 Uttam Kumaran: Hey!

3 00:01:23.360 ⇒ 00:01:24.190 Nicolas Sucari: How are you?

4 00:01:24.680 ⇒ 00:01:25.579 Uttam Kumaran: Good! How are you?

5 00:01:26.390 ⇒ 00:01:27.980 Nicolas Sucari: I’m fine. You

6 00:01:28.230 ⇒ 00:01:31.389 Nicolas Sucari: manage to sleep something or not at all.

7 00:01:31.776 ⇒ 00:01:36.030 Uttam Kumaran: Yeah, yeah, yeah, we all, we always get a little sleep.

8 00:01:38.060 ⇒ 00:01:43.030 Uttam Kumaran: No, I’m okay. I main thing was to get shipments done. And then

9 00:01:43.370 ⇒ 00:01:46.393 Uttam Kumaran: there’s a couple of things to debug today. But

10 00:01:47.690 ⇒ 00:01:52.370 Uttam Kumaran: yeah, we’re good. I mean our con. Our conversation helped yesterday. So that gives me motivation.

11 00:01:53.350 ⇒ 00:01:53.990 Nicolas Sucari: Yeah.

12 00:01:54.570 ⇒ 00:01:55.440 Nicolas Sucari: Okay.

13 00:02:06.210 ⇒ 00:02:08.259 Nicolas Sucari: hey? Can I ask you,

14 00:02:09.060 ⇒ 00:02:11.830 Nicolas Sucari: who is chuck in both parts?

15 00:02:12.640 ⇒ 00:02:15.089 Uttam Kumaran: Chuck is the head of their

16 00:02:15.150 ⇒ 00:02:16.670 Uttam Kumaran: shipping warehouse.

17 00:02:17.270 ⇒ 00:02:17.840 Nicolas Sucari: And there.

18 00:02:17.840 ⇒ 00:02:19.880 Uttam Kumaran: So basically anything to do with shipping?

19 00:02:20.860 ⇒ 00:02:22.170 Nicolas Sucari: Keep me! Check.

20 00:02:22.820 ⇒ 00:02:23.390 Uttam Kumaran: Dad

21 00:02:24.030 ⇒ 00:02:30.229 Uttam Kumaran: Ben knows a lot about it, but Chuck is like the guy in the warehouse managing all the staff, I mean.

22 00:02:33.500 ⇒ 00:02:34.320 Uttam Kumaran: hey.

23 00:02:34.320 ⇒ 00:02:35.459 Robert Tseng: Hey, guys, yeah.

24 00:02:36.310 ⇒ 00:02:37.079 Uttam Kumaran: How’s it going.

25 00:02:38.590 ⇒ 00:02:39.470 Robert Tseng: Good! How are you?

26 00:02:40.490 ⇒ 00:02:41.180 Uttam Kumaran: Good.

27 00:02:42.880 ⇒ 00:02:43.860 Robert Tseng: Sorry, really.

28 00:02:44.560 ⇒ 00:02:46.300 Uttam Kumaran: Yeah, start early.

29 00:02:46.833 ⇒ 00:02:50.389 Uttam Kumaran: I I yeah, I don’t know. I’ve been. I feel like I

30 00:02:50.580 ⇒ 00:02:57.250 Uttam Kumaran: this week. For some reason I I was feeling energized. I haven’t needed a lot of sleep. We also have a lot

31 00:02:58.030 ⇒ 00:02:59.901 Uttam Kumaran: of stuff going on.

32 00:03:01.320 ⇒ 00:03:02.722 Uttam Kumaran: so you know.

33 00:03:03.190 ⇒ 00:03:03.820 Robert Tseng: Deaf.

34 00:03:07.340 ⇒ 00:03:09.240 Uttam Kumaran: cool. So

35 00:03:09.590 ⇒ 00:03:17.780 Uttam Kumaran: I think we just start by like running through everything. I think I’m just gonna try to go through as.

36 00:03:18.460 ⇒ 00:03:19.160 Robert Tseng: Like.

37 00:03:19.490 ⇒ 00:03:24.190 Uttam Kumaran: Quick and gentle as possible. Real evidence and light dash.

38 00:03:24.945 ⇒ 00:03:25.670 Uttam Kumaran: I think

39 00:03:26.770 ⇒ 00:03:28.629 Uttam Kumaran: the 1st 2 will be a little bit

40 00:03:28.780 ⇒ 00:03:36.810 Uttam Kumaran: like out of the ordinary in terms of like, they’re kind of a little bit of new workflow. But like that should be port familiar. Basically, I want to go through a couple of things. I want to go.

41 00:03:37.330 ⇒ 00:03:47.009 Uttam Kumaran: The the primary goal for me and and choosing a Vi tool is always. Is this something that whoever you need to adopt is going to adopt.

42 00:03:47.190 ⇒ 00:03:51.050 Uttam Kumaran: Right? Secondary is our workflow. But

43 00:03:51.090 ⇒ 00:03:52.609 Uttam Kumaran: ideally we marry

44 00:03:52.930 ⇒ 00:03:56.049 Uttam Kumaran: those 2, and we pick something that works for everybody. But

45 00:03:56.120 ⇒ 00:03:58.259 Uttam Kumaran: yeah, the initial goal is like

46 00:03:59.030 ⇒ 00:04:06.609 Uttam Kumaran: doing something that the folks can actually use, that they can log in and find the answers, and that if they do have questions.

47 00:04:06.640 ⇒ 00:04:18.520 Uttam Kumaran: There’s a process by which they can ask us for it, or there’s even better like at least some ability for them to get closer to the answer themselves. So that’s like, always kind of like what’s in in the back of my mind. Of course.

48 00:04:18.660 ⇒ 00:04:21.479 Uttam Kumaran: cost and everything’s a factor. All 3 of these tools.

49 00:04:21.850 ⇒ 00:04:25.389 Uttam Kumaran: I would say, compared to Sigma

50 00:04:25.550 ⇒ 00:04:28.253 Uttam Kumaran: looker. Tableau are cheaper.

51 00:04:28.930 ⇒ 00:04:29.710 Robert Tseng: Yeah.

52 00:04:29.710 ⇒ 00:04:32.519 Uttam Kumaran: And we could talk about the cost. I’ve I’ve talked to

53 00:04:32.870 ⇒ 00:04:38.829 Uttam Kumaran: basically, all these guys, and, you know, have a good understanding of what’s possible. But let’s just jump right into things.

54 00:04:39.240 ⇒ 00:04:39.945 Robert Tseng: Okay.

55 00:04:43.310 ⇒ 00:04:44.390 Uttam Kumaran: So

56 00:04:44.490 ⇒ 00:04:46.849 Uttam Kumaran: see how? What’s my switch?

57 00:05:03.500 ⇒ 00:05:04.250 Uttam Kumaran: Okay?

58 00:05:08.150 ⇒ 00:05:13.580 Uttam Kumaran: So in zoom, you can actually share like 2 windows now. So it’s kinda nice. I’m gonna share my code window.

59 00:05:13.580 ⇒ 00:05:15.130 Robert Tseng: Oh, cool!

60 00:05:15.500 ⇒ 00:05:19.399 Uttam Kumaran: And then I’m gonna try to share this. Let me know how it looks, because I don’t know.

61 00:05:21.110 ⇒ 00:05:22.620 Uttam Kumaran: How does it look on your side?

62 00:05:23.360 ⇒ 00:05:29.960 Robert Tseng: Yeah, if we could zoom in a bit that’d be helpful but yeah, otherwise it works.

63 00:05:30.610 ⇒ 00:05:31.607 Uttam Kumaran: Okay, let me

64 00:05:32.780 ⇒ 00:05:35.739 Uttam Kumaran: I’m just gonna try to. Is the code one better now.

65 00:05:37.232 ⇒ 00:05:38.860 Robert Tseng: Yes, it is.

66 00:05:39.930 ⇒ 00:05:40.550 Uttam Kumaran: Okay.

67 00:05:41.750 ⇒ 00:05:46.301 Uttam Kumaran: thanks. Appreciate it. Cool. So this is

68 00:05:47.440 ⇒ 00:05:53.979 Uttam Kumaran: real. As I mentioned, this is from this company called Real. They’re@realdata.com.

69 00:05:54.877 ⇒ 00:05:56.392 Uttam Kumaran: basically the

70 00:05:57.280 ⇒ 00:06:01.479 Uttam Kumaran: the whole philosophy behind these guys product is

71 00:06:01.990 ⇒ 00:06:17.900 Uttam Kumaran: like, it’s very, very quick. It’s very, very good for heavy operational use cases where you’re doing things very commonly, like filtering by a specific dimension. You’re looking at changes between periods, changes between segments.

72 00:06:17.900 ⇒ 00:06:46.250 Uttam Kumaran: percent of totals, anything where you have a high dimensionality and time series based data. It’s not really great for anything where you’re just looking at like point data, but anything time series, which is most of the stuff that we work with. It’s really, really great. So the basic kind of model of real is that you have sources, models and dashboards. I’ll be walking through kind of like the development process. What you see here is also the same thing that I have on the left.

73 00:06:46.738 ⇒ 00:06:47.871 Uttam Kumaran: So I’ll just

74 00:06:48.500 ⇒ 00:06:50.210 Uttam Kumaran: walk through it everything.

75 00:06:50.330 ⇒ 00:06:54.980 Uttam Kumaran: This sits all on top of our Dbt models. So

76 00:06:55.070 ⇒ 00:07:14.730 Uttam Kumaran: let’s start with something that we’ve been working on internally, which is for shipments. We have a shipments table. That we basically develop that just shows for this client all of their products. They’re shipping where they’re shipping from, and the cost to ship and a couple of other helpful dimensions. What we’re doing in here is

77 00:07:15.074 ⇒ 00:07:26.085 Uttam Kumaran: just running a select star from the shipments. Table. The shipments source is actually defined here in shipments, Dot, Yaml, and all that is also select star from Dbt March shipments.

78 00:07:26.540 ⇒ 00:07:34.599 Uttam Kumaran: so they have the opportunity to do a couple of things. One is they have sources. So sources are just selects on your tables that you need

79 00:07:34.690 ⇒ 00:08:01.140 Uttam Kumaran: models. Actually allow you to join models together. For example, let’s say you have something coming from a postgres database, something coming from Snowflake and something coming from like a Csc file. You can actually join all those 3 together in real. And it’s being built in this technology that you may have heard of called duct dB, basically, what duck dB, does is it takes data, puts it in an in memory database in the browser.

80 00:08:01.230 ⇒ 00:08:05.869 Uttam Kumaran: So one it’s like almost like sequel light. If you’ve heard of sqlite. But basically.

81 00:08:05.890 ⇒ 00:08:25.808 Uttam Kumaran: things are so quick because you’re not saying the query to the warehouse, you’re actually just setting a query to the in memory database, and what it does is every so often it pulls it all the data from the warehouse into this in memory database, and you’ll run queries. You kind of see the quickness that I’m talking about in just a second

82 00:08:26.150 ⇒ 00:08:36.487 Uttam Kumaran: So we set up that we want to select everything from shipments again. We’re not doing any modeling here. So all of our things are pretty much straight selects. And then we set up a dashboard file.

83 00:08:37.230 ⇒ 00:08:56.726 Uttam Kumaran: the nice thing about this dashboard files is really just 2 things to set up. It’s just dimensions and measures. Dimensions are what you know. It’s just basically like dimensions. And then measures are aggregates. So you, they have sums. You have min. Max counts.

84 00:08:57.240 ⇒ 00:08:59.998 Uttam Kumaran: these can have different formats.

85 00:09:01.020 ⇒ 00:09:22.940 Uttam Kumaran: that’s basically it because real and real focus a lot on time series, you have a lot of time series related configurations that you could do. I’ll show you how these kind of come in handy within the dashboard things like setting what time zone people are able to switch things like setting time ranges that people can really quickly select like last 6 h last 5 days last 3 months.

86 00:09:23.654 ⇒ 00:09:27.770 Uttam Kumaran: So let’s just walk through this example within the platform. So

87 00:09:28.440 ⇒ 00:09:33.990 Uttam Kumaran: you have real. This is also running locally. We do have this hosted

88 00:09:34.397 ⇒ 00:09:41.920 Uttam Kumaran: but I just wanted to show you that it’s actually really quick, and it’s free to run this all locally. So a lot of people use this for local development and local analysis.

89 00:09:42.293 ⇒ 00:09:50.706 Uttam Kumaran: As well as customers. So we have. We have our shipments table. We have a shipments model. The nice thing is when you’re developing in real

90 00:09:51.900 ⇒ 00:09:54.840 Uttam Kumaran: and you like, bring in a new table.

91 00:09:54.920 ⇒ 00:09:57.658 Uttam Kumaran: It actually is very nice, because it

92 00:09:58.170 ⇒ 00:10:00.354 Uttam Kumaran: it will tell you.

93 00:10:01.450 ⇒ 00:10:04.459 Uttam Kumaran: It’ll tell you everything about the columns that you’re working with.

94 00:10:04.732 ⇒ 00:10:22.289 Uttam Kumaran: Like, I’m basically for the Me on the data engineering analytics engineering side, when I’m loading data. And I want to see like, Oh, there’s gonna be. There’s a huge thing with noles. I have to go figure that out. Things like that are very nice. I’m just gonna select from a development table that I have that has a little bit more data in it.

95 00:10:25.170 ⇒ 00:10:30.199 Uttam Kumaran: I don’t know whether it exists today. Oh, sorry, not this one. Actually.

96 00:10:31.200 ⇒ 00:10:32.260 Uttam Kumaran: Here we go.

97 00:10:32.670 ⇒ 00:10:34.489 Uttam Kumaran: So this is the shipments table.

98 00:10:36.380 ⇒ 00:10:42.659 Uttam Kumaran: The nice thing is what you could do. Here is we created a model and then from the model we created a dashboard.

99 00:10:43.237 ⇒ 00:10:52.950 Uttam Kumaran: So if we go to shipping models dashboard, here’s the code, basically the same thing we just showed. It’s actually the same file just reading from, and you could see helpful information about each dimension

100 00:10:53.637 ⇒ 00:10:58.160 Uttam Kumaran: and again, just really helpful in the development process. When we go to Preview, which is this is the

101 00:10:58.690 ⇒ 00:11:01.249 Uttam Kumaran: main sauce is like, this is actually what

102 00:11:01.380 ⇒ 00:11:03.219 Uttam Kumaran: you see when you’re in rail.

103 00:11:03.775 ⇒ 00:11:21.209 Uttam Kumaran: So if you strike before, I always talk about how this kind of looks like the stripe dashboard. But basically you just have the ability to quickly filter to different time series quickly filter to different time zones add comparisons.

104 00:11:21.250 ⇒ 00:11:35.229 Uttam Kumaran: change the aggregate to month, week, quarter a year. What you see on the right is all of our dimensions, and you’ll see values associated with it. So, for example, if you want to say great, I want to know how much we spent on usps.

105 00:11:35.390 ⇒ 00:11:38.189 Uttam Kumaran: And I want to change this to total shipping amount.

106 00:11:38.495 ⇒ 00:11:44.104 Uttam Kumaran: You can see here. This is the total shipping amount that we spent on usps in the past year.

107 00:11:44.720 ⇒ 00:11:47.859 Uttam Kumaran: All the helpful dimension all the helpful measures that we set up.

108 00:11:48.511 ⇒ 00:11:53.730 Uttam Kumaran: And quickly you could see that as we click, everything filters very quickly.

109 00:11:53.780 ⇒ 00:12:00.810 Uttam Kumaran: So commonly when you’re doing the sip analysis is is, I might go in here and say, I want to learn about our usps like

110 00:12:01.100 ⇒ 00:12:03.549 Uttam Kumaran: process. Okay, cool. So

111 00:12:03.600 ⇒ 00:12:05.710 Uttam Kumaran: I want to know for usps

112 00:12:05.790 ⇒ 00:12:07.409 Uttam Kumaran: how many shipments we have?

113 00:12:07.743 ⇒ 00:12:12.600 Uttam Kumaran: How much volume we’ve shipped through it. Great what are our top zip codes

114 00:12:12.640 ⇒ 00:12:32.772 Uttam Kumaran: cool. So this is total shipping amount. But let’s actually look at total shipments. So looks like, okay, these zip codes are top. They’re all more shipping from this warehouse. And you know we can see the difference between Amazon shopify and zones. Another interesting thing is, you can go here. And you could actually switch to different models. So let’s go actually look at like all orders, for example.

115 00:12:33.381 ⇒ 00:12:38.730 Uttam Kumaran: This is like all the orders for this client in particular. Again, let’s just look at the last month.

116 00:12:38.970 ⇒ 00:12:45.449 Uttam Kumaran: And let’s just say, like, great, I wanna know, like, basically, what are the top states that we’ve been shipping to this past month?

117 00:12:45.630 ⇒ 00:12:48.869 Uttam Kumaran: Cool? I can see it right here. Texas, Florida, New York.

118 00:12:48.950 ⇒ 00:12:50.550 Uttam Kumaran: I can hit, expand.

119 00:12:50.570 ⇒ 00:13:01.050 Uttam Kumaran: and also really quickly. Here I can look at a lot of helpful dimensions. Not only the measures that we have access to here, but also percent of totals and averages.

120 00:13:01.690 ⇒ 00:13:04.060 Uttam Kumaran: It’s really easy to export this from here.

121 00:13:04.528 ⇒ 00:13:08.509 Uttam Kumaran: You know, I can. You could search for different values.

122 00:13:08.850 ⇒ 00:13:16.881 Uttam Kumaran: This is really helpful for just exploring. The other thing you can do is you can do comparisons. So if I was to do comparison on

123 00:13:18.980 ⇒ 00:13:20.560 Uttam Kumaran: Let’s say

124 00:13:25.830 ⇒ 00:13:29.290 Uttam Kumaran: I was want to do a comparison based on time, and I want to compare

125 00:13:29.666 ⇒ 00:13:41.100 Uttam Kumaran: to the previous year. One thing I can do here is now you really easily get to see comparisons right on the charts here. So as you hover, you can basically see cool between

126 00:13:41.200 ⇒ 00:13:43.070 Uttam Kumaran: whatever it is. May 20th

127 00:13:43.110 ⇒ 00:13:53.450 Uttam Kumaran: this year versus last year. Here is the differences where this really probably comes in handy is, if you were to do like week. And you want to look at like last 3 months. You can basically easily see

128 00:13:53.510 ⇒ 00:14:04.033 Uttam Kumaran: how we did last year versus this year, and on the right. You could see a lot of helpful things. So you could see percent changes between this year and last year. You could also do absolute changes.

129 00:14:06.070 ⇒ 00:14:07.770 Uttam Kumaran: so I mean, I probably

130 00:14:07.830 ⇒ 00:14:22.885 Uttam Kumaran: like I could continue on just with a lot of the features. But I think what I’m trying to impart is that like. It’s just super quick, and it’s very helpful for analysis. I’ll give you one more feature, and then I’ll kind of talk about some of the rooms for improvement.

131 00:14:23.200 ⇒ 00:14:25.859 Uttam Kumaran: so of course, you can click here and you can

132 00:14:26.050 ⇒ 00:14:34.623 Uttam Kumaran: filter by whatever dimension you need to. There’s also pivot table here, so you can bring in

133 00:14:35.520 ⇒ 00:14:42.540 Uttam Kumaran: like the week you can bring in total orders, and you can bring in like the selling platform, and you get a pivot table.

134 00:14:45.930 ⇒ 00:14:47.740 Uttam Kumaran: fairly standard stuff.

135 00:14:48.255 ⇒ 00:14:55.879 Uttam Kumaran: Their add, I think we’re just on the older version. I need to actually update my version. But you could actually do bar charts here.

136 00:14:56.283 ⇒ 00:15:01.436 Uttam Kumaran: and everything. You can kind of export. So the the thing that’s a little bit

137 00:15:02.010 ⇒ 00:15:06.789 Uttam Kumaran: different than common bi platforms is really this is your dashboard.

138 00:15:08.220 ⇒ 00:15:11.490 Uttam Kumaran: there’s not a great ability to drag and drop

139 00:15:12.265 ⇒ 00:15:13.020 Uttam Kumaran: and

140 00:15:13.050 ⇒ 00:15:17.126 Uttam Kumaran: really you just set up those measures, and eventually they pop up here.

141 00:15:17.730 ⇒ 00:15:31.140 Uttam Kumaran: so you can do formatting of the values. But really this is fixed, and the pivot is kind of fixed. They are adding the ability to to add bar graphs here and the ability to customize this a bit. But

142 00:15:31.460 ⇒ 00:15:35.829 Uttam Kumaran: I like real in that. It gets my team out of

143 00:15:35.990 ⇒ 00:15:43.500 Uttam Kumaran: doing like small minute dashboard changes, changing small colors instead, really helps you focus on analysis.

144 00:15:44.640 ⇒ 00:15:45.989 Uttam Kumaran: so I’ll pause there.

145 00:15:46.650 ⇒ 00:15:51.550 Uttam Kumaran: We I can walk through developing one of these. But I’ll just if you have any questions.

146 00:15:51.640 ⇒ 00:15:54.389 Uttam Kumaran: or or even let me know what you think, just from seeing this.

147 00:15:54.880 ⇒ 00:16:03.221 Robert Tseng: Yeah, I mean, yeah, thanks for walking me through this super. Helpful to to see see it in action, and kind of how you use it for analysis.

148 00:16:04.390 ⇒ 00:16:09.339 Robert Tseng: yeah, I mean, it is fast. I think that’s the that’s the 1st thing that that’s fixed out.

149 00:16:10.880 ⇒ 00:16:16.542 Robert Tseng: yeah, I mean, I I totally see how this is. More, it’s like a this is, this is useful for an operator.

150 00:16:17.350 ⇒ 00:16:20.689 Robert Tseng: yeah, I I like that. It is very

151 00:16:22.590 ⇒ 00:16:23.310 Robert Tseng: like

152 00:16:23.880 ⇒ 00:16:33.929 Robert Tseng: anything that you put in there. Like, all, all the dimensions are feature here, and every like time based license, and like that value based like grouping that you that you would expect

153 00:16:34.467 ⇒ 00:16:39.222 Robert Tseng: so that definitely saves saves a lot of time and setting all that up. Typically,

154 00:16:40.100 ⇒ 00:16:49.009 Robert Tseng: yeah. So I think that. Yeah, it’s it is a very different Vi work flow than I would be that that’s like I’m used to. I’m like trying to connect it to

155 00:16:49.020 ⇒ 00:16:53.090 Robert Tseng: kind of like the traditional work flow that that I work through.

156 00:16:55.230 ⇒ 00:16:56.220 Robert Tseng: And

157 00:16:56.680 ⇒ 00:16:58.110 Robert Tseng: yeah, I mean, I think

158 00:17:00.180 ⇒ 00:17:18.130 Robert Tseng: so. I think my my take is like give the end consumer of this dashboard ultimately is is typically a non technical. It’s not taught typically a non technical person. And yeah, I do spend so much time like trying to get them to describe, but like what they want defining requirements, giving like mock ups and doing that whole like.

159 00:17:18.130 ⇒ 00:17:33.709 Robert Tseng: you know it becomes like this, this whole process to get that. So it’s a lot. If if I could, just, you know, get gather all the piece of data I needed. Like, you know, we kind of had that Google, Doc. I’m thinking now for Stella. We will dog with all the dashboards we needed all the fields.

160 00:17:33.830 ⇒ 00:17:54.750 Robert Tseng: Theoretically, if I threw that all into, I guess, like your the the dev part of of this of real but yeah, you would. It would just spend on dashboard with all these things. And I wouldn’t actually have to go in and really like, build any of that. And yeah, instead, I I see how the time is a lot more focused on like trying to identify

161 00:17:55.150 ⇒ 00:18:03.969 Robert Tseng: trans patterns. And like more of the analysis piece. So, yeah, definitely could see how that speed speed speeds things up.

162 00:18:05.120 ⇒ 00:18:20.800 Uttam Kumaran: Yeah. One thing I’ll I’ll I’ll just add in there. Sorry to interrupt. I just some people are sometimes intimidated by how it looks which, as data, people like, I don’t have that cause. I just so this is like how my brain is basically.

163 00:18:20.800 ⇒ 00:18:21.420 Robert Tseng: Yeah, yeah.

164 00:18:21.420 ⇒ 00:18:24.050 Uttam Kumaran: For for sometimes 4 people

165 00:18:24.100 ⇒ 00:18:33.909 Uttam Kumaran: like for executives. I found that they’re like, I don’t know, so I think there, ha! There is probably some level of training or some level of walkthrough.

166 00:18:34.120 ⇒ 00:18:43.029 Uttam Kumaran: but the amount of time this saves us in like having to set up these all as filters again like having to wait for these queries to run

167 00:18:43.400 ⇒ 00:18:46.930 Uttam Kumaran: again to do like, for example, if we were to do.

168 00:18:46.970 ⇒ 00:18:48.550 Uttam Kumaran: I want to see

169 00:18:48.740 ⇒ 00:18:53.060 Uttam Kumaran: this year versus last year for Amazon, for this one thing.

170 00:18:53.680 ⇒ 00:18:55.289 Uttam Kumaran: or like this one

171 00:18:55.310 ⇒ 00:18:56.520 Uttam Kumaran: dimension

172 00:18:56.920 ⇒ 00:18:58.930 Uttam Kumaran: that’s like a 30 min exercise.

173 00:18:59.000 ⇒ 00:19:02.100 Uttam Kumaran: let alone doing it for everything right? And so

174 00:19:02.380 ⇒ 00:19:06.750 Uttam Kumaran: for me. That’s like, when I when I 1st saw this, I was like Whoa.

175 00:19:06.790 ⇒ 00:19:09.539 Uttam Kumaran: like everything’s a parameterized filter.

176 00:19:09.600 ⇒ 00:19:16.609 Uttam Kumaran: and it’s very snappy. And anything time series based. They’re really, really gnarly at

177 00:19:16.964 ⇒ 00:19:18.790 Uttam Kumaran: so that’s what I really liked.

178 00:19:19.400 ⇒ 00:19:20.650 Robert Tseng: Yeah, totally.

179 00:19:22.510 ⇒ 00:19:30.299 Robert Tseng: Yeah, definitely, the the lot. More training on like, how to actually, how to actually use this teaching people to

180 00:19:30.770 ⇒ 00:19:58.209 Robert Tseng: as they ask questions like how they would go about doing like looking for it through through this workflow. I I feel like that’s gonna be where the bulk that would. That’s that would be where the bulk of the time would be spent. But I could also see it being more helpful where, like now, I’m like thinking about. Yeah, it’s been, you know. It’s taken a long time to try to like, get define the requirements with with the cell folks, and they keep asking me, well, we don’t really know what we want. Just like, show us data. We’ll export ourselves.

181 00:19:58.210 ⇒ 00:19:58.630 Uttam Kumaran: Yeah.

182 00:19:58.630 ⇒ 00:20:04.919 Robert Tseng: So this is actually more, it’s it’s more. This is more fit for them, because they don’t exactly know what they’re looking for.

183 00:20:05.321 ⇒ 00:20:17.860 Robert Tseng: And yeah, like, if anything, they just want everything out there, and then they can go. And like they can, they can spend spend their time like looking through it. I don’t actually think that’s like the best way to like.

184 00:20:17.860 ⇒ 00:20:18.210 Uttam Kumaran: Yeah.

185 00:20:18.210 ⇒ 00:20:25.259 Robert Tseng: Use a dashboard. But for exploratory analysis, like, I see that this is this is, this fits, this fits that need

186 00:20:26.186 ⇒ 00:20:26.589 Robert Tseng: so.

187 00:20:26.590 ⇒ 00:20:30.209 Uttam Kumaran: I I also think like for S, for some folks like

188 00:20:30.420 ⇒ 00:20:42.039 Uttam Kumaran: the number. One thing that I want is, I want people to be using the data. And so if us not having a tile on a dashboard, and we have that for some customers that like they’re like, I’m waiting for this thing. I can’t use it yet.

189 00:20:42.150 ⇒ 00:20:42.690 Uttam Kumaran: you know.

190 00:20:42.690 ⇒ 00:20:43.030 Robert Tseng: Yeah, I.

191 00:20:43.030 ⇒ 00:21:08.929 Uttam Kumaran: I want. I don’t want them to like click something something breaks, and they’re like they lost trust and set up. I I used to be very fearful, like people going in and like they like click around. And but again, if you trust that we’re our data model works and that we’ve debugged stuff. I want people to go in and find that like, hey, look, we’re getting. We’re getting some packages that we’re in rocketing shipment heights. Okay? Is that important? Well, what’s the problem there? And then? We can debug that right. But

192 00:21:09.070 ⇒ 00:21:10.180 Uttam Kumaran: I like.

193 00:21:10.340 ⇒ 00:21:13.200 Uttam Kumaran: I really think, just seeing these dimensions and some people

194 00:21:13.450 ⇒ 00:21:22.868 Uttam Kumaran: there. My brain is kind of works like this, where I need to see everything in order to do the analysis where I’m like, we’re. I wanna talk about all of our orders and find trends.

195 00:21:23.170 ⇒ 00:21:33.030 Uttam Kumaran: Okay, great. I need to see these types of things. So I can say, like great, what kind of we have dimensions on geography, we have dimensions on time, we have dimensions on kind of shipments.

196 00:21:33.220 ⇒ 00:21:42.810 Uttam Kumaran: and like, I need this to kind of put together analysis so frankly, even, for when I do analysis for for customers, I actually use this locally.

197 00:21:42.980 ⇒ 00:21:48.570 Uttam Kumaran: commonly because it’s faster than using the wherever the dashboard lives

198 00:21:49.165 ⇒ 00:21:54.679 Uttam Kumaran: right? And so even that is a big enough signal for me that like, okay, it’s helping me a lot

199 00:21:55.030 ⇒ 00:21:57.560 Uttam Kumaran: for for some people I’ve been able to kind of like.

200 00:21:58.080 ⇒ 00:22:03.119 Uttam Kumaran: been able to get them onto this initially for some people. If if we’ve introduced light dash or something.

201 00:22:03.150 ⇒ 00:22:05.940 Uttam Kumaran: it’s it’s like we have to make a decision. Ultimately.

202 00:22:05.950 ⇒ 00:22:07.480 Uttam Kumaran: what tool we use by.

203 00:22:08.340 ⇒ 00:22:14.220 Robert Tseng: Yeah, so walk me through, like, you know. Now, you have this adopted for for some clients like, like.

204 00:22:14.270 ⇒ 00:22:18.909 Robert Tseng: yeah, like, where? How do you see your how do you see your clients using using this.

205 00:22:19.450 ⇒ 00:22:22.005 Uttam Kumaran: Yeah. So for one of our clients,

206 00:22:22.430 ⇒ 00:22:32.669 Uttam Kumaran: the main development workflow is gonna be one making sure that all the necessary dimensions they have are there, and that the values are in the way they expect.

207 00:22:32.730 ⇒ 00:22:38.700 Uttam Kumaran: and that, for example, like, if there’s values that they don’t really know like, what is a what is on hold mean?

208 00:22:38.800 ⇒ 00:22:45.339 Uttam Kumaran: I’m gonna get a question about that. However, that’s fine. We gotta go define that instead of getting a question on like

209 00:22:45.748 ⇒ 00:22:54.799 Uttam Kumaran: I need this one more column or this bar chart is not formatted right? There’s no options for some of those things here. Everything you get

210 00:22:55.448 ⇒ 00:23:01.229 Uttam Kumaran: the other thing is, sometimes this is too much right, I think, on this dashboard it’s sort of fine.

211 00:23:01.250 ⇒ 00:23:03.430 Uttam Kumaran: But, for example, if I go to like

212 00:23:03.630 ⇒ 00:23:05.080 Uttam Kumaran: another dashboard.

213 00:23:05.240 ⇒ 00:23:20.257 Uttam Kumaran: There’s a lot of dimensions. So some sometimes we may need this for internal analysis. Some of these flags, but maybe, like the they don’t care about like exactly how we’re calculating the ups charges. So some of it is just clean up. On. The second thing is

214 00:23:21.252 ⇒ 00:23:34.137 Uttam Kumaran: I need to think I need to think very clearly about how we’re gonna do, how we do onboarding for things like this. For one of our class, we just threw every we just moved everything from light dash into here, and I could tell it’s still very overwhelming for him.

215 00:23:35.165 ⇒ 00:23:35.850 Uttam Kumaran: So

216 00:23:35.880 ⇒ 00:23:38.329 Uttam Kumaran: I need to think about how do we go

217 00:23:38.460 ⇒ 00:23:43.629 Uttam Kumaran: explore from explore, and make sure that, like everything you need on the marketing side is there.

218 00:23:43.790 ⇒ 00:23:53.450 Uttam Kumaran: and and almost sit with him. But again, that’s like a work flow that we would needed to do. Anyways. However, I almost can make those changes in that meeting.

219 00:23:53.520 ⇒ 00:24:01.160 Uttam Kumaran: you know. Instead of saying, I need another day or 2 to like, make a dashboard tweak, or like, you know, when you’re doing stuff on the fly or on live.

220 00:24:01.180 ⇒ 00:24:03.350 Uttam Kumaran: It’s a it’s a nightmare.

221 00:24:03.780 ⇒ 00:24:04.400 Robert Tseng: Yeah.

222 00:24:04.710 ⇒ 00:24:12.890 Uttam Kumaran: I almost can do some of these things. Sh demo it locally and basically get approval. The other thing that’s really nice all this lives as code

223 00:24:12.910 ⇒ 00:24:15.510 Uttam Kumaran: and so selfishly for my team.

224 00:24:15.580 ⇒ 00:24:27.360 Uttam Kumaran: everything can be version controlled. Everything can be associated with a Pr and we’re gonna start to automate some of the deployments and things like that through code. So

225 00:24:27.380 ⇒ 00:24:31.069 Uttam Kumaran: I tell people that on the from the reliability standpoint.

226 00:24:31.130 ⇒ 00:24:37.059 Uttam Kumaran: it’s very good because you’re not. There’s not like logic sitting in a custom field somewhere.

227 00:24:37.620 ⇒ 00:24:51.649 Uttam Kumaran: and you’re not hitting the date. You’re not. I’m not hitting the data warehouse for any of this also, so we haven’t talked about cost yet. But I’m this is none of this is issued as queries to the warehouse, so it’s a 1 time query to run to select Star, and then you’re good. So

228 00:24:52.160 ⇒ 00:24:57.280 Uttam Kumaran: for us, for our customers, it’s it’s not like we don’t have a ton of warehouse costs, anyways. But

229 00:24:58.030 ⇒ 00:24:59.789 Uttam Kumaran: That’s that’s a big plus.

230 00:25:01.140 ⇒ 00:25:01.710 Robert Tseng: Yeah.

231 00:25:03.330 ⇒ 00:25:05.680 Uttam Kumaran: So maybe let me let me walk through evidence.

232 00:25:06.050 ⇒ 00:25:07.400 Uttam Kumaran: Sure.

233 00:25:07.970 ⇒ 00:25:10.669 Uttam Kumaran: and let’s we can take a look at that.

234 00:25:10.920 ⇒ 00:25:13.520 Uttam Kumaran: and then we could end with light dash.

235 00:25:16.460 ⇒ 00:25:19.595 Uttam Kumaran: I like this. I feel like I’m given a little seminar. This is nice.

236 00:25:20.120 ⇒ 00:25:22.219 Robert Tseng: Yeah, no, this is. This is great.

237 00:25:23.835 ⇒ 00:25:24.670 Uttam Kumaran: Ho!

238 00:25:26.040 ⇒ 00:25:29.160 Robert Tseng: I see you’re recording this so I could come back to you later, or something.

239 00:25:29.520 ⇒ 00:25:32.420 Uttam Kumaran: Yes, and so we we can. We can also come back to it later, because.

240 00:25:32.420 ⇒ 00:25:33.130 Robert Tseng: Okay. Great.

241 00:25:33.130 ⇒ 00:25:35.670 Uttam Kumaran: I included Nico. Because I actually.

242 00:25:35.680 ⇒ 00:25:40.989 Uttam Kumaran: I need to do this. And I was like, Oh, perfect like, if I’m doing a walkthrough of like all our tools.

243 00:25:40.990 ⇒ 00:25:41.720 Nicolas Sucari: Yeah.

244 00:25:41.720 ⇒ 00:25:44.680 Uttam Kumaran: This is great, so I will. You totally can ha have this.

245 00:25:44.680 ⇒ 00:25:47.569 Nicolas Sucari: Super, helpful, super helpful for me to

246 00:25:47.590 ⇒ 00:25:50.729 Nicolas Sucari: just to get to know the tool that you’re using. Yeah.

247 00:25:52.690 ⇒ 00:25:58.577 Uttam Kumaran: Cool. So this is this is a tool called evidence. Evidence

248 00:25:59.220 ⇒ 00:26:03.779 Uttam Kumaran: is a tool for building data products.

249 00:26:04.196 ⇒ 00:26:10.600 Uttam Kumaran: The team is like former people from like the New York Times data team. So like, kind of like, they’re used to building very polish

250 00:26:10.920 ⇒ 00:26:18.449 Uttam Kumaran: like data products that need to go out like digitally. And so their focus is on things looking amazing.

251 00:26:18.743 ⇒ 00:26:21.379 Uttam Kumaran: And really being able to tell a story with data.

252 00:26:21.855 ⇒ 00:26:24.300 Uttam Kumaran: Of course, like. I’m not a reporter. But

253 00:26:24.670 ⇒ 00:26:32.494 Uttam Kumaran: one thing that we do commonly is we do analysis where we’re looking at like one thing I’ll show you is we’re looking at

254 00:26:33.369 ⇒ 00:26:36.759 Uttam Kumaran: our refunds program for a client or a warranty program.

255 00:26:36.770 ⇒ 00:26:57.940 Uttam Kumaran: Right? What’s the common way we do that? Okay, we need to run. I need I need to throw somebody at it and like, Go find out like where we’re losing money and warranties, how we lose money and warranties. Follow the rabbit hole, fall down it and then put together like a, I guess a Google slides or Google Doc, that that basically is gonna explain that. But then the problem is that just dies. Basically after that meeting.

256 00:26:57.940 ⇒ 00:26:58.420 Robert Tseng: Hmm.

257 00:26:58.752 ⇒ 00:27:14.049 Uttam Kumaran: You know, and it’s usually kind of janky looking, and the last thing I want is like executives to be looking at queries or like looking at screenshots like they’re they’re gonna get lost in like that and not actually like what’s going on. And so

258 00:27:14.050 ⇒ 00:27:36.509 Uttam Kumaran: I’ve was searching and got recommended this tool and have spent some time with the team there just on. We’ve been actually like trying some local development, and they’ve been really helpful for us and so we kind of began using evidence for self hosting evidence for one of our clients, for pool parts. And I wanna walk through kind of

259 00:27:36.760 ⇒ 00:27:46.559 Uttam Kumaran: the output 1st and then I’ll walk through the development process. So this is like something we put together for their refunds and returns

260 00:27:46.680 ⇒ 00:27:57.659 Uttam Kumaran: impact analysis the other way around. Basically, my task for the engineer on this was like, I want, we want to put together an impact analysis of how refunds and returns affects their business

261 00:27:58.007 ⇒ 00:28:19.510 Uttam Kumaran: holistically, which is primarily on the cost side. But we’re kind of digging deeper on some stuff, as we speak. So the thing is like we’re able to have like headers, footers, everything’s marked down so we can have nice commentary here on here. There’s really nice helpful features about like how you could do hovers the look and styling of all this, as you can see, there’s like

262 00:28:19.951 ⇒ 00:28:43.418 Uttam Kumaran: table contents. There’s a home page. All this is customizable. Ca, all the callers and everything is really customizable. So we basically put together some simple charts. All this actually is coming is using the same duck dB process which it selects a bunch of data from our warehouse, and then they run queries on top of that

263 00:28:43.870 ⇒ 00:28:45.559 Uttam Kumaran: But again I would say.

264 00:28:45.580 ⇒ 00:29:09.439 Uttam Kumaran: out of all the tools I use like this has the potential to be the best looking, and has the most out of the box features for visualizations. So we’re not doing anything complicated here in terms of visualizations. But you could see that, you know, you can do basically everything that you expect. I’m gonna show you one more that we have some interesting visualizations. Are we have these like sparklines that are really.

265 00:29:09.440 ⇒ 00:29:10.150 Robert Tseng: Yes.

266 00:29:10.775 ⇒ 00:29:15.120 Uttam Kumaran: We have like these sort of like football field heat maps,

267 00:29:16.060 ⇒ 00:29:22.448 Uttam Kumaran: kind of like different, like individual calendar fields to show density over time.

268 00:29:23.360 ⇒ 00:29:30.620 Uttam Kumaran: all of like, you know, again, just different ways of showing data. If if we look at their actual docs

269 00:29:30.999 ⇒ 00:29:34.239 Uttam Kumaran: I can show you a little bit about all the components they have.

270 00:29:34.728 ⇒ 00:29:41.291 Uttam Kumaran: So you could do all the things like from big valleys with little small comparisons to line charts, area charts.

271 00:29:41.750 ⇒ 00:29:43.050 Uttam Kumaran: bar charts.

272 00:29:43.920 ⇒ 00:29:46.070 Uttam Kumaran: scatter plots, funnels

273 00:29:46.770 ⇒ 00:29:54.350 Uttam Kumaran: like this is great. I love this I. This is all the stuff that sometimes I really needed to show like something like this.

274 00:29:54.420 ⇒ 00:29:58.080 Uttam Kumaran: right? Or yeah, or a box plot, or like a map. And

275 00:29:58.470 ⇒ 00:30:05.300 Uttam Kumaran: the the functionality is either not there or you have to like, do some sort of thing in like Javascript to get it to work

276 00:30:05.560 ⇒ 00:30:21.640 Uttam Kumaran: like I don’t. I’m not. I’m not on that sort of energy right now. I just, I need the stuff to be there. Additionally, they have a lot of great filters and and components to actually affect the data. So things like dropdowns, text inputs. Date ranges all out of the box. All look

277 00:30:21.680 ⇒ 00:30:23.532 Uttam Kumaran: very, very beautiful.

278 00:30:24.920 ⇒ 00:30:49.120 Uttam Kumaran: so. And then the last thing I’ll kind of mention is, this is all as code as well. So the nice thing is is like, for example, if we sorry if we wanted to show like this 5 times and just add one specific filter to it. We can actually write a for loop to do that in the code. So because all the visualization sits as code. You can do loops

279 00:30:49.885 ⇒ 00:30:51.640 Uttam Kumaran: if else filters.

280 00:30:51.680 ⇒ 00:30:56.450 Uttam Kumaran: And you could actually show different visualizations depending on different parameters.

281 00:30:56.510 ⇒ 00:31:02.100 Uttam Kumaran: So I don’t have, like some great concrete examples of that, because we’re just kinda like getting started with it. But

282 00:31:02.370 ⇒ 00:31:14.400 Uttam Kumaran: you could do a lot of things. The the development workflow for evidence is a bit more complicated, though. One is, they have a cloud hosted version. And they have, like a self hosted version, where

283 00:31:14.440 ⇒ 00:31:19.020 Uttam Kumaran: using the cell phone version, basically, because it was like

284 00:31:19.310 ⇒ 00:31:38.829 Uttam Kumaran: it. I just wanted to start. I didn’t want to talk to any salespeople. I just wanted to like run through it. Their Cloud host version is a lot easier to set up within the way we did. However, the development workflow will be basically the same. You have like connection strings to all your stuff. And then we kind of have the same thing where we have like

285 00:31:38.840 ⇒ 00:31:39.790 Uttam Kumaran: sources.

286 00:31:41.100 ⇒ 00:31:44.344 Uttam Kumaran: and then queries. So let’s start with like

287 00:31:44.980 ⇒ 00:31:49.370 Uttam Kumaran: all order items or refunds like these are all again, just selects from

288 00:31:49.710 ⇒ 00:32:00.729 Uttam Kumaran: our Dbt schemas within up queries, which is like, for example. Let’s look at like days. Let’s look. Let’s just look at one example here. So if we look at refunds.

289 00:32:01.155 ⇒ 00:32:03.410 Uttam Kumaran: let’s look at days to return.

290 00:32:03.480 ⇒ 00:32:10.350 Uttam Kumaran: So days to return is likely pulling from the States return sequel. It’s like select order, date, average days to return.

291 00:32:10.420 ⇒ 00:32:13.610 Uttam Kumaran: So basically, what we’re looking at is we’re looking at. If the days

292 00:32:13.920 ⇒ 00:32:15.300 Uttam Kumaran: to refund

293 00:32:15.580 ⇒ 00:32:21.619 Uttam Kumaran: changes over time like, is it getting longer or shorter over time? This is the query that we write

294 00:32:21.710 ⇒ 00:32:32.040 Uttam Kumaran: if we look at the pages. So everything is as Markdown. So if we go to the refunds page, you can see here that this is the code that actually forms this.

295 00:32:33.030 ⇒ 00:32:42.290 Uttam Kumaran: So you have a title. You have all the queries that you’re gonna be using almost like an import statement. And then each of the charts. Each of the charts pulls from these sequel statements.

296 00:32:42.300 ⇒ 00:32:44.999 Uttam Kumaran: So we have our days from to refund. Which is just this.

297 00:32:45.050 ⇒ 00:32:48.409 Uttam Kumaran: This is a simple statement we’re selecting from days to refund.

298 00:32:48.520 ⇒ 00:32:54.639 Uttam Kumaran: It actually gets used. Down here. Here’s where you have all the configuration for

299 00:32:54.740 ⇒ 00:32:55.810 Uttam Kumaran: the text.

300 00:32:55.950 ⇒ 00:32:57.080 Uttam Kumaran: The charts.

301 00:32:57.240 ⇒ 00:33:01.550 Uttam Kumaran: if we look at days to refund here. It’s a line chart.

302 00:33:01.730 ⇒ 00:33:05.029 Uttam Kumaran: The X axis is order date. The Y axis is this

303 00:33:06.480 ⇒ 00:33:14.399 Uttam Kumaran: There’s a reference line at 19.2, which is the average. It looks like it’s hard coded. And then the label is average.

304 00:33:15.800 ⇒ 00:33:24.209 Uttam Kumaran: the good thing is. And it already went through is that I think it looks great, for, like a tool that does analysis like.

305 00:33:24.360 ⇒ 00:33:34.040 Uttam Kumaran: and for it not being custom react components which you can do in here. It looks great. This is something that an analyst did on our team

306 00:33:34.140 ⇒ 00:33:41.270 Uttam Kumaran: who doesn’t have any front end back end, any experience was able to set this up. The difficult part is.

307 00:33:41.410 ⇒ 00:33:55.530 Uttam Kumaran: it’s like marked down this type code. So the problem is, this is not a great tool for exploratory analysis, this is not a tool you’re gonna go into as a engineer or a data person, and

308 00:33:55.690 ⇒ 00:34:08.308 Uttam Kumaran: like, ask 10 questions in a row. And like figure things out. This is this is this is really, really the last mile for an for my mind, for analysis and and potentially dashboards.

309 00:34:09.010 ⇒ 00:34:24.050 Uttam Kumaran: where you really really wanna convey a story about an analysis, or you want a dashboard that you want people to go look at even for for us. I saw it as a perfect replacement for doing analysis. Anytime we do an analysis.

310 00:34:24.250 ⇒ 00:34:25.430 Uttam Kumaran: I either.

311 00:34:25.440 ⇒ 00:34:32.059 Uttam Kumaran: I I want it to end up in here. I don’t want it to die as a Google, Doc or or a Google slide.

312 00:34:32.179 ⇒ 00:34:44.669 Uttam Kumaran: However, I also know that, like doing, we realize that doing the analysis in here is really really difficult. So I I kind of talked to. You know, the Jacob or my team, who kind of does analysis is I was like.

313 00:34:44.790 ⇒ 00:34:56.260 Uttam Kumaran: I want you to move as fast as you need to, either in excel or in snowflake running queries and finding answers. But when we do the presentation, it’s important that we consider that

314 00:34:56.639 ⇒ 00:35:05.120 Uttam Kumaran: an email with just a bunch of numbers, or like a Google slide with a bunch of stuff is not a great thing. I want the presentation to happen within here.

315 00:35:05.517 ⇒ 00:35:13.199 Uttam Kumaran: And so you will have to have that balance. And he gave the feedback, though, like, Hey doing? This was hard when I didn’t know exactly what I wanted to show.

316 00:35:13.440 ⇒ 00:35:17.739 Uttam Kumaran: But if you know exactly like I want these line charts, I want this sort of view.

317 00:35:17.950 ⇒ 00:35:19.560 Uttam Kumaran: This is a great place to do that.

318 00:35:22.060 ⇒ 00:35:24.579 Uttam Kumaran: so I’ll stop there. Let me know what you think.

319 00:35:25.270 ⇒ 00:35:43.002 Robert Tseng: Yeah. To me, this is like a notebook replacement. Normally, I would be doing this in like Cooper notebooks. Right? I think that’s the most. Yeah. I mean, I would be running everything like this. And in Python the same Markdown style I mean a lot of the components there seem like they’re already there to predefine for you, which is great.

320 00:35:43.290 ⇒ 00:36:03.719 Robert Tseng: I mean, there are like, I mean there. There are python libraries where all these components are somewhat defined as well. So but I think the advantage here is just the the ease of think of it. Yeah. The thinking of the data to your to the like by pulling in the queries very quickly. Because all in the same environment, rather than like having to

321 00:36:04.024 ⇒ 00:36:15.900 Robert Tseng: like, do like export sequel to to python connections is kind of annoying for me when I have to do something like this in in notebook format. But other than that. Yeah, I think totally makes sense. So.

322 00:36:15.960 ⇒ 00:36:24.739 Robert Tseng: yeah, if I wanted to produce a deliverable like this, like as I hope I would typically be using something something like a Jupiter. Or

323 00:36:25.560 ⇒ 00:36:36.920 Robert Tseng: yeah, if it was going to be a more slide based like, Excel charting kind of thing, then, yeah, all those queries that you’re that you’ve described like they’d just be in different tabs on a spreadsheet. And then you’d be, you know.

324 00:36:37.020 ⇒ 00:37:05.310 Robert Tseng: building these building these charts. Using the hunk functionality of of the ui and a tool like that, and then you would be paced to get into like a into a death. So I like that. This is all all in the same environment. But yeah, I I could see that there’s a alerting curve to to to get to that. I don’t really know how you iterate on building all all these components out. Yeah, without knowing exactly what you want, which I think to me that’d probably be like the biggest challenge. Yeah.

325 00:37:06.420 ⇒ 00:37:08.630 Uttam Kumaran: Yeah, I totally agree. You know, I

326 00:37:08.960 ⇒ 00:37:10.689 Uttam Kumaran: in an ideal world.

327 00:37:11.000 ⇒ 00:37:21.379 Uttam Kumaran: and I think, like we should, we could even be good to use this as dashboards. The problem is, again, people need to know how this world works.

328 00:37:21.900 ⇒ 00:37:22.240 Robert Tseng: Yeah.

329 00:37:22.240 ⇒ 00:37:29.639 Uttam Kumaran: That’s tough. This is, we’re actually like having a drag and drop means like anybody can go build these. However.

330 00:37:29.880 ⇒ 00:37:35.998 Uttam Kumaran: like, I don’t think this is, I mean, I don’t know. This might be. I don’t think this is that complicated.

331 00:37:36.710 ⇒ 00:37:51.740 Uttam Kumaran: but you’re right in that. You do need to know what you want. Like you’re, I’m not a visual person. So unless I see what this is, I can’t do this right. And I’ve also found it hard to find people that are like data, this people.

332 00:37:52.310 ⇒ 00:38:03.230 Uttam Kumaran: So commonly, you have analysts. We have engineers. There’s no like, there’s not great like data vis focused people, or what they or they’re working like on like info graphs and stuff

333 00:38:03.430 ⇒ 00:38:07.090 Uttam Kumaran: right? Like, kinda like like marketing, or like

334 00:38:07.120 ⇒ 00:38:16.942 Uttam Kumaran: some sort of activation. So that’s the thing I think a lot about is like the and I’ve told this to the evidence team as like the learning curve is high here.

335 00:38:17.480 ⇒ 00:38:18.490 Uttam Kumaran: but

336 00:38:19.040 ⇒ 00:38:24.525 Uttam Kumaran: I think, like the product looks really good, and it almost like a notion. Look and feel

337 00:38:24.830 ⇒ 00:38:25.490 Robert Tseng: Yeah.

338 00:38:25.490 ⇒ 00:38:28.230 Uttam Kumaran: And again to have something like this.

339 00:38:28.600 ⇒ 00:38:33.030 Uttam Kumaran: You can’t. I don’t know what tool I would do this in.

340 00:38:33.550 ⇒ 00:38:34.929 Uttam Kumaran: Yeah, wait. Can you? Can you show me what.

341 00:38:34.930 ⇒ 00:38:37.830 Robert Tseng: Component looks like for that one specifically looks like in code.

342 00:38:38.310 ⇒ 00:38:46.070 Uttam Kumaran: Yeah. So let me. Then I see all these examples live here as code. I think if I go to, let’s just see what it uses.

343 00:38:48.720 ⇒ 00:38:50.379 Uttam Kumaran: okay, let me actually go

344 00:38:50.690 ⇒ 00:38:52.969 Uttam Kumaran: figure out where, in the example, it is

345 00:39:02.190 ⇒ 00:39:04.760 Uttam Kumaran: so average order value and reference areas.

346 00:39:05.470 ⇒ 00:39:06.149 Robert Tseng: Hmm,

347 00:39:13.990 ⇒ 00:39:15.592 Uttam Kumaran: And then I think it’s using

348 00:39:19.880 ⇒ 00:39:21.020 Uttam Kumaran: me? Look at.

349 00:39:22.010 ⇒ 00:39:24.959 Uttam Kumaran: I think I saw the reference areas somewhere here.

350 00:39:27.320 ⇒ 00:39:28.140 Uttam Kumaran: Yeah.

351 00:39:32.320 ⇒ 00:39:33.740 Uttam Kumaran: like, this is great.

352 00:39:35.440 ⇒ 00:39:39.549 Uttam Kumaran: I didn’t like I I didn’t know I could do something like this, right?

353 00:39:39.820 ⇒ 00:39:40.485 Uttam Kumaran: Yeah,

354 00:39:42.200 ⇒ 00:39:43.510 Uttam Kumaran: and

355 00:39:45.370 ⇒ 00:39:48.810 Uttam Kumaran: I’m glad this tool I’m really glad this tool exists. I

356 00:39:49.370 ⇒ 00:39:58.860 Uttam Kumaran: I am going to start making sure that anybody that does analysis any sort of analysis piece that comes out of our org that’s finalized is gonna come through this I’ve been very, very clear.

357 00:39:59.270 ⇒ 00:40:03.650 Uttam Kumaran: I get pushback that some people are like, Oh, table is very nice or blah blah blah.

358 00:40:04.030 ⇒ 00:40:07.780 Uttam Kumaran: I really care about how things look, especially

359 00:40:08.100 ⇒ 00:40:11.470 Uttam Kumaran: when it comes to analysis that calmly looks so bad

360 00:40:11.670 ⇒ 00:40:12.620 Uttam Kumaran: that

361 00:40:12.930 ⇒ 00:40:21.089 Uttam Kumaran: this looking good is gonna make us look really good. And so, even though the data is just data, right? I think the presentation layer really matters

362 00:40:21.631 ⇒ 00:40:24.159 Uttam Kumaran: ideally. I would use this for dashboards, too. But

363 00:40:25.010 ⇒ 00:40:27.529 Uttam Kumaran: maybe the workflow is like.

364 00:40:27.870 ⇒ 00:40:31.230 Uttam Kumaran: do it in light dash until it’s like, doesn’t move.

365 00:40:31.290 ⇒ 00:40:33.410 Uttam Kumaran: and then bring it to here

366 00:40:33.910 ⇒ 00:40:35.975 Uttam Kumaran: because there’s so many more options. But

367 00:40:36.590 ⇒ 00:40:41.720 Uttam Kumaran: maybe it’s like, if you but some people also give me very, very specific like this, this, this, this.

368 00:40:42.380 ⇒ 00:40:46.750 Uttam Kumaran: and maybe it’s like cool. We’ll build it in. We’ll just build it in evidence, because

369 00:40:46.820 ⇒ 00:40:51.460 Uttam Kumaran: actually, we’re probably limited by a tool like light dash, and that we can’t do certain things.

370 00:40:51.640 ⇒ 00:40:52.690 Uttam Kumaran: I mean.

371 00:40:52.900 ⇒ 00:40:57.640 Uttam Kumaran: And the last thing is you can do all you can. You can bring in custom. You can bring in echard stuff.

372 00:40:57.975 ⇒ 00:41:03.190 Uttam Kumaran: You can bring in custom components, and they have, like other components that they’re working on

373 00:41:03.790 ⇒ 00:41:04.510 Uttam Kumaran: like.

374 00:41:04.510 ⇒ 00:41:06.449 Robert Tseng: This is open source. No, it’s not right.

375 00:41:07.280 ⇒ 00:41:08.420 Uttam Kumaran: This is.

376 00:41:08.820 ⇒ 00:41:11.199 Uttam Kumaran: This is open source, and you can self host it.

377 00:41:11.990 ⇒ 00:41:12.690 Uttam Kumaran: Huh!

378 00:41:15.200 ⇒ 00:41:19.780 Uttam Kumaran: And I could walk you through. We have this host. This one is hosted for 15 bucks

379 00:41:20.220 ⇒ 00:41:21.999 Uttam Kumaran: a month for us right now.

380 00:41:22.560 ⇒ 00:41:25.360 Uttam Kumaran: which is like, I’m paid. I just paid for it. Cause I’m like.

381 00:41:25.940 ⇒ 00:41:29.740 Uttam Kumaran: I wanna, I just I like, I don’t even wanna have. It’s not even a question for me

382 00:41:29.870 ⇒ 00:41:32.469 Uttam Kumaran: that this is something that I want to put in front of people.

383 00:41:32.840 ⇒ 00:41:33.490 Robert Tseng: Yeah.

384 00:41:34.040 ⇒ 00:41:35.160 Uttam Kumaran: And so

385 00:41:36.340 ⇒ 00:41:36.720 Uttam Kumaran: yeah.

386 00:41:37.830 ⇒ 00:41:41.460 Uttam Kumaran: so maybe the last thing I’ll I’ll walk through is I’ll walk through

387 00:41:42.080 ⇒ 00:41:43.549 Uttam Kumaran: I walk through light, dash

388 00:41:47.080 ⇒ 00:41:48.229 Uttam Kumaran: me one sec.

389 00:42:04.390 ⇒ 00:42:05.060 Uttam Kumaran: Look.

390 00:42:07.170 ⇒ 00:42:10.976 Uttam Kumaran: So light dash is, gonna be more of your

391 00:42:12.140 ⇒ 00:42:13.540 Uttam Kumaran: traditional

392 00:42:14.020 ⇒ 00:42:14.860 Uttam Kumaran: like

393 00:42:15.790 ⇒ 00:42:17.950 Uttam Kumaran: workflow for developing dashboards.

394 00:42:18.030 ⇒ 00:42:23.340 Uttam Kumaran: Light dash is my choice with light dash. One was the cost.

395 00:42:23.980 ⇒ 00:42:25.650 Uttam Kumaran: It’s pretty cheap

396 00:42:25.760 ⇒ 00:42:34.150 Uttam Kumaran: compared to looker, which is like it’s just basically a clone of looker. I don’t know whether they they pitch. I don’t know whether they want me to be or whatever, but like

397 00:42:34.310 ⇒ 00:42:40.519 Uttam Kumaran: that’s what it is to be. The second thing is, it sits very closely with dbt, meaning

398 00:42:40.560 ⇒ 00:42:42.840 Uttam Kumaran: you actually specify

399 00:42:42.870 ⇒ 00:42:44.040 Uttam Kumaran: light dash

400 00:42:44.401 ⇒ 00:42:50.720 Uttam Kumaran: models right next to where your Dbt code is, which makes it really nice, because I’m not also

401 00:42:50.890 ⇒ 00:42:53.550 Uttam Kumaran: going into a tool and writing a bunch of code.

402 00:42:54.306 ⇒ 00:42:57.080 Uttam Kumaran: Doing all the coding in one place. So

403 00:42:57.260 ⇒ 00:43:10.759 Uttam Kumaran: the theme you should gather about all these things is like all these are code in somewhat manner, shape, or form, because the way I can assign tickets to the team into myself, the way I can make sure things are reliably deployed

404 00:43:11.390 ⇒ 00:43:13.070 Uttam Kumaran: requires it to be S. Code.

405 00:43:13.240 ⇒ 00:43:26.229 Uttam Kumaran: and we’ll see some of the issues with some of this not being as code. What I want through. So dbt, again, you have like a Dbt project. We have models. So let’s look at like this, for example. So this is coming out of our Mars.

406 00:43:26.250 ⇒ 00:43:27.538 Uttam Kumaran: It’s coming out of

407 00:43:29.903 ⇒ 00:43:31.449 Uttam Kumaran: Daily Kpi.

408 00:43:32.800 ⇒ 00:43:36.110 Uttam Kumaran: I think this is all coming from like this daily Kpi ag, table.

409 00:43:36.760 ⇒ 00:43:40.360 Uttam Kumaran: Basically, this aggregates a bunch of kpis to the daily

410 00:43:41.350 ⇒ 00:43:43.547 Uttam Kumaran: like timeframe.

411 00:43:44.310 ⇒ 00:43:49.419 Uttam Kumaran: So we have our table specified here, and it’s all dbt, and then we also have a yaml file next to it.

412 00:43:49.708 ⇒ 00:43:57.979 Uttam Kumaran: Dbt will light desk actually help you generate this, and then you can go tweak it. But what we have here is we just have the name, the labels Daily Kpis.

413 00:43:58.120 ⇒ 00:44:01.600 Uttam Kumaran: and you’ll see this actually exists here.

414 00:44:02.660 ⇒ 00:44:07.590 Uttam Kumaran: So if we go here and you go to kpis, you’ll see daily. Kpis.

415 00:44:07.850 ⇒ 00:44:10.439 Uttam Kumaran: this is what is pulling from here.

416 00:44:10.650 ⇒ 00:44:14.729 Uttam Kumaran: All this Yaml is is a specification on top of the table.

417 00:44:14.760 ⇒ 00:44:17.390 Uttam Kumaran: So you basically specifying again dimensions.

418 00:44:18.218 ⇒ 00:44:20.461 Uttam Kumaran: Like. Let’s take, for example.

419 00:44:23.015 ⇒ 00:44:23.780 Uttam Kumaran: The

420 00:44:23.910 ⇒ 00:44:29.390 Uttam Kumaran: average same day week orders. This is a dimension. It’s a number I wanted to group under sales.

421 00:44:29.953 ⇒ 00:44:33.390 Uttam Kumaran: And I want to round it to always be thousands.

422 00:44:33.450 ⇒ 00:44:36.349 Uttam Kumaran: So you can specify those types of visual options here.

423 00:44:36.722 ⇒ 00:44:39.807 Uttam Kumaran: And then you also have you also have

424 00:44:40.460 ⇒ 00:44:42.440 Uttam Kumaran: measures here at the bottom.

425 00:44:43.680 ⇒ 00:44:49.280 Uttam Kumaran: So, for example, I have total fees, is a dimension. I also have total fees as a measure.

426 00:44:49.711 ⇒ 00:44:55.850 Uttam Kumaran: That way. I can aggregate to it. So you’ll see a total fees is also here.

427 00:44:58.220 ⇒ 00:45:03.625 Uttam Kumaran: this is a little bit to do with the data model itself. But basically, all you do is you can select, like,

428 00:45:04.350 ⇒ 00:45:05.790 Uttam Kumaran: plug the day.

429 00:45:06.320 ⇒ 00:45:08.299 Uttam Kumaran: Select some of these

430 00:45:08.370 ⇒ 00:45:12.060 Uttam Kumaran: hit run. It’s gonna issue a query to the warehouse.

431 00:45:12.810 ⇒ 00:45:14.800 Uttam Kumaran: Here’s a query down here.

432 00:45:16.540 ⇒ 00:45:19.400 Uttam Kumaran: again, if you use looker, this should be pretty familiar.

433 00:45:19.849 ⇒ 00:45:24.640 Uttam Kumaran: Basically, you have everything here. These are formatted as we set up the formats here.

434 00:45:25.040 ⇒ 00:45:26.350 Uttam Kumaran: Thousands.

435 00:45:28.009 ⇒ 00:45:29.929 Uttam Kumaran: you can filter.

436 00:45:30.470 ⇒ 00:45:36.900 Uttam Kumaran: They have a this is new. They just added the ability to do like. I guess quick calculations can add a table calculation.

437 00:45:38.750 ⇒ 00:45:41.000 Uttam Kumaran: they have the charting functionality here.

438 00:45:41.230 ⇒ 00:45:44.299 Uttam Kumaran: So if we were to say cool, I just wanna look at

439 00:45:44.980 ⇒ 00:45:46.610 Uttam Kumaran: the is

440 00:45:46.830 ⇒ 00:45:48.380 Uttam Kumaran: in the last

441 00:45:48.740 ⇒ 00:45:50.340 Uttam Kumaran: like 10

442 00:45:52.000 ⇒ 00:45:54.080 Uttam Kumaran: weeks I could hit run

443 00:45:54.480 ⇒ 00:45:56.329 Uttam Kumaran: issues occurred in the warehouse.

444 00:45:57.290 ⇒ 00:46:00.529 Uttam Kumaran: Looks like something like this. Okay, I want to say, great. I

445 00:46:00.860 ⇒ 00:46:04.539 Uttam Kumaran: want this to be there. I wanna get rid of a total clicks.

446 00:46:04.810 ⇒ 00:46:10.279 Uttam Kumaran: and then you can make modifications so like have dual axes, different color series.

447 00:46:11.860 ⇒ 00:46:13.749 Uttam Kumaran: you know the usual

448 00:46:15.384 ⇒ 00:46:22.780 Uttam Kumaran: but this is kind of like as good as it looks. You could add like labels and stuff. But again, there’s like sometimes overlap issues.

449 00:46:22.820 ⇒ 00:46:24.799 Uttam Kumaran: And then basically, what you do is you can

450 00:46:24.950 ⇒ 00:46:28.313 Uttam Kumaran: take it from here. You can save the chart.

451 00:46:29.100 ⇒ 00:46:36.540 Uttam Kumaran: and you could save it directly to a dashboard or to a space, and a space can have either charts or dashboards.

452 00:46:36.740 ⇒ 00:46:43.636 Uttam Kumaran: An example of a dashboard that we were just looking at is vital signs. This is like our clients,

453 00:46:44.110 ⇒ 00:46:46.279 Uttam Kumaran: like daily dashboard that they look at.

454 00:46:46.330 ⇒ 00:46:48.810 Uttam Kumaran: Basically they log in here and just look at like

455 00:46:49.160 ⇒ 00:46:51.670 Uttam Kumaran: profit sales, shipping costs.

456 00:46:53.600 ⇒ 00:46:57.360 Uttam Kumaran: you could add, mark down the way this works is like you could hit edit.

457 00:46:57.640 ⇒ 00:47:00.139 Uttam Kumaran: Each of these like you can resize.

458 00:47:00.150 ⇒ 00:47:04.110 Uttam Kumaran: You can go in here and like duplicate or edit the content of the tile.

459 00:47:04.360 ⇒ 00:47:06.940 Uttam Kumaran: Add filters for the dashboard.

460 00:47:07.640 ⇒ 00:47:09.160 Uttam Kumaran: You can add tiles

461 00:47:09.680 ⇒ 00:47:12.029 Uttam Kumaran: if you want to go. Edit a specific chart.

462 00:47:12.390 ⇒ 00:47:15.780 Uttam Kumaran: you would go in here. You can click edit chart from here.

463 00:47:17.510 ⇒ 00:47:20.760 Uttam Kumaran: make whatever change you need to hit. Save

464 00:47:23.070 ⇒ 00:47:25.560 Uttam Kumaran: That’s basically the the gist

465 00:47:28.360 ⇒ 00:47:29.929 Uttam Kumaran: kind of pros and cons.

466 00:47:30.040 ⇒ 00:47:35.339 Uttam Kumaran: I mean, you cut. Probably guess what I’m gonna say is that like, I don’t think it looks particularly great

467 00:47:37.200 ⇒ 00:47:40.870 Uttam Kumaran: it’s this dashboard. I’ve taken our team

468 00:47:40.900 ⇒ 00:47:42.770 Uttam Kumaran: so long to

469 00:47:44.120 ⇒ 00:47:47.899 Uttam Kumaran: configure this way. Get feedback, make adjustments.

470 00:47:48.829 ⇒ 00:48:03.119 Uttam Kumaran: That like I I basically told the client that like I wasn’t gonna find I have to think of as an alternative, because it’s not a good use of your money. I just don’t think we’re making progress on like the real things. I can’t. I don’t want someone really just to be like making

471 00:48:03.310 ⇒ 00:48:06.100 Uttam Kumaran: like small, like visual tweaks.

472 00:48:06.410 ⇒ 00:48:09.750 Uttam Kumaran: They have some good visualizations like be able to comparisons. But

473 00:48:09.760 ⇒ 00:48:14.700 Uttam Kumaran: again, if I just showed you evidence, you’re gonna be like, yeah, this is like table stakes.

474 00:48:17.420 ⇒ 00:48:21.189 Uttam Kumaran: but it’s 600 bucks. There’s no additional user cost.

475 00:48:21.210 ⇒ 00:48:25.590 Uttam Kumaran: They also offer the ability to embed this in your product. If that’s like of interest, any client

476 00:48:27.780 ⇒ 00:48:34.169 Uttam Kumaran: and it’s a traditional dashboarding experience. So for folks that are more used to something like this.

477 00:48:34.710 ⇒ 00:48:37.410 Uttam Kumaran: or if a team has like a dashboard or

478 00:48:37.975 ⇒ 00:48:39.929 Uttam Kumaran: this is kind of like a

479 00:48:40.470 ⇒ 00:48:42.039 Uttam Kumaran: suspect. I save that.

480 00:48:42.836 ⇒ 00:48:43.289 Uttam Kumaran: I th-

481 00:48:44.510 ⇒ 00:48:45.350 Uttam Kumaran: so yeah.

482 00:48:45.350 ⇒ 00:48:46.070 Robert Tseng: Thoroughly.

483 00:48:46.620 ⇒ 00:49:00.460 Robert Tseng: Yeah, I mean, this is the closest to what I’m used to. Right. This is kind of the process that I’ve begun in metabase with with stone. Right? It’s like kind of bring some of the initial tiles out and pulling the requirements in. And then it’s just like iterate iteratively, like

484 00:49:00.870 ⇒ 00:49:22.189 Robert Tseng: changing, making a small visual adjustments, adding new charts in pieces as they’re able to see it. It’s it is very much like a build as you go way of approaching Vi. It’s like you look at it. And as as a stakeholders looking at it, and then trying to talk through like how they would be using it. And you as a as the bi developer hypothetically would be

485 00:49:22.190 ⇒ 00:49:48.206 Robert Tseng: gathering that feedback trying to like, you know, rearrange things like, kind of yeah, we decided every time you you go through it. And yeah, I could see that. You know, it’s it’s it’s yeah. You don’t have everything, all, all, all there and then. It’s it’s less like, yeah, it’s like building from scratch rather than like having everything there and then cutting it down to something that’s that’s like more, I I guess, cutting it down to what they want. So I think this is like 2 different ways of approaching it.

486 00:49:48.600 ⇒ 00:50:04.959 Robert Tseng: yeah, I mean visually, yeah, I mean to be. This is pretty standard. Yeah, it’s not beautiful, like what you describe. But yeah, I mean to me. This is like what people expect things living in tiles. You can move the tiles around. Yeah, like this. This is. This is the most familiar to me. So.

487 00:50:04.960 ⇒ 00:50:08.180 Uttam Kumaran: I think the other thing that like I don’t know if you’ve had success on. But

488 00:50:08.730 ⇒ 00:50:11.530 Uttam Kumaran: even in my like all the companies I’ve worked at.

489 00:50:11.920 ⇒ 00:50:20.529 Uttam Kumaran: They. They say these tools are like self. Serve Bi. But I still have so many problems getting people to go like

490 00:50:20.660 ⇒ 00:50:23.729 Uttam Kumaran: from here. It’s like, oh, what table did I pick?

491 00:50:25.313 ⇒ 00:50:27.530 Uttam Kumaran: Oh, it’s in here! Great!

492 00:50:27.580 ⇒ 00:50:28.830 Uttam Kumaran: What’s what

493 00:50:29.010 ⇒ 00:50:32.070 Uttam Kumaran: right and like, if I it’s just like so much

494 00:50:32.140 ⇒ 00:50:32.899 Uttam Kumaran: that

495 00:50:34.020 ⇒ 00:50:37.220 Uttam Kumaran: I, after doing so many of these. And look around here.

496 00:50:37.550 ⇒ 00:50:38.790 Uttam Kumaran: I’m and like.

497 00:50:38.890 ⇒ 00:50:43.829 Uttam Kumaran: Look, maybe I’m just like going around it the wrong way. But I’m still not convinced that, like

498 00:50:44.620 ⇒ 00:50:47.239 Uttam Kumaran: you need to have a very, very sophisticated

499 00:50:47.400 ⇒ 00:50:50.869 Uttam Kumaran: executive or stakeholder that can come in here and use this.

500 00:50:50.970 ⇒ 00:50:58.349 Uttam Kumaran: Still, to this day I don’t think the client has done anything and explore. I think purely they’re going in and looking at dashboards

501 00:50:58.740 ⇒ 00:50:59.359 Uttam Kumaran: right?

502 00:51:00.630 ⇒ 00:51:01.650 Uttam Kumaran: So

503 00:51:02.280 ⇒ 00:51:03.648 Uttam Kumaran: I don’t know. My

504 00:51:04.250 ⇒ 00:51:07.439 Uttam Kumaran: that’s concerning I don’t like that, like, I want people to

505 00:51:07.840 ⇒ 00:51:14.160 Uttam Kumaran: think about things and be able to go access things beyond us, because then we’re always the blocker.

506 00:51:14.170 ⇒ 00:51:17.670 Uttam Kumaran: you know, and it. Like, I, I don’t worry about like.

507 00:51:17.720 ⇒ 00:51:24.660 Uttam Kumaran: Oh, yeah, like a quote. We have so many other things that we could be doing for them, that helping them like, add a dimension, or like

508 00:51:24.970 ⇒ 00:51:31.350 Uttam Kumaran: like telling where this thing is, or adding a tile is like, not what we need to be spending our time on so

509 00:51:31.360 ⇒ 00:51:32.630 Uttam Kumaran: frankly like

510 00:51:33.030 ⇒ 00:51:35.469 Uttam Kumaran: after discovering evidence and real.

511 00:51:36.000 ⇒ 00:51:41.469 Uttam Kumaran: I’m getting very, very close to making a call about ripping this out.

512 00:51:41.957 ⇒ 00:51:47.789 Uttam Kumaran: I need to think carefully about the way we do it, and make sure that whatever we go to

513 00:51:48.160 ⇒ 00:51:50.720 Uttam Kumaran: is like working really, really well, because I.

514 00:51:50.720 ⇒ 00:51:51.290 Robert Tseng: Yeah.

515 00:51:51.290 ⇒ 00:51:54.679 Uttam Kumaran: Doing these migrations. But

516 00:51:55.290 ⇒ 00:51:57.009 Uttam Kumaran: I don’t know. I’m I just like.

517 00:51:57.150 ⇒ 00:51:58.960 Uttam Kumaran: don’t think this is like

518 00:51:59.890 ⇒ 00:52:02.740 Uttam Kumaran: the like end state. However.

519 00:52:03.080 ⇒ 00:52:05.270 Uttam Kumaran: it works and

520 00:52:05.540 ⇒ 00:52:11.090 Uttam Kumaran: if you’re doing a point solution about something, or it’s like cool people just want a dashboard for a meeting.

521 00:52:11.550 ⇒ 00:52:18.729 Uttam Kumaran: It’s it’s like pretty self explanatory, and it’s easy for developers to kind of run through this sort of process.

522 00:52:19.490 ⇒ 00:52:20.100 Robert Tseng: Yeah.

523 00:52:21.150 ⇒ 00:52:22.040 Robert Tseng: got it?

524 00:52:24.470 ⇒ 00:52:27.711 Uttam Kumaran: Database is probably like a chea. It’s just a cheaper version of this.

525 00:52:28.210 ⇒ 00:52:30.269 Robert Tseng: Yeah, it’s very similar. Yeah.

526 00:52:30.270 ⇒ 00:52:30.760 Uttam Kumaran: I don’t.

527 00:52:30.760 ⇒ 00:52:43.784 Robert Tseng: I hear you. I agree with you. This isn’t really yeah. It’s not from what I, even what we yeah, I’ve been like a big looker guy for for years. Because, like the 2 previous orgs I was working for, like, I set up looking for them. And

528 00:52:44.310 ⇒ 00:52:57.670 Robert Tseng: yeah, yeah, I still ended up, yeah, only the technical users like it would be the engineers, or like technical Pms that would go in and like build things even when they did that it was inconsistent because they weren’t the ones that set up the data model. So maybe.

529 00:52:57.670 ⇒ 00:52:58.069 Uttam Kumaran: Yeah, has it?

530 00:52:58.070 ⇒ 00:53:06.350 Robert Tseng: Metrics. You seem like overlap. And so like governance over this like looker environment was like, you know, a good chunk of of my role on like

531 00:53:06.390 ⇒ 00:53:30.629 Robert Tseng: and to the point where, in the second order I set it up at, I was like, Okay, well, we’re just gonna like limit access. Most people are just viewers. And it’s just gonna be like a select few people that are gonna go in and develop right? So yeah, which kind of defeats the purpose of like even having the explore function if you’re not really even letting people explore. So I totally hear you on that. No, I think it’s really impressive that you’re going in and like looking for new workflows like this is.

532 00:53:30.750 ⇒ 00:53:36.600 Robert Tseng: yeah. I had no idea that, like businesses, code was like kind of at this point, you know, I.

533 00:53:36.600 ⇒ 00:53:37.000 Uttam Kumaran: Yeah.

534 00:53:37.000 ⇒ 00:53:56.540 Robert Tseng: Hear me as I’m like talking through giving my my live reactions. I’m like connecting things to like the tools that I typically use. So having everything on one like code based environment, not having to jump between sequel python and your code like, I think that’s that I I totally okay, i i i more bought in that. This is like

535 00:53:56.740 ⇒ 00:53:58.917 Robert Tseng: where things could be headed. So

536 00:53:59.280 ⇒ 00:54:01.180 Uttam Kumaran: I also think about like.

537 00:54:01.430 ⇒ 00:54:03.160 Uttam Kumaran: I also think about how like

538 00:54:03.960 ⇒ 00:54:07.920 Uttam Kumaran: this is actually primarily where our work gets consumed.

539 00:54:08.760 ⇒ 00:54:19.420 Uttam Kumaran: And if it doesn’t look great. Then it’s like, unfortunately, the human like that’s our work is not going to be like appreciated right? And although I think there’s ton of work that goes on here.

540 00:54:20.131 ⇒ 00:54:31.249 Uttam Kumaran: I want the final bow to be wrapped, or like the plate to be cleaned, and it to to go out in a nice way. And I think more and more. That’s where a lot of data teams fail because

541 00:54:31.280 ⇒ 00:54:40.409 Uttam Kumaran: you have an engineering team. Who’s like, Oh, who cares about this biz like we got the data there, ship it. But then, if you put yourself in the other shoes, they’re like

542 00:54:41.150 ⇒ 00:54:49.170 Uttam Kumaran: the Pm. Or the executive is like. I have. No, I like. They have no idea about this. I can I? I walk them through. But I don’t expect them to like.

543 00:54:49.380 ⇒ 00:54:54.929 Uttam Kumaran: know this right. This is what we’re here to do, but I expect them to love this and be able to sit

544 00:54:54.960 ⇒ 00:54:57.437 Uttam Kumaran: in one of these tools. And

545 00:54:57.980 ⇒ 00:55:07.089 Uttam Kumaran: it’s been a difficult process. So that’s why I kind of went to go. I basically went to talk to like 4 or 5 of these bias code type tools. And

546 00:55:07.140 ⇒ 00:55:23.580 Uttam Kumaran: not only were real and evidence really great products. Both the teams were very, very kind to me, and like very helpful when we weren’t even paying for like months, and just like testing locally. And I was submitting like feedback and fixes. And the support has been really really great.

547 00:55:24.920 ⇒ 00:55:28.149 Uttam Kumaran: and I think about us as like our end products.

548 00:55:28.600 ⇒ 00:55:31.300 Uttam Kumaran: What we do has a visual component to it.

549 00:55:31.310 ⇒ 00:55:36.739 Uttam Kumaran: And I would prefer it to look really, really amazing. And I’m honestly more biased towards that

550 00:55:36.990 ⇒ 00:55:38.520 Uttam Kumaran: than I am towards

551 00:55:38.700 ⇒ 00:55:39.520 Uttam Kumaran: like.

552 00:55:39.920 ⇒ 00:55:41.749 Uttam Kumaran: and then I am towards like

553 00:55:42.410 ⇒ 00:55:46.570 Uttam Kumaran: even costs, even though both of those tools are cheaper than white dash.

554 00:55:46.670 ⇒ 00:55:48.900 Uttam Kumaran: Yeah, cost. But also, just like

555 00:55:49.640 ⇒ 00:55:58.419 Uttam Kumaran: real is a different workflow. And I’m I’m happy to take the risk on explaining it, because I know, like, I just know that, like I have a feeling that that’s a better tool.

556 00:55:58.520 ⇒ 00:56:01.900 Uttam Kumaran: And I’m comfortable showing because the tool looks great. It’s quick

557 00:56:02.110 ⇒ 00:56:02.930 Uttam Kumaran: like

558 00:56:03.540 ⇒ 00:56:23.620 Uttam Kumaran: I’m happy to tell people about it, you know. It’s like, Oh, I bought this fancy new shovel like. Look at this has so many features, I guess, like a light it has like it’s like it’s got an extended poll like I. It’s just nice to talk about, because I I think it actually is more conducive to the work that I do, which is like, I quickly ask questions. I need to see comparison comparative things.

559 00:56:23.690 ⇒ 00:56:24.939 Uttam Kumaran: you know, so.

560 00:56:25.460 ⇒ 00:56:26.870 Robert Tseng: Yeah. Totally.

561 00:56:30.250 ⇒ 00:56:33.279 Uttam Kumaran: Cool. I mean, I think I could talk briefly about like.

562 00:56:33.390 ⇒ 00:56:36.770 Uttam Kumaran: but I mean lightash is like 600 bucks for the platform.

563 00:56:36.960 ⇒ 00:56:42.650 Uttam Kumaran: I think maybe they’ll go a little bit lower. Real. It’s it’s based on

564 00:56:43.680 ⇒ 00:56:45.479 Uttam Kumaran: gigs of data processed.

565 00:56:45.500 ⇒ 00:56:46.400 Uttam Kumaran: which

566 00:56:47.600 ⇒ 00:56:51.929 Uttam Kumaran: we haven’t even hit like we’re not even close to hitting like 10 gigs.

567 00:56:52.398 ⇒ 00:56:54.289 Uttam Kumaran: For one of our clients.

568 00:56:54.320 ⇒ 00:56:55.550 Uttam Kumaran: And

569 00:56:56.440 ⇒ 00:56:57.920 Uttam Kumaran: yeah, basically, I,

570 00:56:58.100 ⇒ 00:57:06.460 Uttam Kumaran: I’m not worried about like us. We’re not processing like terabytes of information, or even like hundreds of gigs of data. We’re looking at basic time series stuff?

571 00:57:09.000 ⇒ 00:57:10.300 Uttam Kumaran: so I don’t know. I don’t.

572 00:57:10.980 ⇒ 00:57:15.410 Uttam Kumaran: Yeah. I think I think if it’s if it’s less than $1,000 for Bi.

573 00:57:15.510 ⇒ 00:57:23.380 Uttam Kumaran: I’m like happy cause. That’s what I was used to light. Dash was the 1st alternative that I kind of found. And I was like, I don’t wanna put $1,000 on them.

574 00:57:23.400 ⇒ 00:57:32.759 Uttam Kumaran: Real is even cheaper. Evidence is also just as cheap. It’s like $15 per user for cloud hosting. Or if you’re like, if people are really cost constrained.

575 00:57:34.570 ⇒ 00:57:46.150 Uttam Kumaran: I just run. I we pay $15 to help so soon like magnify both real and both real and evidence. You could run locally for free, so you could just run this on your laptop.

576 00:57:46.270 ⇒ 00:57:50.189 Uttam Kumaran: Frankly, if it’s just you, you could probably just do for free and just like have it.

577 00:57:50.260 ⇒ 00:57:55.819 Uttam Kumaran: if if unless there, unless if you’re just through presentations. But the people are logging into using it.

578 00:57:56.440 ⇒ 00:57:57.640 Uttam Kumaran: It’s probably worth

579 00:57:58.630 ⇒ 00:58:00.929 Uttam Kumaran: paying for. But 15 bucks is

580 00:58:00.980 ⇒ 00:58:02.340 Uttam Kumaran: pretty reasonable and

581 00:58:02.500 ⇒ 00:58:04.619 Uttam Kumaran: real. They do like 2, 50 a month.

582 00:58:06.880 ⇒ 00:58:07.550 Uttam Kumaran: but

583 00:58:08.290 ⇒ 00:58:09.050 Uttam Kumaran: yeah.

584 00:58:11.630 ⇒ 00:58:16.300 Robert Tseng: Yeah, got it? Yeah, I looked into the pricing. So I think that that made that all made sense to me

585 00:58:16.670 ⇒ 00:58:26.239 Robert Tseng: platform fee. Just, I think, high initial cost, especially since they’re they’re users. I think user based pricing is typically, you know, whatever people people are used to that

586 00:58:26.577 ⇒ 00:58:31.159 Robert Tseng: because you know the cost is small, and then you it scales up. But

587 00:58:31.718 ⇒ 00:58:33.550 Robert Tseng: yeah. So I think

588 00:58:33.870 ⇒ 00:58:43.498 Robert Tseng: just being able to talk through those those 2 like approaches to pricing, I think, is always like a challenge I run into when I’m like evaluating bi tools with with folks.

589 00:58:44.100 ⇒ 00:58:52.370 Uttam Kumaran: One of them would be open to both of them would be open to proof of concepts too like for evidence for one of our clients. We’re going to kick off a proof of concept.

590 00:58:52.450 ⇒ 00:59:00.159 Uttam Kumaran: We’re basically like, I need a month, or I need a month to like kind of run this, and I need it for free to be able to test out. And they’re like, totally whatever

591 00:59:00.440 ⇒ 00:59:01.139 Uttam Kumaran: you know. So.

592 00:59:01.140 ⇒ 00:59:01.470 Robert Tseng: Yeah.

593 00:59:01.470 ⇒ 00:59:02.880 Uttam Kumaran: That was also really nice.

594 00:59:04.020 ⇒ 00:59:04.700 Robert Tseng: Got it

595 00:59:06.230 ⇒ 00:59:07.330 Robert Tseng: good to know.

596 00:59:09.690 ⇒ 00:59:11.159 Robert Tseng: Okay, so yeah, let me.

597 00:59:11.160 ⇒ 00:59:13.184 Uttam Kumaran: Yeah. Why, yeah, I forgot. I forgot. Why, you’re

598 00:59:13.540 ⇒ 00:59:14.560 Uttam Kumaran: kind of sad.

599 00:59:14.960 ⇒ 00:59:24.659 Robert Tseng: Oh, no, I was. I was just talk talking through like the the different types of pricing rather than just like the the number, like the different approaches to pricing. Yeah.

600 00:59:25.680 ⇒ 00:59:27.529 Uttam Kumaran: What’s your gut instinct? What do you think.

601 00:59:28.890 ⇒ 00:59:32.579 Robert Tseng: For for for Stella, or just like in general.

602 00:59:33.040 ⇒ 00:59:47.479 Uttam Kumaran: For Stella, and then also in general, like. And then I wanna talk about the we have. We still have time to talk about on the new opportunity. But like, yeah, instead, for this is just talk about Stella. First, st think about it for this. Whatever we need to do to handy whatever we need. Yeah.

603 00:59:48.380 ⇒ 00:59:53.505 Robert Tseng: Yeah, I mean, I give you my thoughts generally already. So I I think, yeah, I I could see.

604 00:59:54.100 ⇒ 00:59:56.644 Robert Tseng: yeah, that that exploratory

605 00:59:58.970 ⇒ 01:00:07.210 Robert Tseng: yeah, for data exploration. Like, totally think like the real approach. Like, I like that. I’ve I’ve not seen anything like that before. So

606 01:00:08.160 ⇒ 01:00:11.960 Robert Tseng: yeah, and that. But then I think for for Stella specifically.

607 01:00:13.970 ⇒ 01:00:23.034 Robert Tseng: yeah, I mean, I I guess my gut, my gut is still to go with something more traditional, whether it’s like light light dash or or metabase.

608 01:00:23.370 ⇒ 01:00:24.050 Uttam Kumaran: Yeah.

609 01:00:24.050 ⇒ 01:00:49.971 Robert Tseng: I mean, it is very Greenfield, like some of this is like a mature organization. They they they talk about being on power bi. But like I’ve looked at their usage, they’ve had less than like 5 users on power Bi. There are no executive dashboards flowing around. Everything is just spreadsheets excel being shared around the company right? Or they’re using like in app reporting off of like the tools like Netsuite or whatever they’re they’re doing. And that’s and that’s pretty much it. So

610 01:00:50.830 ⇒ 01:00:54.149 Robert Tseng: yeah, I do think that there’s big resistance to like

611 01:00:54.370 ⇒ 01:00:57.050 Robert Tseng: learning like workflows. So.

612 01:00:57.050 ⇒ 01:00:57.520 Uttam Kumaran: Yeah.

613 01:00:58.315 ⇒ 01:00:59.110 Robert Tseng: I

614 01:01:00.630 ⇒ 01:01:23.390 Robert Tseng: I’m totally okay with, I mean, lighthouse. Seems like it’s very easy to plug into like the dB models that we’re doing, and that that’s why, you know, if anything, if we, you know between Madison Lighthouse, if you have a preference on like what works better for, like the data modeling that we’ve done. I’d rather just go with that one give them the illusion that they’re being able to explore data. Yeah, they can click into all the dimensions and like, look at the metrics.

615 01:01:23.724 ⇒ 01:01:37.189 Robert Tseng: Yeah, I mean, I could even do some videos. But at the end of the day, like the the dashboarding or the reports that are going to be consumed more broadly, like we’re like, I would want to own that or like, have have our have our team do that

616 01:01:37.210 ⇒ 01:01:40.919 Robert Tseng: rather than like letting letting there, then then do that internally.

617 01:01:41.070 ⇒ 01:01:42.319 Robert Tseng: So there’s that

618 01:01:42.460 ⇒ 01:01:55.749 Robert Tseng: on the for those consumers. Specifically, there’s then there’s a Sabrina like kind of persona. And like some of these other more technical folks who they do have a lot of business questions. And they want to ask questions of the data. They don’t know how to run sequel, though.

619 01:01:55.860 ⇒ 01:01:58.510 Robert Tseng: Then I’d be like, well, that’s that’s like a good environment.

620 01:01:58.510 ⇒ 01:01:58.850 Uttam Kumaran: Yeah.

621 01:01:58.850 ⇒ 01:02:01.960 Robert Tseng: To to introduce something like a real.

622 01:02:02.270 ⇒ 01:02:07.849 Uttam Kumaran: I think they’re gonna be blown away. Everybody I put this in front of in the business side. They’re like, who has like, who knows.

623 01:02:07.920 ⇒ 01:02:10.740 Uttam Kumaran: kind of like they they know enough to be dangerous, or like

624 01:02:11.360 ⇒ 01:02:14.110 Uttam Kumaran: Whoa! I never. So

625 01:02:14.770 ⇒ 01:02:21.729 Uttam Kumaran: I I was literally like hoping that that was what a good, Scott, I was. Gonna say, the same thing. I’m like, I feel like there’s a traditional part of the business that, like

626 01:02:21.760 ⇒ 01:02:26.630 Uttam Kumaran: you start just just start from the basics of like getting things into metabase.

627 01:02:26.820 ⇒ 01:02:33.709 Uttam Kumaran: having something that they’re like happy with the data. Then we can think about like gaining trust with them, and then thinking about it. But

628 01:02:33.790 ⇒ 01:02:37.459 Uttam Kumaran: you also, if you have a so a person like Sabrina.

629 01:02:37.650 ⇒ 01:02:38.463 Uttam Kumaran: who’s like

630 01:02:38.910 ⇒ 01:02:43.755 Uttam Kumaran: I don’t need like a fixed dashboard. I have so many questions that that’s gonna be changing every day.

631 01:02:44.010 ⇒ 01:02:46.539 Robert Tseng: You don’t care about the dashboard, you just want to answer questions. Yeah.

632 01:02:46.540 ⇒ 01:02:52.169 Uttam Kumaran: She wants. She’s operator right? And so even I have the say, we’re the same thing for for this pool party we have CEO, who’s like.

633 01:02:52.170 ⇒ 01:02:52.550 Robert Tseng: Yeah.

634 01:02:52.550 ⇒ 01:02:54.440 Uttam Kumaran: He’s like, I wanna wake up every day and see something.

635 01:02:54.660 ⇒ 01:02:57.550 Uttam Kumaran: The the other guy who’s a CEO, basically he’s like.

636 01:02:57.780 ⇒ 01:03:01.530 Uttam Kumaran: I don’t use any of that shit. He’s like, I have a feel for the business.

637 01:03:01.610 ⇒ 01:03:08.110 Uttam Kumaran: But then I want to see numbers is like, I’m just gonna ask you. And I’m like, oh, perfect! This would be perfect for you to just have handy.

638 01:03:08.310 ⇒ 01:03:17.530 Uttam Kumaran: so that when you’re kind of like you, their minds racing, and they want to ask much of stuff they could do it, and also if they were to pull that up into a meeting, it makes it them look very, very good

639 01:03:17.600 ⇒ 01:03:25.400 Uttam Kumaran: for for some of our customers, if that doesn’t really matter. But Stella is like a larger organization that like if she was to bring it up

640 01:03:25.480 ⇒ 01:03:26.820 Uttam Kumaran: during a meeting.

641 01:03:27.020 ⇒ 01:03:30.760 Uttam Kumaran: I feel like it would be like huge for for her, like, you know.

642 01:03:31.610 ⇒ 01:03:42.890 Robert Tseng: Yeah. And it already is like she’s she’s relatively new, like past 2 years. Tenure. So her and a couple of people came from tech into this. This is traditionally like a manufacturing client. Right? So.

643 01:03:42.890 ⇒ 01:03:43.330 Uttam Kumaran: Just say.

644 01:03:43.330 ⇒ 01:04:06.739 Robert Tseng: The bias is very intuition based. And so she tried to drive like data drivenness. But then she can’t actually like pull in the data herself. And so it’s been hard for her to really like push for like that change, that paradigm shift. So, being able to equip her, she is the loudest voice. She’s a well respected person now at the table. Least product for them. I think I would want to give her like something like real but then, like the people who are like kind of

645 01:04:06.840 ⇒ 01:04:23.869 Robert Tseng: sponsoring like our contract, are the more traditional folks who just want like more glorified spreadsheet like capability. Being able to plug in something like a light light dash or metadase very quickly, I think, would be just helpful to kind of like. Keep them at bay, for for now, while we

646 01:04:23.950 ⇒ 01:04:30.159 Robert Tseng: work on the higher value stuff cause I agree, like, I feel like it’s been kind of annoying like the past few weeks, just like

647 01:04:30.560 ⇒ 01:04:40.810 Robert Tseng: trying to understand. Like, what do you want to dashboard? And it’s like, yeah, just doing all these like visual changes, or like kind of redesigning mock ups like, I don’t really feel like I’m actually answering questions for them.

648 01:04:40.810 ⇒ 01:04:55.529 Uttam Kumaran: So we know that I was like David. We know that like you’re like, Damn, although I can do this when I get paid for it’s just like I would really like to see you guys just like use this data to move their business. There’s so much opportunity. That’s how I felt, too, and like

649 01:04:55.750 ⇒ 01:04:59.179 Uttam Kumaran: I kind of got over the mental cliff of like

650 01:04:59.230 ⇒ 01:05:01.300 Uttam Kumaran: being nervous about that. I was like.

651 01:05:01.470 ⇒ 01:05:09.779 Uttam Kumaran: if I want to prioritize and be honest about prioritizing that. You guys actually use the data to do it. Maybe you think you need a dashboard.

652 01:05:09.920 ⇒ 01:05:20.769 Uttam Kumaran: What you need is like a date like a weekly meeting with me, where we just like, sit on light dash or sit on rail, and you can ask questions, and we could walk through it right. So I try to cut through the like.

653 01:05:21.260 ⇒ 01:05:48.990 Uttam Kumaran: I need a dashboard, because I’m just like that’s what business that’s like. You’re just like an Mpc. Character here like, I just need that. Instead, I wanna be like, no like what you have. What is your what’s like? The last data question you had? You couldn’t answer. And like, Let’s walk through that. Maybe you’ll find that like real is the best. And the nice thing is, these tools are so cheap now that like. And they’re they’re they’re open and doing proof of concepts and like their Vc. Back. So they don’t. They really don’t care like

654 01:05:49.380 ⇒ 01:05:53.761 Uttam Kumaran: they’ll help, you know, and I’m happy to have the vendor support on that at least, cause they’re still both small.

655 01:05:54.487 ⇒ 01:06:04.479 Uttam Kumaran: That’s really nice. And then they also. Yesterday, you know I was briefly talking about like, do do they have any other manufacturing clients? And they sent me this I’ll send to you on slack.

656 01:06:05.883 ⇒ 01:06:08.410 Uttam Kumaran: This company, Jessie Hardware.

657 01:06:09.230 ⇒ 01:06:14.216 Uttam Kumaran: This is like some sort of like hardware Manufacturing company that’s using rail.

658 01:06:14.600 ⇒ 01:06:15.320 Robert Tseng: Yeah.

659 01:06:15.560 ⇒ 01:06:17.670 Uttam Kumaran: And so if it helps for the pitch.

660 01:06:18.350 ⇒ 01:06:20.790 Uttam Kumaran: I was just like yo, can you tell me if there’s any other like

661 01:06:21.410 ⇒ 01:06:27.810 Uttam Kumaran: manufacturing type companies using this, they’re using this for, like their their e-commerce store. And like.

662 01:06:27.890 ⇒ 01:06:30.389 Uttam Kumaran: first, st all the for, like a good amount of their data.

663 01:06:30.810 ⇒ 01:06:31.530 Robert Tseng: Yeah.

664 01:06:35.930 ⇒ 01:06:37.260 Uttam Kumaran: So, yeah.

665 01:06:38.420 ⇒ 01:06:40.917 Robert Tseng: Okay, no, this is good. This is helpful.

666 01:06:41.580 ⇒ 01:06:59.759 Robert Tseng: yeah. I mean, I already had like the initial conversation, because we just wanted to check in I sent over like that table that I think I sent you and I’m gonna update it. Send her some, doc. I’m gonna I’m already already headlights from the proposal I sent her. We’re in the evaluation I sent her. But yeah, I didn’t really have anything unreal there yet. So

667 01:07:00.282 ⇒ 01:07:01.250 Robert Tseng: yeah, I mean.

668 01:07:01.511 ⇒ 01:07:08.049 Uttam Kumaran: It may be nice honestly, if you can pull real up locally, because I think you could run it locally plug in the Zendesk table.

669 01:07:08.390 ⇒ 01:07:08.969 Robert Tseng: Yeah, I mean.

670 01:07:08.970 ⇒ 01:07:28.709 Uttam Kumaran: Like, I can even like we can help set it up and like, have the code sitting. But like, yeah, I think you walk through that process should just take you maybe an hour or 2 to like, make sure you have everything. It’s just you have to simply put in your snowflake user password, maybe pull. This is the 1st tool where I was like, damn if it’s this click to develop, I can almost like show and tell. I don’t really have to say.

671 01:07:29.120 ⇒ 01:07:31.419 Uttam Kumaran: Yeah, I just have it sell itself almost.

672 01:07:31.795 ⇒ 01:07:39.369 Uttam Kumaran: Here’s a doc. And here’s like screenshots and evidence also takes a little bit of time to set up. They have examples, though, but

673 01:07:39.810 ⇒ 01:07:56.269 Uttam Kumaran: for real. I had a friend of mine where I I sent him. He was like he’s in. He’s in support of manufacturing Pipe Manufacturing Company. He I met him at a thing here in Austin, and I was like you should check out real. He was like, what’s what’s a good data tool? And he’s like, yeah, did I set it up in the afternoon? I just showed it to them like later that day, and then they loved it, and I was like

674 01:07:56.780 ⇒ 01:08:00.129 Uttam Kumaran: sick like. That’s great like that that’s unheard of, you know.

675 01:08:00.130 ⇒ 01:08:13.040 Robert Tseng: Yeah, that is that’s crazy. Yeah, no. I wanna just on a weekly sync with like, Sabrina, I just show her real that she’s she’s only yeah, we kind of walked through already. So yeah, I think just to be clear. So yeah, either. I mean

676 01:08:13.310 ⇒ 01:08:28.120 Robert Tseng: metabasal like, Dash, I I guess 1, one of one of those 2 for for to get the Zendesk dashboard, and just like, have that environment set up for them. But then, like now that we’re having the production data coming into Snowflake.

677 01:08:28.180 ⇒ 01:08:46.012 Robert Tseng: And when I, when we transition to like really working more closely with getting Sabrina answers, I wanna, I wanna set up. I want to demo real for her. So I think that’s that’s the way I I mean, that’s those are my. Those are my thought like, that’s my gut. And seeing Fernando how we should approach this right now.

678 01:08:46.470 ⇒ 01:08:47.749 Robert Tseng: but yeah, so.

679 01:08:47.750 ⇒ 01:08:57.229 Uttam Kumaran: So I’m gonna I’m gonna make sure that the doc I sent. It’s just like we needed that, Doc. Anyways, because I have. I’d have this conversation a bunch. So it’s actually good push to just finish that up.

680 01:08:57.887 ⇒ 01:09:01.470 Uttam Kumaran: And I’ll have this recording and stuff in there. And then.

681 01:09:01.609 ⇒ 01:09:02.819 Uttam Kumaran: yeah, let me know

682 01:09:03.394 ⇒ 01:09:09.700 Uttam Kumaran: timing and everything, and how I can be helpful. So I mean, then maybe, do we want to talk about the.

683 01:09:10.100 ⇒ 01:09:10.470 Robert Tseng: Yes.

684 01:09:11.019 ⇒ 01:09:11.559 Uttam Kumaran: Opportunity.

685 01:09:11.689 ⇒ 01:09:12.319 Uttam Kumaran: Yeah.

686 01:09:12.319 ⇒ 01:09:12.989 Robert Tseng: Yeah.

687 01:09:13.679 ⇒ 01:09:17.509 Robert Tseng: so let me pull up some notes here.

688 01:09:21.739 ⇒ 01:09:36.475 Robert Tseng: Okay, yeah, I mean, so this is like, a, yeah, it’s an Eco company based out of based on New York right now, like, I think I I got connected them because this guy like the V, their Vp of tech and

689 01:09:36.839 ⇒ 01:09:44.399 Robert Tseng: yeah, they have. They’re so just so pretty early on, probably like around what? Like 1010,000,000 revenue or less.

690 01:09:45.039 ⇒ 01:09:46.779 Robert Tseng: And yeah, they’re

691 01:09:46.909 ⇒ 01:10:13.489 Robert Tseng: try. They? They’ve been trying to improve their like user tracking. They they use like a Google Tag manager and Lr for Pixel based tracking but now they’re they recently went into amplitude and for trying to set up, set that stuff up. So that’s usually like an entry point for me to like. Have conversations, because, like we’ll, we’ll like redo their amplitude or mixed panel whatever, and get them setting up. Get them set up on some like AV testing.

692 01:10:13.789 ⇒ 01:10:40.364 Robert Tseng: being able to like, get some of the attribution reporting that they want about for their users in in that environment. So yeah, I mean, I think that’s and because we’re solutions partner for amplitude and mixed panel, I typically use them for like product experimentation, like AV testing kind of like reports. So that’s usually my foot in the door. But then, as someone like talking to him, yeah, I mean, they don’t really have any of the the data stack set up yet. And

693 01:10:40.709 ⇒ 01:11:09.979 Robert Tseng: yeah. So he was just kind of chatting with me. Some experience we’ve had like I shared about like my experience with looker. But then I also mentioned like, Hey, I’m I work with somebody who who built the data stack for ag one I remember you sent me that case study I I kinda I didn’t have it on hand, but like I just talked through a few points of what I remembered from what you what you described there, you know, stuff like dbt, like consulting to resources and all that. So it was like, oh, cool! I think that’s something that we we would want as well. So

694 01:11:09.979 ⇒ 01:11:28.609 Robert Tseng: I was just thinking, okay, well, I’m gonna re-engage with with them, like tomorrow, I’m gonna hop on a call with his CEO and Cmo. And I’d love to be able to like. Give them, give them a pitch and try to try to go for try to get get both in the same. The same contract. But yeah, that’s kind of where I’m at with. With that opportunity.

695 01:11:29.130 ⇒ 01:11:33.030 Uttam Kumaran: Yeah, I think. Yeah, pretty straightforward, I think.

696 01:11:33.633 ⇒ 01:11:37.970 Uttam Kumaran: We can push amplitude through 5 Tran set up snowflake.

697 01:11:38.040 ⇒ 01:11:39.470 Uttam Kumaran: And then

698 01:11:39.940 ⇒ 01:11:49.679 Uttam Kumaran: basically, like, if it depends on how big their data stuff is. But I assume they also want, like shopify, or whatever they’re using for their ecom. Again, everything.

699 01:11:49.680 ⇒ 01:11:50.170 Robert Tseng: Shopify.

700 01:11:50.170 ⇒ 01:11:50.580 Uttam Kumaran: 5 to.

701 01:11:50.580 ⇒ 01:11:51.280 Robert Tseng: Yeah.

702 01:11:51.280 ⇒ 01:12:06.249 Uttam Kumaran: Get. Get it all in in Snowflake. And then it’s kind of the same situation. We’ll start like whatever based on one report or whatever we need to start with and basically begin modeling there. If it’s shopify, then we already have a ton of boilerplate code ready to go same with

703 01:12:06.710 ⇒ 01:12:09.059 Uttam Kumaran: same with all the ad platforms.

704 01:12:09.150 ⇒ 01:12:11.967 Uttam Kumaran: So it should be pretty quick for us to

705 01:12:12.930 ⇒ 01:12:15.939 Uttam Kumaran: like, get started there. And yeah, in terms of like

706 01:12:16.280 ⇒ 01:12:19.709 Uttam Kumaran: people. So yeah, I mean, we’ve done ad measurement.

707 01:12:19.960 ⇒ 01:12:25.819 Uttam Kumaran: like almost every client that we’ve worked for. And then, yeah, at ag one. And we work

708 01:12:26.309 ⇒ 01:12:27.390 Uttam Kumaran: at flow. Code

709 01:12:27.580 ⇒ 01:12:32.440 Uttam Kumaran: did all like add reporting, basically. So everything from pay marketing to out of home TV

710 01:12:32.530 ⇒ 01:12:37.650 Uttam Kumaran: organic, like all that stuff. Nick, is actually really, really talented in that, too.

711 01:12:37.820 ⇒ 01:12:47.709 Uttam Kumaran: So, depending on his his availability, too. I’ll ask him whether. Yes, I that’s that’s really his like bread and butter. It’s a ton of marketing like marketing, ae related work.

712 01:12:48.630 ⇒ 01:12:49.240 Robert Tseng: Okay.

713 01:12:51.820 ⇒ 01:12:53.920 Uttam Kumaran: I mean, I’m down. Yeah. What?

714 01:12:55.260 ⇒ 01:12:58.540 Uttam Kumaran: I’m happy to write some stuff down, or what do you think.

715 01:12:59.689 ⇒ 01:13:04.780 Robert Tseng: Yeah, I mean, I think I I’m I’m gonna I think they’re gonna probably wanna chat through

716 01:13:05.350 ⇒ 01:13:06.340 Robert Tseng: the I can.

717 01:13:06.340 ⇒ 01:13:09.910 Uttam Kumaran: Even give you some sample product like, for example, Ecom thought side, there’s like

718 01:13:10.210 ⇒ 01:13:12.540 Uttam Kumaran: so much other stuff there’s like.

719 01:13:12.590 ⇒ 01:13:14.779 Uttam Kumaran: of course, we walk through shipping today.

720 01:13:15.040 ⇒ 01:13:20.180 Uttam Kumaran: Customer related stuff marketing, related stuff refunds, discounts like

721 01:13:20.610 ⇒ 01:13:25.050 Uttam Kumaran: you could just throw a lot of that at them. Because if they’re not doing anything right now.

722 01:13:25.500 ⇒ 01:13:28.040 Uttam Kumaran: that’s gonna really be like Peak their interest.

723 01:13:29.130 ⇒ 01:13:29.830 Uttam Kumaran: So.

724 01:13:30.190 ⇒ 01:13:58.929 Robert Tseng: Yeah, if you have any like, if you have any like samples that you think would be good good to talk through, just to like kind of showcase, like certain milestones. I think like when when I when I talk to somebody who’s evaluating this for the 1st time, they want to know. Oh, you’ve done something before. They want to hear, just like the highlights of what you’ve done. So I know as you want Peak, your interest, so I’ll go back to your case study. I’m I wouldn’t like to talk through like oh, couple of the milestones you went. You don’t have to pull in that client exactly, but it can be like, Hey, like this was like a big moment we had with them, and like moving through

725 01:13:58.930 ⇒ 01:14:02.020 Robert Tseng: and like, just like some. I don’t know if it

726 01:14:02.660 ⇒ 01:14:24.059 Robert Tseng: I don’t know if there’s it doesn’t necessarily have to be visual, but just like a talking point to be like, yeah, we’d get you to this point where I mean, I guess maybe this just like plugging everything into snowflake. And then from there, like, yeah, this was like the 1st report that we generated for them. And like, yeah, just being able to show them those milestone like Valley moments, I think, is, I, I found to be like good, like

727 01:14:24.510 ⇒ 01:14:31.070 Robert Tseng: point for me in the conversation, to like, pause there and see like how they’re, how they’re responding, and then to pick which direction to go in.

728 01:14:31.470 ⇒ 01:14:34.835 Uttam Kumaran: Okay, okay, cool. So maybe I’ll just slack

729 01:14:35.750 ⇒ 01:14:40.480 Uttam Kumaran: I’ll just yeah. I’ll just slack you with. A couple of these examples of just like

730 01:14:40.600 ⇒ 01:14:43.659 Uttam Kumaran: things that we’ve done in the e-commerce world that’s like.

731 01:14:43.850 ⇒ 01:14:48.105 Uttam Kumaran: really hits that like things that are on everybody’s mind, basically any com

732 01:14:49.390 ⇒ 01:14:50.350 Uttam Kumaran: And

733 01:14:50.520 ⇒ 01:14:52.419 Uttam Kumaran: yeah, we’ll go from there. Go up a bit.

734 01:14:52.420 ⇒ 01:14:52.910 Robert Tseng: Got it.

735 01:14:52.910 ⇒ 01:14:53.530 Uttam Kumaran: Sick.

736 01:14:54.290 ⇒ 01:14:54.890 Robert Tseng: Yeah.

737 01:14:57.590 ⇒ 01:14:59.469 Uttam Kumaran: Okay, cool?

738 01:15:01.160 ⇒ 01:15:08.417 Uttam Kumaran: I guess. Otherwise we’ll catch up on. Still, I ha! I got it. I’m a Texas snow fake guy back.

739 01:15:10.950 ⇒ 01:15:11.850 Uttam Kumaran: And Mike.

740 01:15:13.220 ⇒ 01:15:16.468 Uttam Kumaran: So it’s been a nightmare getting this thing done.

741 01:15:16.940 ⇒ 01:15:21.319 Uttam Kumaran: so let me let me push on him, and then I know you sent some stuff yesterday from Ryan.

742 01:15:22.160 ⇒ 01:15:35.110 Robert Tseng: Yeah, no, I already, I think, I answered Nick. I think I answered my own question. I think he he tried to. Yeah, I think he just got blocked again. He couldn’t find like a connect via private networking. I I think it’s because it’s not enabled. Still. So

743 01:15:36.360 ⇒ 01:15:37.260 Robert Tseng: yeah.

744 01:15:38.090 ⇒ 01:15:46.289 Uttam Kumaran: Okay. And then I send an email yesterday. They just wanted me to send like a confirmation email with someone from Stella that they needed business critical. That was the email I sent.

745 01:15:46.900 ⇒ 01:15:49.549 Uttam Kumaran: I, I, CC, your Stella source email.

746 01:15:49.830 ⇒ 01:15:50.300 Robert Tseng: Oh!

747 01:15:50.300 ⇒ 01:15:50.740 Uttam Kumaran: You just say.

748 01:15:52.720 ⇒ 01:15:55.990 Robert Tseng: we actually all moved off of cell source. So like the.

749 01:15:55.990 ⇒ 01:15:58.629 Uttam Kumaran: I don’t think they need. I don’t think they need. I don’t think they need any.

750 01:15:58.630 ⇒ 01:15:59.839 Robert Tseng: Oh, got it? Got it? Got it? Okay?

751 01:15:59.840 ⇒ 01:16:01.759 Uttam Kumaran: They just wanted me to send like a

752 01:16:02.110 ⇒ 01:16:04.550 Uttam Kumaran: can you formally request? And I was like.

753 01:16:05.130 ⇒ 01:16:07.380 Uttam Kumaran: I’m like talking to him. I’m like Dude.

754 01:16:07.510 ⇒ 01:16:09.879 Uttam Kumaran: Well, can’t you just like, get this done? What are you?

755 01:16:10.280 ⇒ 01:16:15.760 Uttam Kumaran: I understand. But I was like, I’ll send like a email requesting business critical. But yeah, so.

756 01:16:15.760 ⇒ 01:16:17.870 Robert Tseng: Yeah, no. Problem, yeah.

757 01:16:19.070 ⇒ 01:16:21.095 Uttam Kumaran: Okay? And then the oh, the last thing. So

758 01:16:21.330 ⇒ 01:16:33.320 Uttam Kumaran: like we’re good for, like, I wanted to talk to Nick about like the contract and everything between us. So like, we’re, we’re good for another few months like, did you need? Did you need anything there? Or we’re just gonna keep continuing.

759 01:16:34.207 ⇒ 01:16:36.392 Robert Tseng: Yeah. Or do we not.

760 01:16:36.830 ⇒ 01:16:38.629 Uttam Kumaran: On Monday like plan out

761 01:16:38.960 ⇒ 01:16:45.300 Uttam Kumaran: a little bit. Now that Nico’s here, I’m a little bit I really am. We’re trying to get a little bit better with due dates.

762 01:16:45.400 ⇒ 01:16:51.609 Uttam Kumaran: and actually conveying that earlier to the engineers and really sticking to that having some accountability. So

763 01:16:51.710 ⇒ 01:16:53.430 Uttam Kumaran: I want to

764 01:16:53.990 ⇒ 01:17:00.179 Uttam Kumaran: think about like that. And since we’re also kind of like, some of these are in flight, I wanna make sure that we’re all

765 01:17:00.480 ⇒ 01:17:01.970 Uttam Kumaran: comfortable on dates.

766 01:17:03.170 ⇒ 01:17:03.860 Robert Tseng: Yeah.

767 01:17:06.880 ⇒ 01:17:13.110 Robert Tseng: well, yeah, I mean, the contract is extended. But yeah, I would like to try to get would send them a new doc on like

768 01:17:13.480 ⇒ 01:17:20.800 Robert Tseng: timeline, and like the adjustment on the scope of work. But it’s been hard because it’s been moving target. So I think I kind of shared with you some of the updates on, like

769 01:17:20.860 ⇒ 01:17:24.269 Robert Tseng: I talked through all the different deliverables and how they’ve changed yesterday.

770 01:17:24.668 ⇒ 01:17:41.569 Robert Tseng: But yeah, I mean the the main person who would be signing off is out of out of office this week. So I wasn’t like in a rush to like, put that in front of them. Yeah, it’s probably sometime next week I’d like to. If anything. This week. The outstanding things are still. I want to get them. V, 2 of that dashboard.

771 01:17:42.116 ⇒ 01:18:07.599 Robert Tseng: And then, yeah, hopefully, they they can. I can empower them to make a decision on like, okay, we’re gonna go with metabase or light desk. And we’ll we’ll we’ll keep going with that. And then the stuff with Sabrina, and like real like I think that would be in like the next phase, because, yeah, she’s helping drive this because she wants, like my time, more to like help her answer questions. So that’s kind of like. We just need to get this to a point where I can, like.

772 01:18:07.730 ⇒ 01:18:15.969 Robert Tseng: you know, get get this group of stakeholders off my back. And then I think we’ll be in a better. We’ll be head in a better direction after that.

773 01:18:16.520 ⇒ 01:18:20.660 Uttam Kumaran: Great. Okay? So I guess I don’t know. Maybe because sometime

774 01:18:20.680 ⇒ 01:18:26.349 Uttam Kumaran: next week or in the following 2 weeks we’ll do a longer planning session.

775 01:18:26.958 ⇒ 01:18:31.619 Uttam Kumaran: 1st for Stella, basically thinking through both. And that way we

776 01:18:31.740 ⇒ 01:18:35.109 Uttam Kumaran: we’ll start to think about a longer timeline for milestones.

777 01:18:36.560 ⇒ 01:18:39.819 Uttam Kumaran: and that’ll be super helpful for us, so we can make sure we’re on track.

778 01:18:40.440 ⇒ 01:18:41.410 Nicolas Sucari: Yeah, of course.

779 01:18:42.310 ⇒ 01:18:49.320 Nicolas Sucari: Yeah. And then we can like configure all that. And yeah, go follow that milestone so that we can achieve everything.

780 01:18:49.700 ⇒ 01:18:50.730 Nicolas Sucari: Quote button.

781 01:18:51.970 ⇒ 01:18:52.620 Robert Tseng: Like. Nice.

782 01:18:52.620 ⇒ 01:18:54.599 Uttam Kumaran: Okay. Great great meeting.

783 01:18:54.600 ⇒ 01:18:55.070 Robert Tseng: And okay.

784 01:18:55.070 ⇒ 01:18:55.470 Uttam Kumaran: Yeah.

785 01:18:55.470 ⇒ 01:18:55.800 Robert Tseng: Yeah.

786 01:18:55.800 ⇒ 01:18:57.420 Uttam Kumaran: I’m glad we were able to walk through everything.

787 01:18:57.820 ⇒ 01:19:00.309 Robert Tseng: Yeah, learned a lot really appreciate you guys, time.

788 01:19:00.440 ⇒ 01:19:01.370 Robert Tseng: yeah, of course.

789 01:19:01.860 ⇒ 01:19:02.700 Robert Tseng: Yeah.

790 01:19:03.120 ⇒ 01:19:05.009 Uttam Kumaran: Okay. Alright. Talk to you soon.

791 01:19:05.010 ⇒ 01:19:06.480 Robert Tseng: Welcome, slack! Yep, bye, bye.

792 01:19:06.480 ⇒ 01:19:07.260 Nicolas Sucari: But aye.