Meeting Title: Rill-Overview-Direct-Mail Date: 2024-06-12 Meeting participants: Uttam Kumaran, Brian Pei, Nicolas Sucari, Jakob Kagel


WEBVTT

1 00:00:27.840 00:00:28.220 Nicolas Sucari: Hey!

2 00:00:28.810 00:00:29.960 Jakob Kagel: Hey? How’s it going.

3 00:00:32.210 00:00:33.840 Nicolas Sucari: I’m good. How are you, Tom?

4 00:00:47.530 00:00:50.594 Uttam Kumaran: Hey, guys, sorry so late. Just got back of us talking to Jared.

5 00:00:51.500 00:00:52.119 Jakob Kagel: Oh, no worries.

6 00:00:52.120 00:00:52.846 Uttam Kumaran: One second.

7 00:00:56.310 00:00:57.639 Jakob Kagel: Hey, Brian, how are you?

8 00:00:58.220 00:01:00.200 Brian Pei: What’s up? I’m good. How are you.

9 00:01:00.200 00:01:01.300 Jakob Kagel: Nice to meet you.

10 00:01:02.140 00:01:03.259 Brian Pei: Nice to meet you. Man.

11 00:01:12.980 00:01:15.310 Jakob Kagel: What? What city are you based out of?

12 00:01:16.770 00:01:20.020 Brian Pei: I’m in Brooklyn right now. I’m in Vegas, visiting my parents.

13 00:01:20.240 00:01:21.040 Jakob Kagel: Nice.

14 00:01:21.760 00:01:23.070 Brian Pei: Mostly Brooklyn. There.

15 00:01:23.280 00:01:24.560 Brian Pei: I’m a

16 00:01:25.160 00:01:29.370 Brian Pei: it’s my. It’s my 1st week, but not really because I’m

17 00:01:29.870 00:01:35.419 Brian Pei: taking off whatever that means, because I my Dad’s turning 70 tomorrow and tomorrow and Friday

18 00:01:35.788 00:01:43.860 Brian Pei: we’re doing like a family thing in Vegas. Then I’m going back to Brooklyn on Saturday, and then I’ll be. I’ll be full on starting from next week.

19 00:01:44.570 00:01:46.230 Jakob Kagel: Sounds, cool sounds, cool.

20 00:01:48.870 00:01:56.187 Uttam Kumaran: Okay, well, thanks for attending this webinar. I will be walking through real today.

21 00:01:56.810 00:02:07.599 Uttam Kumaran: ideally, I kind of wanted to just take this time to. We haven’t. We have what an hour! Just we could talk through direct mail stuff. But I basically was gonna try and walk through

22 00:02:08.008 00:02:17.960 Uttam Kumaran: how to get a table into real. And then, basically, we can spend the rest of the time just talking about what we need for direct mail. Ideally, Brian on your side. This is kind of like.

23 00:02:18.100 00:02:30.460 Uttam Kumaran: okay, like, how how does real work in terms of like getting tables into there? It’s basically like, kind of like a little look, Amel, type thing, except like way easier. And then, Jacob, on your side, it’s a little bit of like.

24 00:02:30.770 00:02:35.734 Uttam Kumaran: okay, if we need changes. What are what are sort of changes we can make, and then just the general overview of real

25 00:02:36.663 00:02:41.100 Uttam Kumaran: so let’s get started. I will pull some stuff up.

26 00:02:41.667 00:02:43.779 Uttam Kumaran: This is being recorded.

27 00:02:44.072 00:02:47.619 Uttam Kumaran: But there’s also pretty good documentation on how to run all this stuff.

28 00:02:47.770 00:02:52.270 Uttam Kumaran: There, unfortunately, is some stuff you need to do in the cli

29 00:02:52.859 00:02:56.150 Uttam Kumaran: which I know everybody is a big fan of. But

30 00:02:57.110 00:02:59.804 Uttam Kumaran: the product’s great. So deal with it.

31 00:03:03.070 00:03:04.560 Uttam Kumaran: so let me

32 00:03:06.630 00:03:08.010 Uttam Kumaran: let me share

33 00:03:08.300 00:03:10.969 Uttam Kumaran: an arc screen. And

34 00:03:11.190 00:03:12.340 Uttam Kumaran: this

35 00:03:16.150 00:03:16.880 Uttam Kumaran: move.

36 00:03:19.150 00:03:23.610 Uttam Kumaran: okay, I’m just gonna try to share my like main screen and then let me know if it’s like

37 00:03:25.010 00:03:27.360 Uttam Kumaran: way too big. Is this like way too big

38 00:03:31.440 00:03:32.110 Uttam Kumaran: or.

39 00:03:32.310 00:03:33.050 Jakob Kagel: I mean.

40 00:03:33.050 00:03:34.430 Uttam Kumaran: Way too small sorry way too small.

41 00:03:34.430 00:03:35.075 Jakob Kagel: Yeah.

42 00:03:36.000 00:03:36.889 Uttam Kumaran: It’s pretty small.

43 00:03:36.890 00:03:41.259 Jakob Kagel: It’s a little small, but hold on, I can pull it. Let me pull it up on my monitor.

44 00:03:43.420 00:03:43.950 Jakob Kagel: I think.

45 00:03:43.950 00:03:45.780 Brian Pei: Kind of make up what’s going on.

46 00:03:46.180 00:03:48.489 Uttam Kumaran: Well, or what I’m gonna do is I’m gonna share

47 00:03:48.660 00:03:52.736 Uttam Kumaran: here. Let’s just share this. And then until I can get real up,

48 00:03:54.680 00:03:55.440 Uttam Kumaran: Nope.

49 00:04:01.170 00:04:04.749 Uttam Kumaran: and I’ll just share one screen and the role screen and we’ll have both up.

50 00:04:14.230 00:04:16.620 Uttam Kumaran: Okay, how’s this this working?

51 00:04:17.029 00:04:17.860 Uttam Kumaran: Better.

52 00:04:19.700 00:04:20.310 Nicolas Sucari: Yeah.

53 00:04:22.019 00:04:24.490 Nicolas Sucari: Why are you talking for me? I don’t know about.

54 00:04:25.110 00:04:26.989 Jakob Kagel: Yeah, I I can’t see it. Yeah.

55 00:04:28.310 00:04:29.260 Uttam Kumaran: Alright cool.

56 00:04:30.820 00:04:31.995 Uttam Kumaran: Okay,

57 00:04:33.180 00:04:34.940 Uttam Kumaran: great. So

58 00:04:36.280 00:04:41.319 Uttam Kumaran: we are gonna be talking a little bit about close this. Okay.

59 00:04:41.840 00:04:45.180 Uttam Kumaran: about real. So today, I wanna just walk through

60 00:04:45.756 00:04:49.719 Uttam Kumaran: the structure of real how to bring things into real

61 00:04:50.103 00:05:15.200 Uttam Kumaran: and then we could talk a little bit about deployment and things like that. But again, most of this stuff gets handled locally and then just a basic Pr process. So this is real. On the left, real has a couple of different concepts. There is a yaml file here which just has like kind of like your Dbt project. File, Brian. There’s sources, models, and dashboards.

62 00:05:15.576 00:05:25.570 Uttam Kumaran: Jacob, you may be familiar with kind of how like evidence is set up where you have these intermediary steps, but sources is basically direct selects on tables.

63 00:05:25.961 00:05:28.689 Uttam Kumaran: And I’m gonna make this a little wider.

64 00:05:29.640 00:05:33.449 Uttam Kumaran: direct selects on tables. Which is basically likes. Yeah.

65 00:05:33.800 00:05:39.319 Brian Pei: Is what you’re showing on on the right. Is this? A a repo I can clone already in?

66 00:05:40.380 00:05:47.160 Uttam Kumaran: This is in the pool parts to go, Repo, and the folder is pool parts to go real, so you should already have it.

67 00:05:47.340 00:05:48.859 Brian Pei: Okay. Let me alright.

68 00:05:50.211 00:05:53.199 Uttam Kumaran: But yes, you like to follow along at home. It’s all

69 00:05:53.290 00:06:02.629 Uttam Kumaran: our normal pool parts ago. Repo so we have select star here just from all order items, and I’ve limited it just to

70 00:06:03.130 00:06:06.919 Uttam Kumaran: past 2022, because there’s nothing there, although, can probably delete this

71 00:06:06.960 00:06:09.240 Uttam Kumaran: when you create a new one of these.

72 00:06:09.659 00:06:25.549 Uttam Kumaran: Which we will do today, you’ll basically get a lot of helpful stuff about the dimensions. Percent, like things like which presents are null. It’s just nice to kind of see everything in one place. What you’re doing here is just establishing sources. The next step after that is establishing models

73 00:06:26.060 00:06:32.519 Uttam Kumaran: for our for our use cases. I’m basically just running selects from the sources themselves.

74 00:06:32.770 00:06:54.140 Uttam Kumaran: The the theme of this is that in models you can join sources together, and sources sometimes can come from snowflake can come from. Other integrations can come from Csv’s. So in this situation, we just basically have one to one relationship between models and sources. Let’s say you wanna like limit the amount of models by just

75 00:06:54.140 00:07:09.940 Uttam Kumaran: joining, like all orders and all orders items into one thing you can do that, but this is where you really would do any transformations you need to do in terms of joining tables together. For the most part, I don’t believe we have any like transformations.

76 00:07:10.040 00:07:13.520 Uttam Kumaran: These are all just straight, select stars.

77 00:07:14.503 00:07:15.689 Uttam Kumaran: And so

78 00:07:15.740 00:07:18.280 Uttam Kumaran: basically, what does it just do? Yeah, go ahead.

79 00:07:18.280 00:07:29.940 Jakob Kagel: Alright. So I’d norm. But I just a quick question. So how do we get to this view, like in real cause, like the only link I really know is like, I have, like the real dashboards right? But

80 00:07:30.320 00:07:33.769 Jakob Kagel: can you drop the link? Or how? How do we get to this view?

81 00:07:33.770 00:07:36.679 Uttam Kumaran: So this, all this development has to happen locally.

82 00:07:36.690 00:07:47.520 Uttam Kumaran: So let me let me finish just walking through this to you, and then I’ll walk through, how to bring it up locally and then creating a new thing. So you’ll see that whole step. I’ll do it. One big chunk.

83 00:07:47.520 00:07:48.110 Jakob Kagel: Okay. Alright!

84 00:07:48.359 00:07:50.850 Uttam Kumaran: But just walking us through the concept. But good question.

85 00:07:52.700 00:08:02.569 Uttam Kumaran: so we we basically have a select from our model. And then the nice thing is you’ll see there’s a process by which you can actually generate the dashboard code.

86 00:08:02.700 00:08:15.429 Uttam Kumaran: using some AI process they have, which makes it really easy. I’ve already done for some of the models here. But let’s go, for example, to the All Orders model if you go to the All orders model. You can see this is probably

87 00:08:15.430 00:08:33.180 Uttam Kumaran: very similar to light dash Yaml, very similar to look, Amel Brian. But basically, you put the model here. There is an order, Timestamp, here for the time series, and you have a title, and then you have, of course, like what? Yours, all these different dimensions. So these are the dimensions we have the column description.

88 00:08:33.250 00:08:39.989 Uttam Kumaran: You don’t really use descriptions for much. And then you have your measures. So I’m just gonna actually open this.

89 00:08:40.169 00:08:42.799 Uttam Kumaran: can I open this in a

90 00:08:44.140 00:08:45.850 Uttam Kumaran: new window? Maybe be

91 00:08:47.550 00:08:50.310 Uttam Kumaran: probably we’re not gonna see this. So let me hold on one. Sec.

92 00:08:59.840 00:09:00.750 Uttam Kumaran: Okay, here.

93 00:09:06.580 00:09:13.159 Uttam Kumaran: okay, I just wanted to open up the documentation as well, just to show you a couple things so real.

94 00:09:18.240 00:09:19.800 Uttam Kumaran: Okay? Well.

95 00:09:21.320 00:09:25.540 Uttam Kumaran: they have a couple of different connectors. So we are using the snowflake connector.

96 00:09:25.590 00:09:28.770 Uttam Kumaran: They do have a connector for salesforce, for, like

97 00:09:28.930 00:09:38.839 Uttam Kumaran: connecting directly databases. Again, the nice thing about this is you can connect from multiple different sources, and then in the model layer, connect all of them.

98 00:09:39.120 00:09:44.220 Uttam Kumaran: This takes advantage of duct dB, which is like an in browser database, so

99 00:09:44.340 00:10:01.129 Uttam Kumaran: you can connect from wherever you want, bring everything into one model and then basically use it. And so there’s documentation here on what goes into sources, models and dashboards. I wanna just focus on the dashboard piece because there are configurations here, and this is probably where

100 00:10:01.340 00:10:03.450 Uttam Kumaran: most of the time will be spent.

101 00:10:03.760 00:10:08.439 Uttam Kumaran: It’s pretty quick, as you could see, to bring stuff in. Most of the time will be spent adding

102 00:10:08.460 00:10:11.139 Uttam Kumaran: measures, adding dimensions, customizing them.

103 00:10:11.427 00:10:15.590 Uttam Kumaran: And so a couple of things that you can do. So on the dimension side.

104 00:10:15.600 00:10:19.810 Uttam Kumaran: And I’m just gonna pull the code up here on the right so we can kind of. Follow along.

105 00:10:20.184 00:10:35.729 Uttam Kumaran: Let’s look at all order items. Let’s continue with all orders. On the dimension side you have column, which is just specifies the name of the column that you brought in the expression. We don’t. We’re not doing anything like string splits for the most part, all of our column.

106 00:10:36.050 00:10:39.069 Uttam Kumaran: All of our column. Logic is going to be in Dbt.

107 00:10:39.580 00:11:01.620 Uttam Kumaran: This is kind of just like a pattern of development. But Brian and I have dealt with this for a long time, which is basically like trying to keep all logic in the warehouse. If you keep logic here and there’s like stuff where you’re like adding things together. It never makes it so. You can use it at other tables. It’s not like version controlled. And so we’re for the most part keeping everything there if you need to do string things like lowering in it. Cap

108 00:11:01.680 00:11:07.250 Uttam Kumaran: like operating stuff. Probably fine to do it here. But for the most part we’re keeping everything in the warehouse.

109 00:11:07.300 00:11:09.430 Uttam Kumaran: The name, the label.

110 00:11:09.810 00:11:20.879 Uttam Kumaran: These are both just like, so what shows up in the dashboard, which I’ll show you description. And a couple of other things like on nest, which is like, if there’s lists, there’s something happens. And

111 00:11:22.870 00:11:29.479 Uttam Kumaran: yeah, I think unnest basically means like, if a dimension is like a list of items. It’ll automatically detect that. And basically you can

112 00:11:29.750 00:11:31.899 Uttam Kumaran: change how you do filter

113 00:11:32.270 00:11:34.389 Uttam Kumaran: whatever measures.

114 00:11:35.174 00:11:47.799 Uttam Kumaran: This is a, probably this is like, honestly, a little bit more complicated. You have again name, label expression. These are all aggregates. So you see, everything has an average account.

115 00:11:47.810 00:11:49.020 Uttam Kumaran: a sum

116 00:11:50.066 00:11:52.840 Uttam Kumaran: and sometimes an average of like an expression.

117 00:11:53.430 00:12:07.869 Uttam Kumaran: The description. These are fine format presets. So format presets are, basically, how do you format the output. So sometimes we have basically, it’s a difference between currency. Usd, and there’s this thing called humanize, which basically like

118 00:12:08.090 00:12:12.869 Uttam Kumaran: rounds to thousands, millions and billions. I usually only use one of those 2

119 00:12:14.460 00:12:17.079 Uttam Kumaran: and sometimes you if you want to use percentage. But

120 00:12:17.130 00:12:18.800 Uttam Kumaran: basically, it’s like money.

121 00:12:18.930 00:12:22.930 Uttam Kumaran: just basic rounding or percentages don’t use euros

122 00:12:23.300 00:12:30.469 Uttam Kumaran: don’t really like unless we’re doing some stuff in the timestamps. But I don’t really feel like we are. Typically, we don’t have much duration.

123 00:12:30.760 00:12:41.549 Uttam Kumaran: And then a couple of other things that will make a little bit more sense. Once we see the dashboard, there are these time zones. So you could actually specify what time zones users are allowed to switch between.

124 00:12:41.830 00:12:44.540 Uttam Kumaran: I’ve just preferred to allow for these.

125 00:12:44.870 00:12:51.309 Uttam Kumaran: Frankly, we may just be happy just doing New York. I mean I’m here. So sometimes, if you’re debugging timestamps, it helps. But

126 00:12:51.700 00:12:56.590 Uttam Kumaran: and then the other thing is available. Looking at available time zones, time ranges.

127 00:12:57.010 00:12:59.560 Uttam Kumaran: So this is, you’ll see in the dashboard as well.

128 00:12:59.610 00:13:03.049 Uttam Kumaran: Actually, let me just let me just pull this up, going through this

129 00:13:06.300 00:13:12.630 Uttam Kumaran: This is the dashboard that we’re looking at right now. Available time. Zones are here

130 00:13:12.790 00:13:18.904 Uttam Kumaran: right here, so you can see these are the ones that we specified as well as like our default ones.

131 00:13:19.650 00:13:22.210 Uttam Kumaran: you have these available time ranges

132 00:13:22.260 00:13:25.990 Uttam Kumaran: which are here, and you could see I have last 7 days.

133 00:13:26.050 00:13:28.819 Uttam Kumaran: and then comparison offsets.

134 00:13:28.970 00:13:34.209 Uttam Kumaran: I’ll kind of go into, but you could do 7 days, one month, 3 months, which is, you can see here. I have

135 00:13:34.360 00:13:39.710 Uttam Kumaran: P. 3 MP. One Mtd. Which is month to date, and p. 70.

136 00:13:39.930 00:13:45.099 Uttam Kumaran: So again, this just allows you to have those like very simple selectors already predefined

137 00:13:45.380 00:13:48.539 Uttam Kumaran: all the time we’re gonna be doing last 7 days last month.

138 00:13:48.740 00:13:53.950 Uttam Kumaran: 24 months ago, 12 months ago, like, you can super easily put that in here.

139 00:13:53.980 00:13:56.989 Uttam Kumaran: Click last 3 months to basically get everything filtered to it.

140 00:13:58.710 00:14:06.340 Uttam Kumaran: the comparison offsets are here, which is basically like ability to go great. I wanna compare everything I see here by time.

141 00:14:06.860 00:14:08.702 Uttam Kumaran: And I want to look at

142 00:14:09.060 00:14:13.649 Uttam Kumaran: like your custom rate. You’re like you wanted, basically, like, okay, I wanna look at the previous year.

143 00:14:13.800 00:14:16.990 Uttam Kumaran: So for example, the last 3 months

144 00:14:17.500 00:14:20.959 Uttam Kumaran: range. I have a comparison offset of previous year.

145 00:14:21.220 00:14:30.600 Uttam Kumaran: So if I wanted to say, I also want to include previous month previous 3 months, I could do that. So what this is doing is, it’s saying, cool. Look at the last 3 months. March 1st to June first, st

146 00:14:30.630 00:14:32.009 Uttam Kumaran: and then also look at

147 00:14:32.290 00:14:35.380 Uttam Kumaran: March 1st to June 1st of 2023.

148 00:14:35.540 00:14:38.850 Uttam Kumaran: We’re in 2024 for the recording. So

149 00:14:39.490 00:14:43.019 Uttam Kumaran: if I were to switch this to last one month.

150 00:14:43.170 00:14:49.459 Uttam Kumaran: I could continue to say, compare by time and look at the previous year. It’s gonna look at May versus May year over year.

151 00:14:50.320 00:14:57.700 Uttam Kumaran: I think this is pretty cool. This helps for a lot of questions. We’re gonna get from pool parts. Which is typically, how did we do versus last year?

152 00:14:57.700 00:14:58.089 Jakob Kagel: Yeah, d-.

153 00:14:58.090 00:15:00.000 Uttam Kumaran: Because they’re a very seasonal business.

154 00:15:00.270 00:15:27.060 Uttam Kumaran: They basically ask questions like this all the time. They don’t care versus last month. Necessarily, they care about same month last year, and basically same month last few years. So this is gonna be like, really powerful. And I think everybody on this call, maybe, except for Nico. But Nico’s probably learning that this is hardest shit to do like in light dash and kind of annoying to write a sequel query to basically put this all in one place visually.

155 00:15:29.010 00:15:35.270 Uttam Kumaran: continuing on this like comparison thing. And let me just go to 3 months because it just adds a little bit more data.

156 00:15:36.042 00:15:40.140 Uttam Kumaran: You can actually add on the right here

157 00:15:40.240 00:15:51.700 Uttam Kumaran: the changes between periods for any of these different dimensions. So what you’re seeing here is you’re seeing total shipping costs for shopify for the last 3 months this year versus last 3 months last year.

158 00:15:52.600 00:16:10.860 Uttam Kumaran: So think about how many layers we had to go through to answer that question. And instead, you just have it available. So what you’re seeing here is like we’re down 70% in shipping costs for this last 3 months this year versus last 3 months last year. And you can basically see that here, which is last 3 months.

159 00:16:11.150 00:16:13.429 Uttam Kumaran: You could see that this is our total shipping cost

160 00:16:13.730 00:16:15.319 Uttam Kumaran: the sum here.

161 00:16:16.160 00:16:20.369 Uttam Kumaran: And this is the 72%. You’re basically, which is a sum of all this.

162 00:16:20.670 00:16:22.110 Uttam Kumaran: Yeah. So

163 00:16:22.710 00:16:28.380 Uttam Kumaran: it just it gets you where you basically like you want. And you probably wanted to see this. It just like, Get throws it all at you.

164 00:16:30.340 00:16:33.619 Uttam Kumaran: the other things you could do is change this to absolute change.

165 00:16:33.680 00:16:37.610 Uttam Kumaran: You can also have this be percent of total, which is basically percent of total within this.

166 00:16:38.130 00:16:45.239 Uttam Kumaran: within these little boxes you can also change what you’re showing. So let’s say, I want to look at total sales.

167 00:16:45.500 00:16:52.690 Uttam Kumaran: You can see that like great. I want to say total sales. In that same period you could see that we’re down 26% on total sales

168 00:16:52.750 00:16:56.520 Uttam Kumaran: or down 52% within the same 3 month range.

169 00:16:58.540 00:17:03.489 Uttam Kumaran: the other things that are very, very hard to do normally, but really easy to do here is like.

170 00:17:03.540 00:17:06.019 Uttam Kumaran: have these days where you can basically say, like

171 00:17:06.240 00:17:15.429 Uttam Kumaran: common questions, like, How do we do on Tuesdays versus this day? I wanna look at like this specific city versus this very easy to do here.

172 00:17:15.849 00:17:19.549 Uttam Kumaran: everything real does well is on time series. Based data

173 00:17:19.569 00:17:22.090 Uttam Kumaran: doesn’t really do well on just

174 00:17:22.530 00:17:26.679 Uttam Kumaran: tables with like fax in them that don’t have a time series attached.

175 00:17:26.890 00:17:30.780 Uttam Kumaran: Most all of our data is basically a time series, anyways, which is really great.

176 00:17:30.880 00:17:39.579 Uttam Kumaran: The nice thing that you can also do. Here is, let’s say, let’s say I just want to do all time. Okay, let’s just let’s do custom thing.

177 00:17:39.840 00:17:41.900 Uttam Kumaran: Let’s do custom from

178 00:17:42.260 00:17:44.430 Uttam Kumaran: the start of this year until

179 00:17:45.270 00:17:47.190 Uttam Kumaran: what’s June, 12th

180 00:17:49.200 00:17:50.170 Uttam Kumaran: rates?

181 00:17:50.740 00:17:53.601 Uttam Kumaran: And then let’s say, I want to look at

182 00:17:53.940 00:17:56.829 Uttam Kumaran: I wanna look at Texas. And I wanna particularly look at like

183 00:17:57.352 00:18:06.090 Uttam Kumaran: let’s just start with. I just wanna look at like how text is doing year to date. Right? So let’s start here. Let’s just remove all these comparisons.

184 00:18:06.230 00:18:13.629 Uttam Kumaran: and we have no contacts of no filters. And we’re going. And this is actually the aggregation. So let’s start by week.

185 00:18:14.060 00:18:18.410 Uttam Kumaran: I wanna scroll down and I wanna go. Okay, where is my like state?

186 00:18:18.460 00:18:22.709 Uttam Kumaran: Okay, zone state, this is around for shipping.

187 00:18:22.900 00:18:28.860 Uttam Kumaran: Is there a state here? I can try to find shipping state. So let’s look at shipping state.

188 00:18:29.160 00:18:31.019 Uttam Kumaran: Let’s expand this.

189 00:18:31.530 00:18:33.679 Uttam Kumaran: and we can see all of our states here.

190 00:18:33.930 00:18:39.229 Uttam Kumaran: The other nice thing that you can really do is you basically get a ton of dimensionality right off the rip.

191 00:18:39.430 00:18:45.860 Uttam Kumaran: So not, I can see exactly what we’re looking at, which is like these metrics. You could see percent changes.

192 00:18:45.970 00:18:50.249 Uttam Kumaran: And let’s say, I want to search for Tx perfect.

193 00:18:50.400 00:18:54.299 Uttam Kumaran: great. So I can see all this stuff about Tx. But I want to filter the whole thing to Tx.

194 00:18:54.450 00:18:57.709 Uttam Kumaran: I’m gonna click on this. It’s gonna filter everything to Texas.

195 00:18:57.860 00:18:59.759 Uttam Kumaran: And I can click X here.

196 00:18:59.780 00:19:07.519 Uttam Kumaran: I can see that Texas, the only one selected, and we go back to all dimensions. Once you’re in all dimensions, everything here is actually filtered. Texas.

197 00:19:07.660 00:19:15.681 Uttam Kumaran: the orders, the customers, the days, all these values are now filtered to where is where is test? State shipping state? Is Texas.

198 00:19:16.050 00:19:21.300 Uttam Kumaran: So now it’s really just say, cool like this is our average like this looks like our average like

199 00:19:21.390 00:19:22.900 Uttam Kumaran: sales for the.

200 00:19:23.120 00:19:25.210 Uttam Kumaran: I guess that’s maybe just average like

201 00:19:25.570 00:19:26.590 Uttam Kumaran: orders.

202 00:19:26.770 00:19:33.249 Uttam Kumaran: Average order value. Probably I don’t know. This is not a good measure name, but we’re probably gonna be looking more like total sales amount.

203 00:19:33.440 00:19:37.090 Uttam Kumaran: Also. This is not. This should be probably Usd.

204 00:19:40.690 00:19:43.421 Uttam Kumaran: I’ll just let’s just keep going off. Make these changes later.

205 00:19:45.080 00:19:57.380 Uttam Kumaran: and so great. So we have everything filtered to Texas. One thing that we we could see a right off the bat. Let’s just let’s talk about Texas here. We can see that most of our sales coming from shopify. We can see that

206 00:19:57.410 00:20:11.080 Uttam Kumaran: like kind of like, there’s not a great breakdown of like when people are ordering but also a thing we can see is we can look at. Hey? Most of our Texas packages are actually zone 7,

207 00:20:11.270 00:20:12.383 Uttam Kumaran: meaning it’s

208 00:20:13.100 00:20:17.700 Uttam Kumaran: Texas is probably farthest from any of our existing warehouses.

209 00:20:17.830 00:20:39.489 Uttam Kumaran: We have a warehouse in New York. We have a warehouse in California, we warehouse in Florida. It’s clearly the farthest meaning zone 7. There’s only 8 zones, so everything’s in 6 and 7, meaning, this is probably one of the mo most expensive places to ship. And so let’s say, Okay, great. I wanna look at this versus last year. So let’s do time, and let’s do previous period. So, Jan, one to June 13th of last year.

210 00:20:39.720 00:20:43.699 Uttam Kumaran: perfect like, we now have year to day. And look, the data even goes further, because

211 00:20:45.180 00:20:48.509 Uttam Kumaran: like we don’t even have or I don’t know what is this? This is June.

212 00:20:48.810 00:20:50.940 Uttam Kumaran: Oh, this may be like 2 days delayed.

213 00:20:51.900 00:20:55.759 Uttam Kumaran: Yeah, I maybe even having like, I don’t know if the data is not refreshed. But either way.

214 00:20:56.250 00:21:00.250 Uttam Kumaran: you could just look at versus this past year. And so you could see great like

215 00:21:00.610 00:21:22.419 Uttam Kumaran: we’re about in line with what we were doing. We actually have like way more. We have way, less zone 8. So we actually improved. We’ve moved people to Zone 7 or Zone 6 looks like. But in fact, there’s still probably we need to move both of these, probably like Zone 3. So this is probably gonna get tackled by our our ticket, where we’re figuring out where their next warehouse should be.

216 00:21:22.430 00:21:26.379 Uttam Kumaran: And there’s a technical backlog for that. Yeah, so this is the exact like

217 00:21:26.670 00:21:40.510 Uttam Kumaran: through line. Just to that. Get a lot of other stuff like, look at shipments. There’s a lot of great data here. Refund reason. So hopefully, this one we replace with our like AI thing. Eventually. Another thing that would be great here to look at.

218 00:21:41.067 00:21:47.859 Uttam Kumaran: Who like what’s what? Orders by percentage are coming from pool pros versus not same period last year.

219 00:21:49.840 00:21:59.740 Uttam Kumaran: yeah. So I mean, I think, like, you guys probably get the gist, I wanna try to do one change, live and then spend the rest of the time talking about

220 00:22:00.160 00:22:03.989 Uttam Kumaran: like how we can do this for direct mail, and then kind of like hand

221 00:22:04.060 00:22:13.400 Uttam Kumaran: the rains off. So if so, this is all local. And I’m just gonna honestly, just like, get out of this. So this is kind of this is, this will be like dead if I try to open this again.

222 00:22:13.790 00:22:19.143 Uttam Kumaran: Yeah, this won’t even load. And so basically, we’re here in

223 00:22:20.850 00:22:22.489 Uttam Kumaran: We’re here in

224 00:22:22.500 00:22:24.291 Uttam Kumaran: and full parts to go.

225 00:22:25.050 00:22:26.910 Uttam Kumaran: I’m in my like

226 00:22:27.230 00:22:47.434 Uttam Kumaran: pie end, like virtual environment. And these, this is all I kind of tackled in the engineering onboarding Doc, and like how to set this up. But basically just have a good environment for every client. Right now, you can see I’m in just the main repo. I’m going to actually CD into pool parts to go slash real.

227 00:22:47.990 00:22:54.550 Uttam Kumaran: You can see I’m there. And there’s a couple of helpful commands for real. If you just type in real, you’re gonna kind of see this come up.

228 00:22:54.911 00:22:58.010 Uttam Kumaran: You’re gonna see real. So the main things we’ll be looking at is real.

229 00:22:58.380 00:23:00.420 Uttam Kumaran: We’ll start real deploy

230 00:23:00.440 00:23:12.639 Uttam Kumaran: real login. If if you’re not able, you’ll basically, it’ll tell you that you’re not logged in when you hit start. The other things that I do is I will add users and then we also have some variables for the project.

231 00:23:13.297 00:23:15.000 Uttam Kumaran: We do have

232 00:23:15.530 00:23:16.883 Uttam Kumaran: like a

233 00:23:18.890 00:23:27.579 Uttam Kumaran: We do have a live version of real up right now, and there are variables there, things like snowflake passwords and stuff.

234 00:23:27.710 00:23:32.279 Uttam Kumaran: Locally, I have like a dot em file that I use, but of course, like we don’t

235 00:23:32.310 00:23:42.409 Uttam Kumaran: we have that in our get ignore? So it doesn’t get published the repo. So we have local credentials, and the cloud version of real has credentials and nothing gets stored in Github.

236 00:23:42.490 00:23:47.239 Uttam Kumaran: So these are things that like it’s all really well documented in terms of like how to

237 00:23:47.330 00:23:54.755 Uttam Kumaran: set these things up. And I can help walk through PE people with it. So basically to to get going like, we just run real start

238 00:23:56.048 00:23:59.040 Uttam Kumaran: it’ll basically should pull this up.

239 00:24:01.640 00:24:07.699 Uttam Kumaran: Here, I’m just going to. Yep. It’s back up. And you’ll get put in. You get put in here.

240 00:24:08.452 00:24:10.967 Uttam Kumaran: The nice thing is we

241 00:24:11.820 00:24:14.810 Uttam Kumaran: are going to look at bringing in a new

242 00:24:15.179 00:24:21.129 Uttam Kumaran: source today. So let’s let’s just say we want to bring in the direct mail source.

243 00:24:21.140 00:24:22.325 Uttam Kumaran: So let me.

244 00:24:23.340 00:24:30.170 Uttam Kumaran: I don’t think we, I think, don’t think we even have like us. We don’t have really like. It’s not a connector that we have set up.

245 00:24:34.530 00:24:43.149 Uttam Kumaran: yeah. So okay, let’s just do this. Let’s do it the way we’ve do. They’ve kind of this is kind of like brand new. So I may need to look into like. Maybe I can bring in all the snowflake stuff here. But

246 00:24:45.020 00:24:47.740 Uttam Kumaran: let’s just show everybody where

247 00:24:48.480 00:24:50.369 Uttam Kumaran: this stuff is and snowflake

248 00:24:51.570 00:24:53.040 Uttam Kumaran: one sec.

249 00:24:53.190 00:24:54.509 Uttam Kumaran: The questions.

250 00:24:54.620 00:24:57.439 Uttam Kumaran: So far, I’m just gonna navigate to the direct meal stuff.

251 00:24:57.440 00:25:05.570 Jakob Kagel: Yeah, I guess like. So I mean, obviously, you’re very familiar with real like is, who else on the team is familiar.

252 00:25:07.390 00:25:11.530 Uttam Kumaran: It’s probably me, Ryan and Patrick. No good amount.

253 00:25:12.476 00:25:16.629 Uttam Kumaran: It’s probably me and Patrick can probably, like Patrick is gonna be

254 00:25:16.730 00:25:20.200 Uttam Kumaran: really gnarly at any of these connection type situations.

255 00:25:20.200 00:25:21.270 Jakob Kagel: Okay. Cool.

256 00:25:21.270 00:25:22.320 Uttam Kumaran: I am.

257 00:25:22.850 00:25:27.559 Uttam Kumaran: I like I really like the tool. But Patrick’s probably like the go to, I would say.

258 00:25:27.560 00:25:45.609 Jakob Kagel: I mean it makes sense, I mean, I kind of feel the same way. I mean all the filters and everything like are great. I mean, you know, it’s a lot better than light dash already, like, in my opinion. But yeah, I kind of feel the same way, like about evidence is this where it’s like? It’s a lot we gotta do back end versus like, you know.

259 00:25:45.910 00:25:49.389 Jakob Kagel: But it’s all good. I mean, I’m I’m with it.

260 00:25:50.260 00:25:57.629 Uttam Kumaran: Yeah, I think it. I think it’s definitely gonna take a little bit longer. But again, the the kind of the thing I was talking to Jacob yesterday was about, which is mainly like.

261 00:25:58.010 00:26:10.459 Uttam Kumaran: if we can get it in a view in in real. And then basically say, like, you guys can go get it. It allows Kim and these guys to somehow run by themselves. But also, like I actually find it helpful when I’m answering questions for them.

262 00:26:10.520 00:26:27.269 Uttam Kumaran: Because, like I typically wanna cut things like 90 ways. And I wanna like, I try to go really quickly. And it’s you guys know, it’s hard to write these like queries, really fast and kinda like, answer the next thing, and I don’t know whether like our end analysis lives here. But

263 00:26:27.630 00:26:37.600 Uttam Kumaran: I think this is a little bit more helpful than like just going in raw into snowflake and writing stuff, but I also think there is some times where, like we don’t have everything in there. So

264 00:26:38.010 00:26:45.080 Uttam Kumaran: I think in that situation, if there should be a process by which Jacob, you’re basically like, Hey, these things don’t exist there. I still need to move forward.

265 00:26:45.360 00:26:49.380 Uttam Kumaran: And so we there’s both a ticket to like, add it to real. And there’s like

266 00:26:49.390 00:26:54.490 Uttam Kumaran: you can continue, basically. And then, once you have piece of logic, you’re basically like, Hey, here, this needs to get into

267 00:26:55.000 00:26:56.260 Uttam Kumaran: grill somehow.

268 00:26:56.500 00:26:58.229 Jakob Kagel: Okay, yeah, makes sense.

269 00:26:58.810 00:27:00.241 Uttam Kumaran: So let me just

270 00:27:02.190 00:27:04.319 Uttam Kumaran: just gonna call this directly

271 00:27:08.020 00:27:09.879 Uttam Kumaran: and then so let’s look at

272 00:27:10.550 00:27:17.490 Uttam Kumaran: Dbt, we’re gonna look at Dbtmart. I don’t know where direct mail is. Okay. Got lucky here.

273 00:27:20.220 00:27:24.800 Uttam Kumaran: it’s in this file. So if we just select Star from

274 00:27:25.460 00:27:29.880 Uttam Kumaran: this, you should see all of our great direct mail stuff.

275 00:27:31.640 00:27:32.870 Uttam Kumaran: and

276 00:27:33.420 00:27:36.580 Uttam Kumaran: let’s go ahead and start to create our 1st source. So we.

277 00:27:36.580 00:27:40.350 Jakob Kagel: Sorry in the table, real quick in the direct mail table. Can you just

278 00:27:40.420 00:27:43.979 Jakob Kagel: can we look and see? Do we have order number in this Table

279 00:27:44.560 00:27:47.529 Jakob Kagel: cause? That was a question we had like in the standup.

280 00:27:47.970 00:27:49.418 Uttam Kumaran: Yeah, I’m pretty sure.

281 00:27:49.780 00:28:01.145 Jakob Kagel: Old. She had order number, like in the Csv. That she sent over to us the other day, or he’s basically order reference Id, but it mapped our order number like I checked it.

282 00:28:02.790 00:28:03.319 Jakob Kagel: yes, and.

283 00:28:03.320 00:28:05.179 Uttam Kumaran: Go look at like the actual.

284 00:28:05.180 00:28:05.750 Jakob Kagel: Right.

285 00:28:05.750 00:28:08.831 Uttam Kumaran: Yeah, if we go look at the actual logic for

286 00:28:12.040 00:28:13.170 Uttam Kumaran: for this

287 00:28:17.440 00:28:19.500 Uttam Kumaran: we go to direct mail.

288 00:28:22.910 00:28:24.440 Uttam Kumaran: go there, please.

289 00:28:26.510 00:28:27.155 Uttam Kumaran: Hello!

290 00:28:36.520 00:28:38.110 Uttam Kumaran: What is going on.

291 00:28:43.110 00:28:44.740 Uttam Kumaran: Let’s just go to the folder.

292 00:28:47.550 00:28:48.490 Uttam Kumaran: Nothing.

293 00:28:51.730 00:28:53.840 Uttam Kumaran: Okay? Well, there it is.

294 00:28:54.410 00:28:56.050 Uttam Kumaran: Models.

295 00:28:56.540 00:28:57.620 Uttam Kumaran: Lawrence

296 00:28:58.540 00:29:00.070 Uttam Kumaran: marketing.

297 00:29:00.830 00:29:02.960 Uttam Kumaran: not in there. And

298 00:29:07.280 00:29:09.300 Uttam Kumaran: he’s just open, please.

299 00:29:11.260 00:29:15.950 Uttam Kumaran: Great. So yeah, here’s basically, it so looks like we’re just selecting

300 00:29:16.700 00:29:19.529 Uttam Kumaran: day direct mail campaign week of your

301 00:29:19.710 00:29:22.803 Uttam Kumaran: sends cost and revenue roas from Google sheets.

302 00:29:23.460 00:29:26.349 Uttam Kumaran: This Google sheets is coming from

303 00:29:27.240 00:29:30.519 Uttam Kumaran: parts. Google sheets are.

304 00:29:31.790 00:29:32.750 Uttam Kumaran: yeah, there’s

305 00:29:33.630 00:29:39.328 Uttam Kumaran: this Google sheets has, like, basically all the Google sheets we’re bringing in. And there’s this direct mail thing. And here’s basically all the data we’re getting.

306 00:29:40.480 00:29:45.710 Uttam Kumaran: This is coming from a this is coming from an output. We’re getting from

307 00:29:46.260 00:29:47.520 Uttam Kumaran: postpilot.

308 00:29:47.640 00:29:53.486 Uttam Kumaran: And so they’re probably there may be it. We pro. We may need to have to modify what we’re getting from them for this exercise.

309 00:29:53.900 00:29:54.250 Jakob Kagel: Yeah.

310 00:29:54.650 00:29:55.450 Uttam Kumaran: Order id.

311 00:29:55.697 00:30:01.040 Uttam Kumaran: but we have a we have a through line to their it. So this is something that, Brian. You could probably just let me know, and I’ll

312 00:30:01.570 00:30:02.580 Uttam Kumaran: hook you up with that.

313 00:30:02.580 00:30:11.699 Nicolas Sucari: Probably we should. We should also ask him like, where is she getting that? Csv, see if she’s like extracting it directly from pro post pilot, or from where.

314 00:30:11.700 00:30:15.481 Uttam Kumaran: She’s probably getting it directly from them. The problem is.

315 00:30:16.000 00:30:18.579 Uttam Kumaran: we can we could do it one time. But

316 00:30:18.840 00:30:24.680 Uttam Kumaran: I, basically, we have. We now have access to the it. People at Post Pilot, and they were like we could set up any report you want.

317 00:30:25.107 00:30:28.912 Uttam Kumaran: This is the 1st one we set up because we need this for hit crazy dashboard.

318 00:30:29.240 00:30:34.370 Uttam Kumaran: Now that we need order Id probably like something better we can get from them.

319 00:30:34.660 00:30:35.840 Uttam Kumaran: But actually, I think.

320 00:30:36.100 00:30:40.700 Uttam Kumaran: Jacob, for your question. It’s helpful to see? Like, yeah, okay, we don’t have it. Here’s the logic. And then

321 00:30:41.350 00:30:42.410 Uttam Kumaran: theory spaces like that.

322 00:30:42.410 00:30:50.429 Nicolas Sucari: This is. This is all the manual. This is all the manual work that Ryan does like to load. All of those emails that we are receiving through slack.

323 00:30:50.980 00:30:51.730 Uttam Kumaran: Yes.

324 00:30:52.830 00:30:53.430 Nicolas Sucari: Okay.

325 00:30:55.020 00:30:57.889 Uttam Kumaran: Which I know there’s a ticket there for that right now.

326 00:30:57.890 00:30:58.410 Nicolas Sucari: Yeah, yeah.

327 00:30:58.639 00:31:04.139 Uttam Kumaran: So let’s just talk about bringing what we have for direct mail right now. Looks like there’s probably more stuff we need to do

328 00:31:04.150 00:31:07.320 Uttam Kumaran: so for sources. I’m gonna go ahead and

329 00:31:07.640 00:31:09.299 Uttam Kumaran: let’s go back to here.

330 00:31:10.610 00:31:15.049 Uttam Kumaran: you can do. You can add it right here. You can add it right here. I’m just gonna go ahead and do it here.

331 00:31:16.095 00:31:17.010 Uttam Kumaran: And

332 00:31:18.290 00:31:20.990 Uttam Kumaran: let’s go ahead and do new file.

333 00:31:21.140 00:31:23.899 Uttam Kumaran: Let’s call it direct mail.

334 00:31:24.820 00:31:27.929 Uttam Kumaran: Yeah. Well, we don’t have naming conventions yet for this

335 00:31:28.580 00:31:32.090 Uttam Kumaran: we probably will eventually. But for now it’s okay.

336 00:31:32.685 00:31:38.264 Uttam Kumaran: I’m just gonna select star from Dbt. Dot Dvt. Mart direct mail

337 00:31:39.220 00:31:40.820 Uttam Kumaran: and hit save

338 00:31:41.540 00:31:43.470 Uttam Kumaran: we should see it pop up

339 00:31:43.660 00:31:44.780 Uttam Kumaran: here.

340 00:31:47.610 00:31:50.679 Uttam Kumaran: Something happened incorrectly. Oh.

341 00:31:51.190 00:31:53.499 Uttam Kumaran: I totally forgot to do this stuff.

342 00:31:56.430 00:32:00.930 Uttam Kumaran: So yes, you do need these kind of like format strings.

343 00:32:01.390 00:32:03.180 Uttam Kumaran: Let’s go back here.

344 00:32:03.880 00:32:09.630 Uttam Kumaran: as you can see how wonderful brings in all these columns with little helpful little things.

345 00:32:09.700 00:32:11.399 Uttam Kumaran: Let’s go to models.

346 00:32:11.730 00:32:14.089 Uttam Kumaran: Let’s go ahead and

347 00:32:14.743 00:32:18.460 Uttam Kumaran: let’s go ahead and do it here for me.

348 00:32:18.560 00:32:23.360 Uttam Kumaran: I’m gonna go hit new files. I’m gonna do direct mail

349 00:32:24.030 00:32:25.000 Uttam Kumaran: model.

350 00:32:28.420 00:32:31.360 Uttam Kumaran: Pretty sure here. You don’t really need to do anything except

351 00:32:31.680 00:32:33.980 Uttam Kumaran: right to select so select

352 00:32:36.440 00:32:38.719 Uttam Kumaran: direct mail.

353 00:32:40.680 00:32:43.159 Uttam Kumaran: save that’s auto complete.

354 00:32:43.670 00:32:45.960 Uttam Kumaran: You should see the direct mail model here.

355 00:32:46.010 00:32:47.490 Uttam Kumaran: All the columns.

356 00:32:47.980 00:32:49.220 Uttam Kumaran: All looks great.

357 00:32:50.105 00:33:06.344 Uttam Kumaran: The nice thing that you can do from here is there is a really helpful generate dashboard with AI. This is the best part of this is that you don’t have to write any of those. You don’t have to basically write any of the dashboard code for the 1st go around. Probably just need to make modifications.

358 00:33:09.610 00:33:10.240 Jakob Kagel: This is.

359 00:33:10.240 00:33:10.670 Uttam Kumaran: Pop up.

360 00:33:10.670 00:33:13.160 Jakob Kagel: Okay with the key metrics, or whatever like it can.

361 00:33:13.160 00:33:18.509 Uttam Kumaran: Yeah, basically, it looks at each column is like, this is a dollar. This is like a metric. This is a string

362 00:33:18.550 00:33:19.670 Uttam Kumaran: creates that.

363 00:33:21.720 00:33:24.720 Uttam Kumaran: I’m sure it created, and it’s just so.

364 00:33:24.720 00:33:26.950 Jakob Kagel: Once we have this data source, like

365 00:33:27.070 00:33:43.989 Jakob Kagel: as like select star from direct mail, we just now, you just basically need to just make the underlying change. Like to the direct mail table. Right? I guess. I mean, we’d also have to probably join for what she wants, like we have to join the all order items. So we can actually get the products to. Yeah.

366 00:33:44.490 00:33:56.229 Uttam Kumaran: That’s correct. Yeah. So it’s probably sub direct. So Brian, for you, I mean, you’re probably following. But say, yeah, it’s probably something you need to make either make a direct mail ag table, which is what exists now.

367 00:33:56.550 00:34:01.780 Uttam Kumaran: and probably have, like a better direct mail source table that, like brings the orders into one place.

368 00:34:02.482 00:34:03.810 Uttam Kumaran: That didn’t work.

369 00:34:04.410 00:34:08.920 Uttam Kumaran: I’m gonna try one more time. If not, I’m gonna just close this and bring it back up.

370 00:34:09.270 00:34:17.140 Jakob Kagel: So when you’re creating like the data sources in real like, are you always just doing select star from the table? Or are you ever going to do like

371 00:34:17.530 00:34:26.150 Jakob Kagel: the custom sequel kind of like as the data source you just always want to make like the underlying table 1st and then do select stars, that sort of best practice.

372 00:34:26.159 00:34:32.489 Uttam Kumaran: Yeah, it’s they do have like some opinionated things about how

373 00:34:32.729 00:34:36.119 Uttam Kumaran: like, what a source is versus what a model is.

374 00:34:36.120 00:34:36.520 Jakob Kagel: Sure.

375 00:34:36.520 00:34:39.180 Uttam Kumaran: Basically they try to say that like

376 00:34:42.120 00:34:50.409 Uttam Kumaran: I like I. I asked them about this and discord. They didn’t really give me a great answer. I think, basically, for our situation.

377 00:34:50.489 00:34:56.010 Uttam Kumaran: I would prefer that all logic stays in the warehouse, and I think, Brian, you probably agree with me.

378 00:34:56.060 00:34:58.020 Uttam Kumaran: The nice thing is is like.

379 00:34:58.060 00:35:05.179 Uttam Kumaran: then, yeah, the nice thing is is like we can join 2 sources together. For example, let’s say we we don’t want to have like.

380 00:35:05.760 00:35:11.359 Uttam Kumaran: let’s say we just want to have all orders here, and you want to join to something. And you want to do that here.

381 00:35:11.420 00:35:15.300 Uttam Kumaran: I would you could do that here. But in terms of like

382 00:35:16.230 00:35:28.290 Uttam Kumaran: heavy like business logic, I prefer that to all. Be in the warehouse. But let’s say like we don’t. You don’t need to create a brand new source where you do the join. The join can happen directly here.

383 00:35:28.440 00:35:30.949 Uttam Kumaran: And the other thing that’s nice is like

384 00:35:31.180 00:35:33.920 Uttam Kumaran: real actually isn’t hitting the warehouse

385 00:35:34.040 00:35:42.270 Uttam Kumaran: every time we query something. It hits the warehouse brings all this data into that in memory database, and then handles any joins you want.

386 00:35:42.890 00:35:51.040 Uttam Kumaran: So maybe, like I’m again like, maybe, Jacob, we kind of like, once we find a situation, we need to make a decision. We all kind of like come to a conclusion on like what’s the best route

387 00:35:51.050 00:35:54.490 Uttam Kumaran: for now, I think, having these straight selects is fine, I think.

388 00:35:54.490 00:35:54.950 Jakob Kagel: The.

389 00:35:54.950 00:36:01.569 Uttam Kumaran: How we did refunds and evidence. Maybe you do need to do some joins right here or like create some little aggregates.

390 00:36:01.720 00:36:06.339 Uttam Kumaran: If the aggregates don’t need to go anywhere else. Then I I think that’s fine to make them here.

391 00:36:06.941 00:36:08.188 Uttam Kumaran: I probably just like

392 00:36:09.150 00:36:12.390 Uttam Kumaran: message, Brian, or misses the Channel, and just confirm, but

393 00:36:12.450 00:36:13.658 Uttam Kumaran: probably not that bad.

394 00:36:14.880 00:36:20.950 Brian Pei: I. I tend to think of things like this, even though this is my 1st time using real. But any

395 00:36:21.190 00:36:24.440 Brian Pei: bi tool that can

396 00:36:24.550 00:36:25.890 Brian Pei: store.

397 00:36:26.640 00:36:29.470 Brian Pei: What is it like? Extra SQL or

398 00:36:29.730 00:36:31.890 Brian Pei: SQL. For use cases?

399 00:36:32.460 00:36:34.650 Brian Pei: if it’s for simple

400 00:36:35.010 00:36:39.129 Brian Pei: aggregates and functions is fine, because

401 00:36:39.400 00:36:40.370 Brian Pei: in

402 00:36:40.520 00:36:46.619 Brian Pei: Dbt, in in Snowflake we’re not gonna create like 8 aggregate tables on top of like.

403 00:36:46.860 00:37:05.590 Brian Pei: I don’t know direct mail or or whatever it is. That’s the bi tools job is to take specific columns and do averages and sums by whatever dimensions that you want. When it gets to. If you need a sum where it’s like a sum if so I only wanna get

404 00:37:05.660 00:37:16.319 Brian Pei: direct direct mail for for a State, or for an invoice type, or anything like that. If you save it, and it becomes a really

405 00:37:16.520 00:37:28.593 Brian Pei: important metric for the business in just the real layer, and we don’t have it in the snowflake layer. Then it’s confined to live and be edited in real

406 00:37:29.230 00:37:32.009 Brian Pei: where we would like it to be in snowflake

407 00:37:32.280 00:37:43.259 Brian Pei: for any cases where they don’t use real or for analysts who are just looking in snowflake. They don’t have access to a column that has kind of like extra business logic. That’s important to the client

408 00:37:43.803 00:37:47.080 Brian Pei: and then I think Utah mentioned this before. But even things like.

409 00:37:47.290 00:37:51.534 Brian Pei: Oh, for this join to work. I need to trim the white space and and lower

410 00:37:51.870 00:38:02.467 Brian Pei: the it all has to be lower case. You can do that in real to make a join work. But then, in snowflake it wouldn’t work. It would probably need the same logic.

411 00:38:02.940 00:38:07.479 Brian Pei: and yeah, and then it depends on speed of development. If you really need it

412 00:38:07.570 00:38:30.220 Brian Pei: in real, in like a day. It’s easier to do it on bi layer, and then you can just tell Nico or somebody, or myself, or an ae like, hey? I I needed to get this join to work. There’s white space. I need replace strings. I’m doing it in real now, but maybe create a ticket to do it in the warehouse, so that I don’t have to do it. In the bi layer is also, I think, okay.

413 00:38:32.500 00:39:02.009 Nicolas Sucari: Yeah, I have a question. Like, why should like Jacob would like be using a lot more of querying like for that like analytic stuff here? Or should he like still keep using, like the data warehouse and snowflake, to like some like other analysis? And that stuff like, because if we have these is what just, Brian was saying, like we can do like the analysis here without hitting the data warehouse. And then, if we see that something is like really needed to.

414 00:39:02.080 00:39:07.950 Nicolas Sucari: Yeah. And we need to have that even persisted in the at the warehouse. We can then like add the change there, right.

415 00:39:08.530 00:39:11.249 Uttam Kumaran: Yeah, I have a couple of opinions on. This one is like.

416 00:39:11.330 00:39:13.629 Uttam Kumaran: it’s a balance one. I I don’t want.

417 00:39:13.820 00:39:22.720 Uttam Kumaran: I want to leverage the tools for analysis. Because this is what we’re putting in front of customers for them to do. Analysis. We’re not able to use these tools.

418 00:39:22.840 00:39:51.110 Uttam Kumaran: Then we’re Doa, we’re we’re having that problem with light dash where it’s kind of ass. And so like, we’re not able to use that. And so I’m glad that we’re finding these new tools, and we’re able to push these. On the other hand, everybody has their own workflows. And so I know we got into situation with evidence where, you know, I was like, Hey, Jacob, can we do everything in evidence? And it just took a long time, and evidence isn’t geared for a data expiration. I think this tool is a little bit more geared for data exploration. But I also want people to have access to whatever they need.

419 00:39:51.130 00:40:05.179 Uttam Kumaran: There’s difference between getting to the answer and delivering it to the client that I wanna make a clear distinction on getting to the answer and having the discussion internally, just like as fast as we need to. It could be, whatever. But when we want to hand it over to the client.

420 00:40:05.477 00:40:23.259 Uttam Kumaran: It’s all, of course, client by client, but it is really nice when we can say, Hey, we got to the answer. And also here’s how we found it, using the tool that we’ve we’re we’re in. We’ve implemented for you, and also, for you know, for folks like Kim and Chuck, who are, who can be, you know in data and use it.

421 00:40:23.260 00:40:41.777 Uttam Kumaran: I would like to try and empower them right? This is always a common thing that we’re gonna have where it’s a mix of. Do we do the work to Dave do the work it’s always gonna come back to. We’ll handle it. But I want to share with them that it’s possible here. But there’s gonna be stuff we can’t do in here. There’s gonna be stuff that doesn’t need to be done in the year that can be done quicker. So that’s up to.

422 00:40:42.000 00:40:44.510 Uttam Kumaran: That’s up to Jacob, and anybody who’s doing the analysis.

423 00:40:44.510 00:40:44.840 Nicolas Sucari: Perfect.

424 00:40:44.840 00:40:45.670 Uttam Kumaran: Judgment.

425 00:40:45.790 00:40:57.319 Uttam Kumaran: The thing that Brian made a point about which is when there’s logic that needs to get used downstream when there’s actual kpis and definitions like, for example, the definition of profit should not live here, should live in the warehouse.

426 00:40:57.320 00:40:57.850 Nicolas Sucari: Yeah, so, so.

427 00:40:57.850 00:41:09.438 Uttam Kumaran: Because we’re gonna have alerting that’s going to be used elsewhere. So it’s always an ebb and flow but again, like we’re all senior here, like, you guys know, like the trade off. So,

428 00:41:10.111 00:41:14.029 Uttam Kumaran: yeah, I think we’re all pretty aware of like how to handle this

429 00:41:14.040 00:41:15.469 Uttam Kumaran: to continue on this direct mail.

430 00:41:15.470 00:41:17.920 Jakob Kagel: I totally agree with you. Sorry. Yeah. Go ahead.

431 00:41:17.920 00:41:18.550 Uttam Kumaran: Cool

432 00:41:19.330 00:41:25.324 Uttam Kumaran: to correct. To come. To continue this direct mail piece, I press that. AI button. I got all this great stuff.

433 00:41:25.760 00:41:28.590 Uttam Kumaran: I immediately kind of remove these

434 00:41:29.557 00:41:32.599 Uttam Kumaran: like all of these time zones.

435 00:41:32.620 00:41:41.389 Uttam Kumaran: and you’ll see that like these time ranges are kind of od but I want to just walk through what it created. It created this type called metrics view.

436 00:41:41.450 00:41:45.439 Uttam Kumaran: actually don’t know. Are there other? Are there other types of dashboards?

437 00:41:49.140 00:41:49.810 Jakob Kagel: Must be.

438 00:41:49.810 00:41:54.399 Uttam Kumaran: Okay, I think they’re. I think they’re creating some other dashboards. So they added this because

439 00:41:54.560 00:41:56.119 Uttam Kumaran: I honestly didn’t even

440 00:41:57.230 00:42:02.887 Uttam Kumaran: no, that this was there before. But but they’re changing this product like almost like every week or so.

441 00:42:03.645 00:42:07.359 Uttam Kumaran: Which is great cause it gets better. Better. So direct mail model.

442 00:42:07.870 00:42:12.781 Uttam Kumaran: I don’t really like mail model. It’s kind of sauce, but I guess I’ll just say direct meal.

443 00:42:13.617 00:42:18.720 Uttam Kumaran: date Channel campaign name. Looks like there’s only 2

444 00:42:19.516 00:42:25.910 Uttam Kumaran: week number. It’s not probably like not really that necessary, but looks like we have total records.

445 00:42:25.960 00:42:28.250 Uttam Kumaran: Honestly, that’s probably don’t need

446 00:42:28.678 00:42:34.941 Uttam Kumaran: just account. Star cost. I want this to be currency. Usd. Oh, look! They added, like little

447 00:42:35.770 00:42:39.440 Uttam Kumaran: auto complete summer revenue. Let’s make this currency. Usd

448 00:42:40.594 00:42:42.410 Uttam Kumaran: and some of Roas.

449 00:42:42.850 00:42:45.680 Uttam Kumaran: Honestly, we don’t need a sum of roas.

450 00:42:45.850 00:42:48.710 Uttam Kumaran: Maybe average of Roaz is a better

451 00:42:50.920 00:42:52.070 Uttam Kumaran: and

452 00:42:53.510 00:42:57.780 Uttam Kumaran: campaign name is already there. Revenue. So revenue sends.

453 00:42:57.970 00:42:59.930 Uttam Kumaran: I also want to look at sends.

454 00:43:00.030 00:43:01.726 Uttam Kumaran: So let’s look at

455 00:43:05.640 00:43:06.540 Uttam Kumaran: I should leave

456 00:43:07.580 00:43:12.279 Uttam Kumaran: to even do another demo. Let’s go to the direct. Let’s go to the direct mail model. Here

457 00:43:12.490 00:43:13.830 Uttam Kumaran: I want to look at

458 00:43:16.080 00:43:17.030 Uttam Kumaran: name.

459 00:43:17.200 00:43:19.109 Uttam Kumaran: This sends

460 00:43:20.470 00:43:21.620 Uttam Kumaran: and

461 00:43:22.430 00:43:25.010 Uttam Kumaran: use AI to auto complete that

462 00:43:25.430 00:43:27.549 Uttam Kumaran: that makes this way easier.

463 00:43:27.700 00:43:28.460 Uttam Kumaran: And

464 00:43:31.780 00:43:35.195 Uttam Kumaran: yeah, I guess I could also just add the week number here.

465 00:43:45.300 00:43:53.525 Uttam Kumaran: Cool. Okay, we should see that I’ll pop up here if there’s not like if there’s an issue. For example, like, if I if I ditch this.

466 00:43:54.390 00:43:57.100 Uttam Kumaran: you’re gonna see, it’s gonna scream.

467 00:43:59.020 00:44:07.560 Uttam Kumaran: so fix that. This is honestly, I’ve been stuck trying to find stuff like that. This is where it’s probably helpful to call me or call Patrick.

468 00:44:07.560 00:44:07.960 Jakob Kagel: Sorry.

469 00:44:07.960 00:44:11.956 Uttam Kumaran: So like. Look at like what we need to do. Because

470 00:44:12.300 00:44:18.000 Uttam Kumaran: I’ve been stuck for like hours trying to fix like one small thing, and their error reporting authority gone better in the last 2 months.

471 00:44:18.000 00:44:18.430 Jakob Kagel: Yeah.

472 00:44:18.430 00:44:19.060 Uttam Kumaran: But

473 00:44:19.170 00:44:20.770 Uttam Kumaran: it’s not worth it like.

474 00:44:20.770 00:44:21.590 Jakob Kagel: Not for sure.

475 00:44:21.730 00:44:25.119 Uttam Kumaran: Yeah, it’s gonna be deadly. So we go to preview.

476 00:44:25.140 00:44:31.569 Uttam Kumaran: And then you have it. Right? So we went from creating a source creating a model creating a dashboard.

477 00:44:31.730 00:44:46.669 Uttam Kumaran: adding what we needed to add. And then, basically, here we are. So you can see, like we didn’t do direct mail. We did some direct mail in the past. Kim’s really just ramped up. Looks like in the past year. So let’s and you could

478 00:44:46.690 00:44:53.379 Uttam Kumaran: you see, these are all the really helpful time ranges, we added. So let’s say, I want to just look at the last 12 months.

479 00:44:53.600 00:44:59.790 Uttam Kumaran: and this is on. Let’s say I wanna share the East Coast time. Great. We only have one channel.

480 00:44:59.820 00:45:02.789 Uttam Kumaran: We want to look at our campaigns, and you can look at sum of cost.

481 00:45:03.130 00:45:04.950 Uttam Kumaran: Go with that summer revenue

482 00:45:05.770 00:45:07.360 Uttam Kumaran: again for Kim.

483 00:45:07.490 00:45:12.270 Uttam Kumaran: I don’t know how long it took her previously to. Just look at you know what’s my best campaign that we’ve ran

484 00:45:12.656 00:45:21.020 Uttam Kumaran: but now it’s way easier. You can also add a context calm, which is like percent of total. You can say, like great like, these are attribute to like this percent of the total

485 00:45:21.440 00:45:28.820 Uttam Kumaran: week number. It’s she just needed it for her reporting. It’s probably not so necessary here. But you can see like total sends

486 00:45:28.840 00:45:30.430 Uttam Kumaran: our average row as

487 00:45:31.200 00:45:35.500 Uttam Kumaran: total revenue. And so let’s say, we want to change the summer revenue. Is not that great?

488 00:45:35.520 00:45:38.554 Uttam Kumaran: I just wanna do revenue. Here

489 00:45:40.560 00:45:42.220 Uttam Kumaran: we do. Total revenue

490 00:45:42.340 00:45:43.030 Uttam Kumaran: couple.

491 00:45:44.480 00:45:48.177 Uttam Kumaran: Bam changed. Let’s just do total cost here

492 00:45:49.580 00:45:50.440 Uttam Kumaran: Bam

493 00:45:50.750 00:45:53.189 Uttam Kumaran: changed. And then let’s do

494 00:45:54.000 00:45:56.258 Uttam Kumaran: email sentence for for direct

495 00:45:57.330 00:45:58.400 Uttam Kumaran: sense.

496 00:45:58.990 00:46:03.600 Uttam Kumaran: Great average of roas. Let’s make this average

497 00:46:04.590 00:46:05.370 Uttam Kumaran: class

498 00:46:08.240 00:46:13.059 Uttam Kumaran: great, and this has to be a dollar sign. So we’ll do currency. That’s the

499 00:46:13.610 00:46:14.500 Uttam Kumaran: great.

500 00:46:15.330 00:46:22.590 Uttam Kumaran: very nice. So then this is all like, basically set you can then deploy this.

501 00:46:23.490 00:46:26.260 Uttam Kumaran: But but we haven’t actually pushed this stuff yet.

502 00:46:26.300 00:46:30.329 Uttam Kumaran: So what I would do is, I would basically control C,

503 00:46:30.450 00:46:35.759 Uttam Kumaran: and that, let’s we have like a couple of minutes. Let me just let’s just walk through that process real quick.

504 00:46:36.960 00:46:40.426 Uttam Kumaran: Just doing this job on like speed. Run.

505 00:46:42.000 00:46:44.529 Uttam Kumaran: let’s just have Github open.

506 00:46:45.710 00:46:47.039 Uttam Kumaran: Yes, I do.

507 00:46:49.120 00:46:49.790 Uttam Kumaran: Awesome.

508 00:46:51.990 00:46:53.280 Uttam Kumaran: Let’s open.

509 00:46:54.470 00:46:55.970 Uttam Kumaran: Do the desktop.

510 00:46:57.690 00:47:01.539 Uttam Kumaran: I’m just hoping. Let’s continue to keep this other stuff open.

511 00:47:08.220 00:47:12.359 Uttam Kumaran: Okay, great. And does this look like super weird? I have 3 of these open.

512 00:47:17.400 00:47:21.449 Nicolas Sucari: I think that’s fine, Github. The window is kind of small, but that’s okay.

513 00:47:21.970 00:47:25.520 Uttam Kumaran: Okay. So I just wanna go to the actual

514 00:47:25.780 00:47:27.300 Uttam Kumaran: ticket for this.

515 00:47:27.700 00:47:29.279 Uttam Kumaran: which is on

516 00:47:29.950 00:47:31.650 Uttam Kumaran: Brian’s plate.

517 00:47:33.038 00:47:37.750 Uttam Kumaran: This is something to do with direct mail.

518 00:47:39.300 00:47:39.730 Nicolas Sucari: 16.

519 00:47:39.730 00:47:41.949 Uttam Kumaran: Yeah, yeah, this one.

520 00:47:45.400 00:47:55.960 Uttam Kumaran: Cool. So what I’m gonna do is actually, this is probably the law. I’m just gonna do it to this one, bring direct mail conversions into real. So I’m gonna go here. Let me go create a branch.

521 00:47:57.940 00:48:01.869 Uttam Kumaran: I’m gonna go to this. But it’s probably not gonna work.

522 00:48:03.760 00:48:07.000 Uttam Kumaran: Yeah. Okay, so let’s refresh this.

523 00:48:09.010 00:48:13.679 Uttam Kumaran: This is 5, 76. So I want to switch to 5, 7, 6.

524 00:48:13.920 00:48:16.920 Uttam Kumaran: There, it’s gonna ask me to bring my changes.

525 00:48:17.280 00:48:18.350 Uttam Kumaran: great

526 00:48:18.660 00:48:19.889 Uttam Kumaran: switch branch.

527 00:48:20.730 00:48:23.690 Uttam Kumaran: adding direct mail

528 00:48:23.830 00:48:26.120 Uttam Kumaran: to Mel.

529 00:48:26.870 00:48:27.912 Uttam Kumaran: So it’s not

530 00:48:28.370 00:48:29.240 Uttam Kumaran: versions

531 00:48:30.820 00:48:31.670 Uttam Kumaran: I have

532 00:48:32.870 00:48:33.943 Uttam Kumaran: order. So

533 00:48:35.050 00:48:36.500 Uttam Kumaran: commit that

534 00:48:36.640 00:48:38.250 Uttam Kumaran: we’re gonna push, that

535 00:48:38.690 00:48:40.840 Uttam Kumaran: I’m going to

536 00:48:41.620 00:48:43.529 Uttam Kumaran: create a pull request.

537 00:48:45.430 00:48:49.229 Uttam Kumaran: just gonna open a different tab.

538 00:48:49.590 00:48:51.159 Uttam Kumaran: Bring that here.

539 00:48:52.750 00:48:54.809 Uttam Kumaran: I’m going to create this.

540 00:48:56.270 00:48:57.530 Uttam Kumaran: I will

541 00:48:57.970 00:49:04.699 Uttam Kumaran: probably assign Brian. But for this situation. I’m just gonna push it through because it’s working locally.

542 00:49:05.373 00:49:09.309 Uttam Kumaran: As you can see, we have our 3 files here.

543 00:49:10.420 00:49:13.429 Uttam Kumaran: I’m going to go ahead and merge this.

544 00:49:16.150 00:49:19.739 Uttam Kumaran: And then what I’m gonna do is I’m also going to

545 00:49:21.450 00:49:26.710 Uttam Kumaran: Go back to the Cli, and I’m gonna hit real. And then let’s actually make sure that.

546 00:49:27.220 00:49:29.557 Uttam Kumaran: Let’s go back to our pool parts overview.

547 00:49:31.168 00:49:35.630 Uttam Kumaran: Looks like it’s not here yet. So what I’m gonna do is, I’m gonna say, real

548 00:49:35.720 00:49:37.150 Uttam Kumaran: deploy.

549 00:49:37.720 00:49:39.599 Uttam Kumaran: It’s going to

550 00:49:41.960 00:49:43.349 Uttam Kumaran: do something.

551 00:49:51.190 00:49:52.130 Uttam Kumaran: Maybe.

552 00:49:55.490 00:49:55.935 Uttam Kumaran: Hmm.

553 00:50:14.310 00:50:15.959 Uttam Kumaran: gonna try login again.

554 00:50:31.160 00:50:32.229 Uttam Kumaran: There it is.

555 00:50:33.330 00:50:34.170 Uttam Kumaran: Yay.

556 00:50:35.200 00:50:38.149 Uttam Kumaran: I don’t know honestly know if this had anything to do with this

557 00:50:38.790 00:50:43.110 Uttam Kumaran: kind of iffy on the deploy process. I’ve just kind of merged stuff, and it’s usually worked

558 00:50:46.290 00:50:47.210 Uttam Kumaran: But

559 00:50:47.760 00:50:48.809 Uttam Kumaran: let’s see.

560 00:50:48.880 00:50:51.399 Uttam Kumaran: dashboard quick start deploy dashboards.

561 00:50:51.620 00:50:55.810 Uttam Kumaran: local code files. Put it to Github. It then pushes it to real cloud.

562 00:50:56.872 00:50:59.350 Uttam Kumaran: Yeah, you just run real deploy.

563 00:51:00.130 00:51:01.120 Uttam Kumaran: and

564 00:51:01.660 00:51:02.460 Uttam Kumaran: I think.

565 00:51:02.620 00:51:04.720 Uttam Kumaran: just assume it gets taken care of.

566 00:51:05.455 00:51:13.079 Uttam Kumaran: If we go to direct mail dashboard. Here this is again. Ui real data. Pull parts ago. Pull parts to go click on direct mail.

567 00:51:13.320 00:51:14.769 Uttam Kumaran: There we go.

568 00:51:15.400 00:51:17.220 Uttam Kumaran: The cake is baked.

569 00:51:19.020 00:51:20.950 Jakob Kagel: So I guess like.

570 00:51:21.310 00:51:21.670 Uttam Kumaran: Cool.

571 00:51:21.670 00:51:29.760 Jakob Kagel: Let’s maybe align real quick, like on the next steps. Then I mean, I guess the 1st thing is right. We just want to bring the order Id into the direct mail team.