Meeting Title: Zoom Meeting Date: 2024-12-18 Meeting participants: Luke Daque, Nicolas Sucari, Uttam Kumaran, Payas Parab
WEBVTT
1 00:01:33.290 ⇒ 00:01:34.370 Luke Daque: Hi. Nicholas.
2 00:01:37.650 ⇒ 00:01:39.320 Nicolas Sucari: Here I am. How are you.
3 00:01:39.880 ⇒ 00:01:41.250 Luke Daque: I’m doing well. How are you?
4 00:01:44.240 ⇒ 00:01:45.170 Nicolas Sucari: I’m good.
5 00:01:46.630 ⇒ 00:01:47.430 Luke Daque: Nice.
6 00:01:49.030 ⇒ 00:01:50.299 Payas Parab: Hey? What’s up, gents?
7 00:01:51.290 ⇒ 00:01:51.869 Nicolas Sucari: If they ask.
8 00:01:51.870 ⇒ 00:01:53.640 Luke Daque: See you guys, I pass.
9 00:01:53.870 ⇒ 00:01:55.249 Payas Parab: How are you guys doing.
10 00:01:57.100 ⇒ 00:01:58.120 Luke Daque: Doing well, how are you.
11 00:01:59.870 ⇒ 00:02:02.874 Payas Parab: I’m doing good. Just been cranking on some
12 00:02:03.610 ⇒ 00:02:08.779 Payas Parab: trying to figure out this weather thing took me a while. I was like. Had to spend more time this morning.
13 00:02:10.020 ⇒ 00:02:32.649 Payas Parab: every Api. I found the cost just blow up so quickly. So I’ve been trying to like. Figure out what the best kind of approaches. I can kind of walk you guys through like where I’m at. So far so we can give an update. But yeah, can before before I jump in, happy to just kind of kick off like what you guys know about this or like, have any of you guys tried this before is this like, I’m just kind of starting.
14 00:02:34.610 ⇒ 00:02:45.699 Nicolas Sucari: So I I know we started once one time ago. I don’t know when we started trying to get some data in Snowflake. There was an Api.
15 00:02:45.800 ⇒ 00:02:57.830 Nicolas Sucari: I think it was that tried to bring the data into Snowflake once, but we haven’t do like, yeah, we didn’t do any analysis on that, and try to compare what? What with our data? So
16 00:02:58.180 ⇒ 00:02:58.700 Nicolas Sucari: yep.
17 00:02:58.700 ⇒ 00:03:01.239 Payas Parab: Is there data in Snowflake for weather.
18 00:03:02.620 ⇒ 00:03:07.030 Nicolas Sucari: I’m not sure. Let me know there was something, but I don’t know if we
19 00:03:07.590 ⇒ 00:03:10.189 Nicolas Sucari: we finally got all the the data.
20 00:03:10.800 ⇒ 00:03:38.390 Payas Parab: Got it. Okay? So I can kind of walk you guys through like, I kind of put together that prospecting doc where we’re at. Unfortunately, I will give the spoiler alert, unfortunately, is like, we’re not at the point of like, okay, this weather drives. This type of thing is just the reality. There’s a couple of reasons for that which I’ll kind of walk you through. I’m hoping to potentially drive through to there. I I need to like potentially purchase some Api access. So I wanted to kind of like run through the approach and kind of walk you guys through it and then see if like.
21 00:03:38.460 ⇒ 00:03:45.940 Payas Parab: this is the right way, and then we can kind of figure out you know what to do from it. Does that? Does that all sound good.
22 00:03:48.440 ⇒ 00:03:48.880 Luke Daque: Sounds good.
23 00:03:50.030 ⇒ 00:03:53.301 Payas Parab: Awesome. Sorry. I’m also eating chips. I’m hungry.
24 00:03:53.960 ⇒ 00:03:57.960 Payas Parab: awesome. So I’ll kind of start at the beginning right now. Still, we’re still in a
25 00:03:58.450 ⇒ 00:04:02.399 Payas Parab: kind of Ipython mode right here, right? Like, I think I mentioned part of
26 00:04:02.690 ⇒ 00:04:08.200 Payas Parab: the beginning of the approach. So
27 00:04:09.000 ⇒ 00:04:12.909 Payas Parab: basically like the voiceover, right? Is like we started with.
28 00:04:13.050 ⇒ 00:04:25.329 Payas Parab: we know we have this internal data, right? We have their sales data. And we have, you know, we can find 3rd party data sets right to basically give us the weather based on where that data is
29 00:04:25.929 ⇒ 00:04:29.280 Payas Parab: and so the starting point sort of was
30 00:04:29.990 ⇒ 00:04:32.209 Payas Parab: And again, like I, Nico. By the way, I like
31 00:04:32.510 ⇒ 00:04:46.120 Payas Parab: distill this into like bullet points for the client, so we won’t be like sharing a voice over Ipython notebook, but kind of just like where we’re at. I’ll I’ll make sure we get it in client, friendly format. But essentially like, I basically broke broke down.
32 00:04:46.678 ⇒ 00:04:56.250 Payas Parab: Okay, all of the orders we have, and by date, what the revenue was. And then like, what type of standardized thing we could use, I decided to opt for zip code, because it seems like
33 00:04:56.360 ⇒ 00:05:07.269 Payas Parab: the easiest level. But it’s granular enough. Most weather data relies on like what weather station the data is coming from. But even there, there’s like variance. And like.
34 00:05:08.020 ⇒ 00:05:13.280 Payas Parab: like, like, I basically was like, okay, like, what if we group together things we basically like.
35 00:05:13.460 ⇒ 00:05:14.340 Payas Parab: So
36 00:05:14.460 ⇒ 00:05:28.679 Payas Parab: the starting point was in an ideal world. What I wanted was for every Zip code where there was a sale I wanted 60 days prior, 60 days after, and I wanted to create like rolling metrics. Essentially right? So that was kind of where I started, where I was like, okay in an ideal world.
37 00:05:28.790 ⇒ 00:05:48.260 Payas Parab: I want to see all their sales. I want to see, you know where they’re at. And this is like we essentially landed on like we’d need date zip combinations. There’s 9.7 million essentially dates and zip codes that we need data for that’s obviously like, just really not tenable. Because Apis, we’re gonna call based on
38 00:05:48.925 ⇒ 00:06:08.619 Payas Parab: like per call, right? So like, even if it costs like just half a cent per Api call like, you can see, like 9 million Api calls like, really adds up really, really quickly. So what I tried to do is basically figure out, how do we distill this data set into smaller sections, right? So we wanna create like something that’s a representative sample of their customer base.
39 00:06:08.740 ⇒ 00:06:13.519 Payas Parab: And my approach was 1st like, let’s try and break it down by zip code. See if we can isolate
40 00:06:13.860 ⇒ 00:06:18.339 Payas Parab: a certain number of zip codes that represent a meaningful amount of their revenue.
41 00:06:18.520 ⇒ 00:06:25.699 Payas Parab: So you know that way, we don’t collect data. We don’t have to collect Api data for zip codes that aren’t as relevant. I found
42 00:06:26.310 ⇒ 00:06:33.316 Payas Parab: essentially the top 5,000 is like a sweet spot of like decent coverage on their data set, and like not
43 00:06:33.850 ⇒ 00:06:40.350 Payas Parab: not you know, not as extensive as asking for all their every zip code they’ve ever sold to.
44 00:06:41.212 ⇒ 00:06:59.310 Payas Parab: The next thing I did was essentially okay. Let’s try and isolate by Zips. Let’s isolate to those few. We still have a lot of a lot of dates. We need data for just to give you an example, the Api calls for the one I found to be the cheapest. We’re looking at like $7,500 for that data.
45 00:06:59.937 ⇒ 00:07:04.699 Payas Parab: So I basically kept going down being like, okay, we need to isolate even further.
46 00:07:05.325 ⇒ 00:07:14.489 Payas Parab: One attempt was like, what about zip codes that are close enough to each other? Or we take representative samples from every State right. So, instead of doing like
47 00:07:14.620 ⇒ 00:07:20.199 Payas Parab: every state will do like. Okay. We know that like maybe trends in Florida are similar to trends in.
48 00:07:20.530 ⇒ 00:07:27.159 Payas Parab: like, like, you know, if they’re they have 10 customer, 10 states. And we can isolate. I thought maybe we could isolate. So
49 00:07:27.270 ⇒ 00:07:32.490 Payas Parab: this is just basically kind of the process. So basically like the core issue here. To do this analysis, right is like
50 00:07:33.370 ⇒ 00:07:43.709 Payas Parab: we need to determine a good sampling method for the amount of dates that we’re going to pull data for, and then try and isolate the universe of zip codes right to make it as simplified as possible.
51 00:07:43.910 ⇒ 00:07:50.690 Payas Parab: So ultimately what I landed on after kind of cutting it. I was like, Okay, what if we just isolate it to certain states?
52 00:07:51.336 ⇒ 00:07:57.909 Payas Parab: Like, ultimately, what I landed on is like monthly data for the top 5,000 zip codes
53 00:07:58.582 ⇒ 00:08:00.749 Payas Parab: and pulling that from
54 00:08:01.030 ⇒ 00:08:11.850 Payas Parab: a like a monthly aggregated data, and then only pulling the months that like there’s a customer. And then there’s like months, 2 months behind 2 months in front of it. So we basically try and create this like
55 00:08:12.070 ⇒ 00:08:16.010 Payas Parab: range essentially of like, okay, the top 5,000 zip codes.
56 00:08:16.310 ⇒ 00:08:20.960 Payas Parab: I was able to bring the cost down essentially to like, okay, now, we’re only at 179,000.
57 00:08:22.990 ⇒ 00:08:32.250 Uttam Kumaran: Price. What’s this? What’s the cost? Like? What’s the cost? Like, based on like, what do you mean by Api calls? Do you have, like an Api, that
58 00:08:32.570 ⇒ 00:08:33.769 Uttam Kumaran: you are looking at.
59 00:08:34.070 ⇒ 00:08:34.760 Payas Parab: Yes, yeah.
60 00:08:38.309 ⇒ 00:08:40.589 Payas Parab: I tried a few. I tried
61 00:08:41.439 ⇒ 00:08:48.179 Payas Parab: I tried open weather. There’s like a list of a few that I tried. This. This Api call is like this is from open open weather
62 00:08:48.599 ⇒ 00:08:53.619 Payas Parab: data set, which on a per call basis.
63 00:08:53.620 ⇒ 00:08:55.110 Uttam Kumaran: Okay, I think one thing.
64 00:08:55.980 ⇒ 00:08:56.400 Payas Parab: Yeah.
65 00:08:56.400 ⇒ 00:09:04.969 Uttam Kumaran: Yeah, I think one thing at the end of this is if you just tell me, like, Hey, I need, I need data in this shape. Basically, I can probably go find
66 00:09:05.260 ⇒ 00:09:09.820 Uttam Kumaran: it for you as well. But okay, keep keep going. Keep going.
67 00:09:09.820 ⇒ 00:09:16.789 Payas Parab: How? How? So can I? Can? I ask, like, cause? I’m I’m just basically trying to get that 3rd party data. Is there like some other source that I don’t know about, or some better way.
68 00:09:17.111 ⇒ 00:09:23.860 Uttam Kumaran: Yeah. So like, there’s yeah. I mean, there’s there’s sources in Snowflake, like within snowflake marketplace. You can actually go get
69 00:09:24.090 ⇒ 00:09:25.930 Uttam Kumaran: pre-modeled weather data.
70 00:09:26.130 ⇒ 00:09:26.480 Uttam Kumaran: Okay.
71 00:09:26.480 ⇒ 00:09:47.219 Uttam Kumaran: I’ve done some research on it. I just need to know the shape that you need, because the weather data will include a whole bunch of other stuff. So up to this point, like, I was kind of like, okay, what do we need? Do we need like at a zip level like like weather incidents. Do we need like temperature like like, what? What is it? Exactly so if you tell me the shape
72 00:09:47.631 ⇒ 00:09:52.328 Uttam Kumaran: then I will go also. Do one more search to make to see if I can get that
73 00:09:52.590 ⇒ 00:09:53.270 Payas Parab: Okay.
74 00:09:53.480 ⇒ 00:09:58.800 Uttam Kumaran: You know, from there. But frankly, this cost, this cost is like pretty marginal, so.
75 00:09:58.800 ⇒ 00:09:59.180 Payas Parab: Yeah.
76 00:09:59.180 ⇒ 00:10:02.659 Uttam Kumaran: Even just seeing this, I feel like we. This is totally fine.
77 00:10:03.010 ⇒ 00:10:08.910 Payas Parab: Okay, got it? Yeah, I think so. Basically, the ideal format I landed on. And this is just like playing around with, like.
78 00:10:09.380 ⇒ 00:10:27.300 Payas Parab: you know what there’s like. There’s sort of the ideal state which is like daily data 30, 60 days before and after per zip code. They have a shit ton of zip codes that they ship to. And a lot of them aren’t super close. So you actually do need like zip code level data. So I think that could add up, I don’t know. Maybe Snowflake has, like a lot of cheaper data sets. But like.
79 00:10:27.850 ⇒ 00:10:34.539 Payas Parab: I think these Api calls that I was looking at, we’re adding up, and I assume the pricing would be similar unless Snowflake is like crazy, cheap.
80 00:10:37.160 ⇒ 00:10:43.440 Uttam Kumaran: It may be just like it may be free and stuff like basically people list data sets in the marketplace. So you can just download
81 00:10:43.990 ⇒ 00:10:56.589 Uttam Kumaran: so if I find one that’s there, I’ll just let you know, and that saves us some time in modeling. But just tell me the shape after this, just tell me the shape of the actual data that you need. Or if you give me.
82 00:10:56.890 ⇒ 00:11:02.279 Uttam Kumaran: yeah, basically like, what are the mandatory fields that you need from that, and I’ll just go do another pass through.
83 00:11:02.280 ⇒ 00:11:06.310 Payas Parab: So. So this is the. And I can show you essentially like what I landed on.
84 00:11:06.520 ⇒ 00:11:08.279 Payas Parab: As like the ideal. So
85 00:11:08.969 ⇒ 00:11:19.800 Payas Parab: in this you don’t need all of this. I’m just saving everything down from this Api. But essentially, if you can get me the temperature min, Max, date precipitation by month
86 00:11:20.330 ⇒ 00:11:23.899 Payas Parab: per zip code. That would get the job done.
87 00:11:30.150 ⇒ 00:11:31.110 Payas Parab: Does that?
88 00:11:32.490 ⇒ 00:11:33.040 Payas Parab: Okay?
89 00:11:33.040 ⇒ 00:11:34.010 Uttam Kumaran: That makes sense.
90 00:11:34.500 ⇒ 00:11:50.040 Payas Parab: Yeah. So I mean, I I ultimately did find like where we can get this data from right now, it’s like pulling it. And I hit the Api limit for the free trial just to, I just wanna take run a quick test on like, okay, is there any? Does this correlate to enough customer data that this would be like worthwhile at all.
91 00:11:50.450 ⇒ 00:11:59.320 Payas Parab: And so right now, where I’m at is basically like, yeah, let me know if you find a better data source for this. This one was like the one I landed on was like.
92 00:11:59.720 ⇒ 00:12:03.030 Payas Parab: if we can geocode for free, and then we can
93 00:12:03.140 ⇒ 00:12:11.519 Payas Parab: pull in temperature. Min average Max precipitation rate and then there’s like a humidity score. If we can pull that in
94 00:12:11.770 ⇒ 00:12:19.120 Payas Parab: per I forget what. It’s not humidity, it’s something else. But or sorry. There’s like days of sunshine or percentage sunshine
95 00:12:19.250 ⇒ 00:12:41.029 Payas Parab: that this data set has. If I can find all of those that would be superior to like what I’ve built here. But like this also what I built here also kind of gets the job done. So I feel like decently about it where it’s like I’m pulling in those metrics. I’m able to get a full time series, for as long as a customer has been active for the top 5,000 zip codes
96 00:12:41.190 ⇒ 00:12:46.080 Payas Parab: in a reasonable what I believe to be a reasonable cost. I I basically just have to upgrade to this like.
97 00:12:46.100 ⇒ 00:12:46.610 Uttam Kumaran: 30.
98 00:12:46.610 ⇒ 00:12:52.559 Payas Parab: $40 a month Api tier, and then pull all of it down, and it would just take like an hour or 2 to pull it all down.
99 00:12:59.870 ⇒ 00:13:04.989 Uttam Kumaran: Okay, just just shoot me one of these like, just literally shoot me one of these Csv’s in slack.
100 00:13:05.750 ⇒ 00:13:11.050 Uttam Kumaran: And I will see if I can get I will. I’ll see if I can find something. But
101 00:13:11.220 ⇒ 00:13:15.510 Uttam Kumaran: yeah, you can continue on that. That cost like 300 bucks, or whatever that’s whatever that
102 00:13:15.960 ⇒ 00:13:17.200 Uttam Kumaran: they’ll be. Okay with that.
103 00:13:17.480 ⇒ 00:13:21.510 Payas Parab: Yep, okay. Got it. Yeah. Nico, you wanna share you.
104 00:13:21.900 ⇒ 00:13:28.869 Nicolas Sucari: This is no fake. I just found this one in the marketplace for the source. Or see, you have here global.
105 00:13:28.870 ⇒ 00:13:32.530 Uttam Kumaran: Yeah, there’s a couple yeah to go confirm. Like.
106 00:13:32.990 ⇒ 00:13:35.399 Uttam Kumaran: yeah, I’m gonna go confirm which ones
107 00:13:35.500 ⇒ 00:13:41.039 Uttam Kumaran: work. But we don’t have to do that now, like I just wanna, I just wanna hear about more about the analysis piece and like.
108 00:13:42.120 ⇒ 00:13:50.699 Nicolas Sucari: Cool. Okay, and pay us. This is the one that we had here in in Snowflake, like there is a lot of information. I don’t know if this is something that we can use.
109 00:13:50.700 ⇒ 00:13:51.360 Nicolas Sucari: Buddy.
110 00:13:51.360 ⇒ 00:13:55.520 Nicolas Sucari: What we used before or we were trying to is okay.
111 00:13:55.980 ⇒ 00:13:57.429 Payas Parab: Got it. Got it?
112 00:14:00.050 ⇒ 00:14:09.069 Payas Parab: Got it? Yeah, I can review those. I actually didn’t know we had those in Snowflake. So I was just trying to like. Figure out what that 3rd party data source was. I I think, like.
113 00:14:09.480 ⇒ 00:14:28.420 Payas Parab: I think the core of it is just monthly before and after precipitation temperature. And then, basically, what I was hoping to do is pump that through a random forest, once we have that full feature set. So it’s like basically the final feature set would be. And you know, right now, it’s just pulling that data and making the joins from what I have which I can do, which is like, basically
114 00:14:28.650 ⇒ 00:14:29.710 Payas Parab: you have.
115 00:14:30.220 ⇒ 00:14:36.000 Payas Parab: yeah, like per month. There’s a sale. There’s like a revenue number. And then there’s a 2 month prior
116 00:14:36.110 ⇒ 00:14:45.749 Payas Parab: one month prior, one month after one month or 2 month after, and you would just pull every single one precipitation min average max
117 00:14:46.490 ⇒ 00:14:47.589 Payas Parab: and then,
118 00:14:48.140 ⇒ 00:14:49.970 Uttam Kumaran: It every day of the month.
119 00:14:50.910 ⇒ 00:14:52.699 Payas Parab: No, I’m doing aggregated.
120 00:14:55.690 ⇒ 00:14:58.490 Uttam Kumaran: Dude. How’s that gonna work? Bro, the weather changes in a month.
121 00:14:59.710 ⇒ 00:15:01.070 Payas Parab: Wait! What do you mean?
122 00:15:02.284 ⇒ 00:15:08.309 Uttam Kumaran: Like the weather on October first, st has nothing to do with the weather. On October 31.st
123 00:15:09.490 ⇒ 00:15:14.270 Payas Parab: Yeah. But you can take like an aggregated for like, what’s part of the season? Right? That it does.
124 00:15:14.270 ⇒ 00:15:15.390 Uttam Kumaran: Boyfriend, 19 6.
125 00:15:16.980 ⇒ 00:15:19.380 Uttam Kumaran: I like them a lot. I think those things like that.
126 00:15:19.990 ⇒ 00:15:24.208 Uttam Kumaran: I mean, I’m open to being convinced. But like I feel like
127 00:15:25.330 ⇒ 00:15:27.699 Uttam Kumaran: if we’re trying to look at. I mean, I think.
128 00:15:27.850 ⇒ 00:15:40.079 Uttam Kumaran: like I guess I guess what I’m saying is that like I would be. I could be convinced that, like you can look at the weather for like a week. But if the weather is sunny like in Texas here, if it’s sunny.
129 00:15:40.400 ⇒ 00:15:51.939 Uttam Kumaran: for example, if it’s sunny on December first, st but then it starts to rain on December 21st people place orders on December 21, st then.
130 00:15:52.350 ⇒ 00:15:53.520 Uttam Kumaran: like
131 00:15:53.740 ⇒ 00:16:07.560 Uttam Kumaran: I don’t know you you would. We would lose that context right? Like basically, the hypothesis is that when the the kind of the hypothesis is that when one, if there are weather, is there? If there is weather, activity, or forecasted weather, activity.
132 00:16:07.560 ⇒ 00:16:07.970 Payas Parab: Sure.
133 00:16:07.970 ⇒ 00:16:16.280 Uttam Kumaran: Then sales for either one opening pools. So people need pumps and stuff like that could change, or
134 00:16:16.560 ⇒ 00:16:23.650 Uttam Kumaran: people buy covers and things like that to close their pools right? So it’s basically like trying to trace, try to like, follow the weather patterns and say, like.
135 00:16:23.650 ⇒ 00:16:24.190 Payas Parab: Sure, sure.
136 00:16:24.190 ⇒ 00:16:33.499 Uttam Kumaran: Okay, broadly, you know. So I mean, I think on on a monthly basis. Probably you’ll get a sense of like, it’s getting hotter and colder. But I think, like we already know.
137 00:16:33.720 ⇒ 00:16:37.209 Uttam Kumaran: like you don’t need you, don’t. We don’t need to do this to know that already.
138 00:16:37.629 ⇒ 00:16:46.599 Uttam Kumaran: That’s already true. That’s a known fact that like when it gets hot, people open the pools and we see them. The data. This is, I think this is more about
139 00:16:46.700 ⇒ 00:16:49.920 Uttam Kumaran: if we know a storm is gonna hit
140 00:16:50.260 ⇒ 00:17:08.909 Uttam Kumaran: in like 2 weeks, then could ideally, what actions can they take? Maybe they can run a promo. Maybe they can do some sort of marketing right? Same thing. If we know that in an area like here’s a good example. They they probably they do a huge amount of their sales during this opening and closing of the season.
141 00:17:08.970 ⇒ 00:17:34.129 Uttam Kumaran: Right? So I think, like April, may think like September, October, they do a shitload of sales there during during the winter is really really closed down. And then during the summer, it’s still pretty good. Because people are opening up their pools, basically like across the country as the weather moves. The nice thing is, let’s say, in one year the weather turns early in an area we could probably get ahead of that by starting to launch marketing and things like that. That’s the real.
142 00:17:34.660 ⇒ 00:17:35.450 Payas Parab: Yeah, for sure.
143 00:17:35.450 ⇒ 00:17:53.929 Uttam Kumaran: Key thing here. And another example. Last example that is like, if we know Atlanta is gonna get hit by a storm, then we should start running campaigns for them to buy pool covers and stuff like that. Instead, they, some people may do it, but then us, taking advantage of that could lead to some more sales. That’s the hypothesis.
144 00:17:53.930 ⇒ 00:18:18.949 Payas Parab: I. I hear you on that. And the the question, then it more is more of like a data availability thing of like, if we have that daily, like I, I started with like the thesis being like daily data, and then being able to model that out. This like similar logic, as I mentioned, right is like you could take the team in T. Max T. Average. And it’s just that like that data availability. For like, if we need to get their top 5,000 zip codes which they have sales and like
145 00:18:19.020 ⇒ 00:18:24.499 Payas Parab: for recurring, or like multiple for certain zip codes, you need like multiple months of range.
146 00:18:24.600 ⇒ 00:18:29.340 Payas Parab: and the data is coming in a per day. And we’re sort of being charged on a per day basis.
147 00:18:29.530 ⇒ 00:18:34.020 Uttam Kumaran: But I guess that’s what I’m saying. Don’t I want to talk about that limitation
148 00:18:34.170 ⇒ 00:18:38.129 Uttam Kumaran: versus like I didn’t know. We went ahead and said, like, Oh, we
149 00:18:38.270 ⇒ 00:18:40.570 Uttam Kumaran: it’s expensive, so let’s not do it.
150 00:18:40.650 ⇒ 00:18:43.400 Payas Parab: Like what what I’m saying is like, I wanna.
151 00:18:43.460 ⇒ 00:18:51.880 Uttam Kumaran: I want to push back and say, I think there’s gonna be. There’s 2 ways around it. One I just need to like this is more of like a data engine problem like, can we source
152 00:18:51.930 ⇒ 00:18:55.020 Uttam Kumaran: this daily data for you? Basically.
153 00:18:55.030 ⇒ 00:19:18.930 Uttam Kumaran: So there’s a couple of ways. One is you, you already have some research on Apis. We have some costing. Second is, I want to do another search on my end, because I’ve never had, like the clear requirements from anybody on what they need here. This is the 1st time I’ve seen like. So I’m gonna go call some friends and be like Yo, where do we get weather data these days of the cheap? So let’s let’s keep going with that. Because let’s set the requirement as we need daily data.
154 00:19:18.930 ⇒ 00:19:27.280 Uttam Kumaran: And the the last backup we can do is if with the cost, I can go to them and say, Hey, we need to buy this data. Are you guys, okay with it. But I don’t want to move past.
155 00:19:27.620 ⇒ 00:19:33.869 Uttam Kumaran: I don’t want to sacrifice the quality of the analysis based on data availability. I want to hit the data availability problem first.st
156 00:19:34.410 ⇒ 00:19:36.260 Payas Parab: I see? Okay? So I can summarize.
157 00:19:36.260 ⇒ 00:19:37.150 Uttam Kumaran: Does that make sense.
158 00:19:37.150 ⇒ 00:19:40.480 Payas Parab: Yes, I can summarize what we can do if we can both take a stab at it. I just like.
159 00:19:41.250 ⇒ 00:19:47.970 Payas Parab: Yeah, maybe I’m just like coming like my, my, basically like, in order to pitch the like
160 00:19:48.560 ⇒ 00:19:56.089 Payas Parab: like. So you’re saying, like, if we’re willing to go spend money on data and go get it right. I just wanna like, before you make that pitch. You wanna make sure that like
161 00:19:56.090 ⇒ 00:19:57.240 Payas Parab: there’s some
162 00:19:57.240 ⇒ 00:20:04.240 Payas Parab: grounds to it, right? So I’m like in the cheapest way possible. Can I determine if there’s some grounds, for it was kind of my approach, but if I like.
163 00:20:04.240 ⇒ 00:20:05.040 Uttam Kumaran: Hmm.
164 00:20:05.390 ⇒ 00:20:07.569 Payas Parab: Like if I framed it for you. Okay, first, st let’s just see.
165 00:20:07.570 ⇒ 00:20:12.150 Uttam Kumaran: The the weekly.
166 00:20:13.180 ⇒ 00:20:21.089 Uttam Kumaran: Yeah, I see what you mean. But dude, there’s also different different ways we can get. I can get you a 3 month daily data set for one Zip code or the highest Zip code or
167 00:20:21.330 ⇒ 00:20:33.350 Uttam Kumaran: Florida, like a piece of Florida. Right? So there’s a couple of different ways. We can do that if if like, if your thing is okay, I wanna make sure that this works. Then why don’t we just go take the top 3 zips in Florida.
168 00:20:33.480 ⇒ 00:20:37.759 Uttam Kumaran: Get the 3 months day by day by day, and then you run it on that.
169 00:20:38.110 ⇒ 00:20:52.559 Payas Parab: I can do that. I can do that. I just. I don’t think there’s enough sampling in like a given zip, though that’s the issue. So like, if you look the number one, the number one zip, you have, like point 2 7% point 1% of orders. Right?
170 00:20:52.560 ⇒ 00:20:53.089 Uttam Kumaran: Doesn’t it.
171 00:20:53.090 ⇒ 00:20:55.200 Payas Parab: So to try and like build some type of a relationship.
172 00:20:55.200 ⇒ 00:20:59.760 Uttam Kumaran: But why but why don’t you do? Why don’t you do top zips by like state
173 00:20:59.980 ⇒ 00:21:02.170 Uttam Kumaran: like, take Florida, and then take the top.
174 00:21:03.050 ⇒ 00:21:09.629 Uttam Kumaran: Take like wherever 60% of sales coming from, and then just start there. That way limits the zip codes right.
175 00:21:09.630 ⇒ 00:21:10.030 Payas Parab: Yeah.
176 00:21:10.030 ⇒ 00:21:14.470 Uttam Kumaran: Go one layer, one large cause, because weather weather is sort of centralized by
177 00:21:15.020 ⇒ 00:21:18.536 Uttam Kumaran: like on this sort of like on a Geo basis. Right? So.
178 00:21:19.020 ⇒ 00:21:27.225 Payas Parab: Let me quickly show you just like when I ran this in Snowflake. Right? I think that was like part of my my initial inclination was that as well. But it was that like, there’s
179 00:21:27.460 ⇒ 00:21:28.060 Uttam Kumaran: Okay.
180 00:21:28.060 ⇒ 00:21:32.969 Payas Parab: There’s not as much of a concentration as you would like hope like, even if I filter into
181 00:21:33.766 ⇒ 00:21:35.750 Payas Parab: you see, my snowflake.
182 00:21:39.260 ⇒ 00:21:39.610 Uttam Kumaran: Yes.
183 00:21:41.650 ⇒ 00:21:46.740 Payas Parab: Excellent. So I have here. It’s like, Okay, the zipping shipping zip code by this, and then I can
184 00:21:47.742 ⇒ 00:21:49.400 Payas Parab: filter that final.
185 00:21:49.400 ⇒ 00:21:49.950 Uttam Kumaran: That’s true.
186 00:21:50.320 ⇒ 00:21:51.919 Payas Parab: Kind of data set by
187 00:21:55.250 ⇒ 00:21:57.269 Payas Parab: like, let’s just try Florida right.
188 00:22:02.490 ⇒ 00:22:03.220 Uttam Kumaran: He did it.
189 00:22:05.680 ⇒ 00:22:06.300 Payas Parab: Oh!
190 00:22:06.300 ⇒ 00:22:07.310 Uttam Kumaran: It’s great.
191 00:22:08.500 ⇒ 00:22:11.350 Uttam Kumaran: Oh, I see what you mean. Meaning like, yeah, there’s a.
192 00:22:11.350 ⇒ 00:22:13.720 Payas Parab: But there’s a there’s a lot of zip codes, and
193 00:22:15.360 ⇒ 00:22:16.809 Payas Parab: and on a given day it’s like.
194 00:22:16.810 ⇒ 00:22:17.440 Uttam Kumaran: But then.
195 00:22:18.890 ⇒ 00:22:19.590 Payas Parab: Go ahead!
196 00:22:20.820 ⇒ 00:22:22.950 Uttam Kumaran: Okay, okay, I kind of. I see what you mean. I see what you mean.
197 00:22:22.950 ⇒ 00:22:33.749 Payas Parab: So if you look at the revenue percentage, the the order percentage and like a given zip in Florida, right like here, like, there’s actually just such a small amount of data. And this is over 2 and a half years.
198 00:22:33.750 ⇒ 00:22:34.180 Uttam Kumaran: Something.
199 00:22:34.180 ⇒ 00:22:37.260 Payas Parab: If you split this out over 2 and a half years, the idea that you’d get like.
200 00:22:37.260 ⇒ 00:22:37.830 Uttam Kumaran: That’s like.
201 00:22:37.830 ⇒ 00:22:48.319 Payas Parab: I guess you could again. We can try it right like, I think you’re right. Maybe for the purposes of just trying it. I just think like in general for building a model. You’re not gonna really get a meaningful output on like a daily basis. But.
202 00:22:48.320 ⇒ 00:22:51.950 Uttam Kumaran: Can we cluster by anything in between Zip and State.
203 00:22:51.950 ⇒ 00:22:56.722 Payas Parab: That’s I’m gonna try and do that right now, I’ve I’ve been trying to do that. I tried to basically try
204 00:22:58.070 ⇒ 00:22:58.580 Payas Parab: cluster. Maybe.
205 00:22:58.580 ⇒ 00:23:02.790 Uttam Kumaran: Because zip, because basically look, Zip has zip has lat long.
206 00:23:02.790 ⇒ 00:23:03.130 Payas Parab: Yep.
207 00:23:03.130 ⇒ 00:23:07.840 Uttam Kumaran: And then you can kind of cluster by like something there, right.
208 00:23:07.840 ⇒ 00:23:08.740 Payas Parab: Yeah, yeah.
209 00:23:11.640 ⇒ 00:23:15.069 Uttam Kumaran: I’ll let you think about it. But I don’t want. I don’t want cost to be. The reason we
210 00:23:15.210 ⇒ 00:23:18.460 Uttam Kumaran: don’t do daily like this isn’t gonna really be meaningful
211 00:23:19.110 ⇒ 00:23:41.340 Uttam Kumaran: other than other than having that daily data. So I want to start there. If you give me the shape of the data, I’ll go find it. Otherwise we’ll estimate the cost, and I’ll propose it to them. Look, I don’t want to do extra work in order to Co. I’d rather just be like, hey, for us to do this. It’s gonna cost this. If you want, we can start smaller and just make sure it works, or you can. We can buy it. Let me handle that
212 00:23:42.190 ⇒ 00:23:53.829 Uttam Kumaran: But I wanna do I wanna go? I wanna do the complicated stuff like I wanna go after the more complicated answer here. What was if it if it works or not, cause cause dude? What they’re gonna say is, if I go to them and tell them
213 00:23:54.340 ⇒ 00:23:58.299 Uttam Kumaran: sales correlates with weather gonna be like, yeah fucking duh.
214 00:23:58.300 ⇒ 00:23:59.520 Payas Parab: I get that, I get that. Yeah.
215 00:24:00.030 ⇒ 00:24:02.120 Uttam Kumaran: But yeah.
216 00:24:03.550 ⇒ 00:24:05.930 Payas Parab: If we can find a relationship.
217 00:24:06.960 ⇒ 00:24:17.769 Payas Parab: because because their orders span so much right? Like, maybe we can try and cluster these based on the distance between zip codes. Maybe we can try and do that and find like a grouping that makes a ton of sense, and then just
218 00:24:18.320 ⇒ 00:24:21.689 Payas Parab: test that one but it just like, because there’s such a small.
219 00:24:21.690 ⇒ 00:24:31.480 Uttam Kumaran: I would be surprised, dude, if, like someone in data hasn’t done this. Where there’s something there’s some sort of Geo sort of clustering between Zip
220 00:24:31.690 ⇒ 00:24:39.240 Uttam Kumaran: and State. There’s gotta be some sort of sector or something like that where it’s like
221 00:24:39.600 ⇒ 00:24:47.328 Uttam Kumaran: Southwest Florida, these sectors, and maybe there’s some sort of segmentation. We can do that may get us a little bit further.
222 00:24:48.380 ⇒ 00:24:53.004 Uttam Kumaran: because we do have with the their business is really great state concentration.
223 00:24:53.929 ⇒ 00:25:02.879 Uttam Kumaran: but you’re right like there is a lot of zips so that has to be something in between. We could try to aggregate to. And hopefully, that’s all. If not, then
224 00:25:03.780 ⇒ 00:25:04.380 Uttam Kumaran: basically.
225 00:25:04.750 ⇒ 00:25:09.469 Payas Parab: Like insane like, it’s surprising, like, for some reason, I mean, okay, they do have like one or 2.
226 00:25:09.770 ⇒ 00:25:10.340 Uttam Kumaran: Okay.
227 00:25:10.340 ⇒ 00:25:15.869 Payas Parab: Permanent ones. But it’s like still is like among the top 10 States, like a pretty good spread.
228 00:25:16.659 ⇒ 00:25:18.239 Uttam Kumaran: Okay. Okay.
229 00:25:18.500 ⇒ 00:25:23.929 Uttam Kumaran: there, yeah, there is a. It’s just there’s a good spread among the top 10. But beyond the top 10 it dies.
230 00:25:25.210 ⇒ 00:25:28.100 Payas Parab: Sure. Sure, I think it’s 63% as well.
231 00:25:28.140 ⇒ 00:25:34.879 Uttam Kumaran: Most of their businesses like Arizona. It’s like Arizona, New York, Florida, California.
232 00:25:36.970 ⇒ 00:25:38.579 Uttam Kumaran: A couple of others, Texas.
233 00:25:38.580 ⇒ 00:25:40.109 Payas Parab: Yeah, I have that.
234 00:25:40.369 ⇒ 00:25:44.519 Uttam Kumaran: We’ve done this sort of second. We’ve done this sort of 80 20 for that before.
235 00:25:47.320 ⇒ 00:25:50.559 Payas Parab: Makes sense. Okay. So maybe the the best action item is just like.
236 00:25:50.560 ⇒ 00:25:51.120 Uttam Kumaran: Thank you.
237 00:25:52.170 ⇒ 00:25:53.010 Uttam Kumaran: Leave my school.
238 00:25:53.010 ⇒ 00:26:05.245 Payas Parab: The best action. So Florida and Texas, so we can isolate a set of zip codes that are close to each other, and then do daily data and then do a quick like quick analysis on those team in team, Max team minus
239 00:26:05.730 ⇒ 00:26:06.420 Payas Parab: and just.
240 00:26:06.420 ⇒ 00:26:08.809 Nicolas Sucari: Precipitation. Maybe maybe precipitation.
241 00:26:08.810 ⇒ 00:26:15.749 Payas Parab: Precipitation, and like days of sunshine, or something on a 60 day minus 60 day plus 60 day rolling basis.
242 00:26:16.400 ⇒ 00:26:18.610 Payas Parab: Then we could talk that through, and see what that.
243 00:26:18.610 ⇒ 00:26:19.260 Uttam Kumaran: Yeah.
244 00:26:19.420 ⇒ 00:26:32.240 Payas Parab: Small, isolated set, and then we can propose the larger project. I I would be curious if, like, I’ll give you the exact format I need on the data I did look at like quite a few sources. So I’m like I would pricing.
245 00:26:32.240 ⇒ 00:26:35.740 Uttam Kumaran: No, no, I I trust you on the costing. I just want to check the one.
246 00:26:35.910 ⇒ 00:26:40.589 Uttam Kumaran: I don’t. I don’t really care about Price. I care about you having the
247 00:26:42.050 ⇒ 00:26:44.300 Uttam Kumaran: the data that you need
248 00:26:44.880 ⇒ 00:26:55.102 Uttam Kumaran: But, like again, if it’s like, gonna be like 1,000, and I don’t care about it. I know we haven’t set really like budgets or conversation on that. But
249 00:26:55.720 ⇒ 00:26:57.599 Uttam Kumaran: I also I’m
250 00:26:58.320 ⇒ 00:27:09.599 Uttam Kumaran: there’s a couple of ways we can attack it. Look, if we know that there’s gonna be signal. They’ll fund us so that’s why I want to start with at least one of the top 4 States, one of the top 5 States that you sent ideally.
251 00:27:09.840 ⇒ 00:27:35.160 Uttam Kumaran: ideally something with actual weather like Texas doesn’t have crazy weather. So it’s really like Florida. It’s probably I would. I don’t know. Honestly, I would just go, Florida, because that every there they know their zip codes like Super. Super. Well, you do, Florida, New York. They’ll be like, oh, yeah, I know that zip code that city? So I would say, let’s start with Florida. That way. Gives us the best odds of seeing something. And then, yeah, if you can send me the shape. Basically, I need.
252 00:27:35.250 ⇒ 00:27:43.030 Uttam Kumaran: I need zip these columns daily for this range. I’ll just do a quick check on Snowflake. And then otherwise
253 00:27:43.363 ⇒ 00:27:45.919 Uttam Kumaran: you can tell me the bill, and I’ll just confirm with them.
254 00:27:46.090 ⇒ 00:27:46.660 Payas Parab: Okay.
255 00:27:46.660 ⇒ 00:27:47.960 Payas Parab: Sounds good. Alright.
256 00:27:47.960 ⇒ 00:27:55.619 Uttam Kumaran: And then, if you need help, actually doing the like, running the Api and moving it we’ll get the AI team to help basically
257 00:27:56.725 ⇒ 00:28:00.154 Uttam Kumaran: run those requests for you and store that in a in a
258 00:28:00.690 ⇒ 00:28:03.359 Uttam Kumaran: yeah, in Snowflake, or in something.
259 00:28:03.890 ⇒ 00:28:12.949 Payas Parab: Okay. Sounds good. I can take that right now. That’ll just be like what I’ll do for the next hour or 2 is just. I’ll send you the format and see if we can isolate that one thing and just get the daily data
260 00:28:12.950 ⇒ 00:28:14.820 Payas Parab: back to me that looks like and go from there.
261 00:28:14.820 ⇒ 00:28:15.530 Uttam Kumaran: Dope.
262 00:28:16.010 ⇒ 00:28:17.170 Payas Parab: Alright, sounds good.
263 00:28:17.170 ⇒ 00:28:18.269 Uttam Kumaran: It’s out. Bye.
264 00:28:18.270 ⇒ 00:28:18.659 Payas Parab: Alright!
265 00:28:19.930 ⇒ 00:28:34.449 Nicolas Sucari: Great. Thank you. Pay us just for to send something to the client tomorrow. Maybe we can. Yeah, create some kind of a description of what we are doing and some bullet points so that we can share. That would be great. Okay, Tom, what do you think about that? So that we can share with Ben and Dan.
266 00:28:34.450 ⇒ 00:28:35.462 Uttam Kumaran: Yeah, that’s perfect.
267 00:28:35.800 ⇒ 00:28:36.550 Nicolas Sucari: Update.
268 00:28:36.550 ⇒ 00:28:44.149 Uttam Kumaran: Exactly. I just wanna share. I just want to kind of send them what we talked about today. And then basically what we’re starting to try first.st
269 00:28:44.490 ⇒ 00:28:48.619 Uttam Kumaran: I mean again, like they’re gonna skip anything that’s like
270 00:28:48.790 ⇒ 00:29:01.179 Uttam Kumaran: they they want to see the answer, basically so as fast as we can move towards that and then. But I’m glad we had this conversation. Yeah, for me, it’s like on the analyst side. Look, I I want you guys to have
271 00:29:01.300 ⇒ 00:29:11.868 Uttam Kumaran: as much support and tools to actually get to like the real meat of the problem. If the data access is an issue, then we’ll bust through that like. I don’t want to sacrifice the quality of that, so we’ll find something
272 00:29:12.120 ⇒ 00:29:16.460 Payas Parab: That makes sense. Yeah, I that was kind of like my assumption is I. I sort of assumed that I was starting the project from.
273 00:29:16.460 ⇒ 00:29:17.200 Uttam Kumaran: No, it’s fair.
274 00:29:17.200 ⇒ 00:29:25.240 Payas Parab: Like. And maybe it was like again, that that’s just like a coordination oversight of like, okay, maybe I need to like, tell you, this is what I’m.
275 00:29:25.240 ⇒ 00:29:26.319 Uttam Kumaran: No, no, it’s totally fair.
276 00:29:26.320 ⇒ 00:29:28.610 Payas Parab: Yeah, this is fucking dumb like, we can’t do this.
277 00:29:28.840 ⇒ 00:29:32.449 Uttam Kumaran: It’s more like I wanna have a conversation about about like
278 00:29:33.091 ⇒ 00:29:36.600 Uttam Kumaran: adjusting like our expectations for the outcome
279 00:29:37.026 ⇒ 00:29:40.650 Uttam Kumaran: before just being like, oh, it’s gonna cost money. So we’re not gonna do it like
280 00:29:42.650 ⇒ 00:29:53.699 Uttam Kumaran: that’s I think that’s a conversation we could. We’re having here today, which is like, oh, I actually don’t want to adjust it. Let’s find a way to make the money work and get you the right data. So cool. Okay, let’s do that.
281 00:29:54.230 ⇒ 00:29:59.290 Payas Parab: Okay, sounds good. Let’s do that subsample thing, and then I’ll send you the data requirements, and then we’ll move from there.
282 00:30:00.400 ⇒ 00:30:01.370 Uttam Kumaran: And then you don’t get.
283 00:30:01.370 ⇒ 00:30:02.130 Payas Parab: The.
284 00:30:02.130 ⇒ 00:30:04.069 Uttam Kumaran: Cool, Nico. We’ll talk in slack about it. Yeah.
285 00:30:05.470 ⇒ 00:30:06.020 Payas Parab: I guess.
286 00:30:08.450 ⇒ 00:30:09.170 Uttam Kumaran: Basis.
287 00:30:11.420 ⇒ 00:30:14.999 Nicolas Sucari: Hey, Ryan, do you wanna keep one win here?
288 00:30:15.000 ⇒ 00:30:16.490 Luke Daque: Yeah, sure. We can.
289 00:30:17.064 ⇒ 00:30:23.955 Nicolas Sucari: This index task. Yeah, this index task is is that directly
290 00:30:25.370 ⇒ 00:30:35.050 Nicolas Sucari: it, it has to be has to be something around on the Daily Kpis report updating or anything regarding that. Because I think
291 00:30:35.260 ⇒ 00:30:40.310 Nicolas Sucari: I don’t know why. But that’s the only report that is not Updating. I was just checking that
292 00:30:40.800 ⇒ 00:30:42.729 Nicolas Sucari: the scheme is saying that.
293 00:30:42.950 ⇒ 00:30:43.630 Luke Daque: Yeah.
294 00:30:43.630 ⇒ 00:30:44.760 Nicolas Sucari: It’s not saying that.
295 00:30:44.760 ⇒ 00:30:47.789 Luke Daque: She’s not 12, something.
296 00:30:48.180 ⇒ 00:30:54.100 Nicolas Sucari: And that happened like on December 10.th So it should be that right.
297 00:30:55.310 ⇒ 00:31:09.779 Luke Daque: Let me check if, like, daily Kpi has anything really like using any of the sources from
298 00:31:12.500 ⇒ 00:31:17.429 Luke Daque: Zendesk tickets doesn’t look like it.
299 00:31:19.000 ⇒ 00:31:20.059 Luke Daque: Yeah, it doesn’t
300 00:31:23.060 ⇒ 00:31:27.030 Luke Daque: look like it’s using Zendesk.
301 00:31:27.770 ⇒ 00:31:33.850 Luke Daque: But yeah, it’s weird. If I look at even in Snowflake.
302 00:31:35.450 ⇒ 00:31:37.830 Luke Daque: Yeah, let me share my screen so you can see.
303 00:31:37.830 ⇒ 00:31:38.540 Nicolas Sucari: Okay.
304 00:31:44.050 ⇒ 00:31:47.009 Nicolas Sucari: Oh, yeah. Other reports, too. We don’t have.
305 00:31:48.760 ⇒ 00:31:49.080 Luke Daque: Yeah.
306 00:31:49.080 ⇒ 00:31:50.320 Nicolas Sucari: Don’t have data.
307 00:31:51.410 ⇒ 00:31:51.920 Luke Daque: Okay.
308 00:31:51.920 ⇒ 00:31:53.330 Luke Daque: Viewing it, my screen.
309 00:31:53.560 ⇒ 00:31:54.340 Nicolas Sucari: Yes.
310 00:31:55.030 ⇒ 00:31:56.020 Luke Daque: So I’m
311 00:31:56.910 ⇒ 00:32:03.080 Luke Daque: trying to look yeah, in in Snowflake. Right? It does have it. The data is only up to
312 00:32:04.400 ⇒ 00:32:09.100 Luke Daque: December 11, th right? So which is weird.
313 00:32:10.268 ⇒ 00:32:16.435 Luke Daque: I did try to run that in in
314 00:32:17.430 ⇒ 00:32:22.389 Luke Daque: Dbt, development. So daily, Kpi, what was that?
315 00:32:28.160 ⇒ 00:32:31.579 Luke Daque: And yeah, it looks like, we actually do have
316 00:32:32.090 ⇒ 00:32:43.290 Luke Daque: December 18 data. So I wonder if there’s any error in our runs that’s causing. This
317 00:32:47.300 ⇒ 00:32:51.279 Luke Daque: doesn’t look like we have any errors, though. So it’s.
318 00:32:52.110 ⇒ 00:32:56.190 Nicolas Sucari: No, yeah. I don’t know what. Why, what happens.
319 00:32:56.950 ⇒ 00:33:01.740 Luke Daque: The only error here is the the Pr. But that’s a different thing.
320 00:33:14.700 ⇒ 00:33:17.240 Luke Daque: I thought this was fixed last time right.
321 00:33:18.500 ⇒ 00:33:22.030 Nicolas Sucari: Yeah, me, too. I don’t know what happened.
322 00:33:22.430 ⇒ 00:33:25.230 Nicolas Sucari: I’m just seeing your snowflake, by the way, but don’t worry.
323 00:33:26.230 ⇒ 00:33:28.321 Luke Daque: Oh, you can only see Snowflake.
324 00:33:29.950 ⇒ 00:33:31.510 Luke Daque: Let me try to share again.
325 00:33:58.420 ⇒ 00:34:03.330 Nicolas Sucari: I don’t know what happened, but it’s weird.
326 00:34:12.270 ⇒ 00:34:18.880 Nicolas Sucari: maybe something with the with the credentials in Fivetran or
327 00:34:19.179 ⇒ 00:34:22.360 Nicolas Sucari: something. But we are still getting like we have all the
328 00:34:23.310 ⇒ 00:34:25.620 Nicolas Sucari: that are coming through. I don’t know why.
329 00:34:46.550 ⇒ 00:34:47.190 Luke Daque: Hello!
330 00:34:47.929 ⇒ 00:34:48.479 Nicolas Sucari: Yep.
331 00:34:55.109 ⇒ 00:34:56.779 Luke Daque: Yeah. Can you see my screen? Now?
332 00:34:57.240 ⇒ 00:34:58.020 Nicolas Sucari: Yes.
333 00:34:59.690 ⇒ 00:35:02.392 Luke Daque: Yeah, this is a different error, though.
334 00:35:03.510 ⇒ 00:35:09.420 Luke Daque: so. But if we look at the latest run
335 00:35:10.220 ⇒ 00:35:16.690 Luke Daque: I haven’t checked this yet. But oh, so it did fail.
336 00:35:20.180 ⇒ 00:35:21.200 Luke Daque: Yeah, this is sweet.
337 00:35:21.200 ⇒ 00:35:22.939 Luke Daque: It feels good.
338 00:35:22.940 ⇒ 00:35:26.770 Luke Daque: It was. It’s showing like it was okay.
339 00:35:27.830 ⇒ 00:35:28.979 Nicolas Sucari: Yeah, that’s weird.
340 00:35:33.410 ⇒ 00:35:35.352 Luke Daque: Yeah, I’ll have to work on listening.
341 00:35:35.920 ⇒ 00:35:41.641 Luke Daque: fix this the the tricky part here is that we are actually using
342 00:35:43.460 ⇒ 00:35:47.720 Luke Daque: a model that’s coming from a package.
343 00:35:49.130 ⇒ 00:35:53.520 Luke Daque: This one, the 5 trans Zendesk package.
344 00:35:54.030 ⇒ 00:35:56.980 Luke Daque: And it looks like the failure happened in
345 00:35:57.460 ⇒ 00:36:04.000 Luke Daque: in this one. So if if we look at this, there’s.
346 00:36:07.200 ⇒ 00:36:07.890 Nicolas Sucari: Okay.
347 00:36:08.450 ⇒ 00:36:10.979 Nicolas Sucari: It got installed correctly. But then.
348 00:36:13.990 ⇒ 00:36:14.340 Luke Daque: Yeah.
349 00:36:14.340 ⇒ 00:36:15.130 Nicolas Sucari: It’s like.
350 00:36:15.130 ⇒ 00:36:21.640 Luke Daque: Really in this one the in Zendesk failed. History. Scd.
351 00:36:21.990 ⇒ 00:36:28.540 Luke Daque: and this is coming from one of the packages here in Zendesk.
352 00:36:32.287 ⇒ 00:36:33.720 Luke Daque: I think here.
353 00:36:38.070 ⇒ 00:36:44.299 Luke Daque: yeah, who’s that in the field history.
354 00:36:48.410 ⇒ 00:36:49.689 Nicolas Sucari: We don’t have it there.
355 00:36:50.910 ⇒ 00:36:54.179 Nicolas Sucari: Oh, yeah, there it is. Send the ticket fields. History sequel.
356 00:36:58.320 ⇒ 00:36:59.160 Luke Daque: Yeah.
357 00:37:03.480 ⇒ 00:37:10.379 Luke Daque: yeah. I I wonder if there’s any update to the package, though? Maybe that’s causing the the error.
358 00:37:13.060 ⇒ 00:37:25.930 Luke Daque: Oh, packages 14 to 2.26 point 14 to 26.
359 00:37:36.730 ⇒ 00:37:38.129 Luke Daque: Looks like we are.
360 00:37:39.270 ⇒ 00:37:40.810 Luke Daque: We should be good.
361 00:37:42.810 ⇒ 00:37:46.719 Luke Daque: Yeah, I’ll I’ll work on this. So I think this is
362 00:37:46.830 ⇒ 00:37:51.839 Luke Daque: if we can fix this Zendesk thing. Then, yeah, we should be good.
363 00:37:52.670 ⇒ 00:37:57.779 Nicolas Sucari: With the data like the the. That’s why it’s not updating. Yeah, I think it’s.
364 00:37:58.450 ⇒ 00:38:02.250 Luke Daque: Yeah. Cause when, if we look at the actions here.
365 00:38:02.710 ⇒ 00:38:03.520 Nicolas Sucari: Okay.
366 00:38:03.520 ⇒ 00:38:09.149 Luke Daque: It was a it failed, and in because it failed, it did not run. The succeeding.
367 00:38:09.920 ⇒ 00:38:10.970 Nicolas Sucari: Okay.
368 00:38:11.750 ⇒ 00:38:18.160 Luke Daque: Yeah models. So maybe, like the Daily Kpi was not rat as well
369 00:38:19.083 ⇒ 00:38:25.210 Luke Daque: maybe, for now, for now maybe I can just run manually. So that real?
370 00:38:25.210 ⇒ 00:38:26.170 Luke Daque: Yeah, it’s.
371 00:38:26.520 ⇒ 00:38:29.499 Nicolas Sucari: Let’s try to do that. But if we.
372 00:38:30.350 ⇒ 00:38:31.020 Luke Daque: But
373 00:38:32.890 ⇒ 00:38:36.029 Luke Daque: Oh, yeah, I can just run everything that should be
374 00:38:38.370 ⇒ 00:38:41.070 Luke Daque: so that it will be everything will be updated.
375 00:38:41.560 ⇒ 00:38:46.790 Nicolas Sucari: Yeah, okay, perfect. Yeah, let’s try doing that, because I’m also seeing that
376 00:38:47.850 ⇒ 00:38:54.510 Nicolas Sucari: if I go to like teams weekly report it. Also, it doesn’t have like any data.
377 00:38:54.510 ⇒ 00:38:58.440 Luke Daque: Data, I guess, since December 11 or something.
378 00:38:58.670 ⇒ 00:39:02.710 Nicolas Sucari: Yes, in December. Yeah, since December 8th or yeah, I think.
379 00:39:02.710 ⇒ 00:39:04.430 Luke Daque: Hmm, okay.
380 00:39:07.251 ⇒ 00:39:10.020 Nicolas Sucari: Let me check all orders. Maybe.
381 00:39:11.390 ⇒ 00:39:19.130 Nicolas Sucari: So, all orders. So everything that’s coming from shopify and Amazon. I think that’s okay. Because we have data until yesterday.
382 00:39:19.960 ⇒ 00:39:26.140 Nicolas Sucari: but the other reports no marketing.
383 00:39:26.140 ⇒ 00:39:26.480 Luke Daque: Okay.
384 00:39:26.480 ⇒ 00:39:27.720 Nicolas Sucari: One let me see.
385 00:39:30.890 ⇒ 00:39:32.640 Luke Daque: What’s the previous
386 00:39:36.210 ⇒ 00:39:40.310 Nicolas Sucari: Shipments. Also, we don’t have like the that updated.
387 00:39:40.310 ⇒ 00:39:40.650 Nicolas Sucari: Yeah.
388 00:39:41.580 ⇒ 00:39:46.980 Luke Daque: It looks like they had a new release last week. This this 19.1 release.
389 00:39:47.340 ⇒ 00:39:47.780 Nicolas Sucari: Okay.
390 00:39:47.780 ⇒ 00:39:49.819 Luke Daque: And maybe that.
391 00:39:51.590 ⇒ 00:39:53.649 Nicolas Sucari: Yeah, maybe that screwed us. Yeah.
392 00:39:53.650 ⇒ 00:39:57.279 Luke Daque: So maybe I can revert to the previous release
393 00:39:57.790 ⇒ 00:40:00.209 Luke Daque: that worked, which is, maybe this one.
394 00:40:00.980 ⇒ 00:40:01.600 Nicolas Sucari: Okay.
395 00:40:03.120 ⇒ 00:40:07.880 Luke Daque: Yeah, but for now yeah, I’ll I’ll do that. I’ll do the manual run for everything.
396 00:40:08.020 ⇒ 00:40:12.070 Luke Daque: and then I’ll let you know. Once it’s done. I’ll we can check
397 00:40:12.750 ⇒ 00:40:13.720 Nicolas Sucari: Okay.
398 00:40:14.050 ⇒ 00:40:16.880 Luke Daque: One more thing, though, I think, are you?
399 00:40:17.110 ⇒ 00:40:29.690 Luke Daque: Yeah, this is regarding their real. There’s 2 real projects for pool. Partridge is weird, the real one.
400 00:40:29.690 ⇒ 00:40:35.419 Luke Daque: Yeah, you can see there’s like pool parts to go, and there’s pool parts to go real.
401 00:40:36.350 ⇒ 00:40:37.330 Nicolas Sucari: Let me check.
402 00:40:37.550 ⇒ 00:40:43.670 Luke Daque: And even if I look at in the terminal, so if I do real project list.
403 00:40:44.670 ⇒ 00:40:47.149 Nicolas Sucari: No, I can’t see it. I don’t know why.
404 00:40:47.510 ⇒ 00:40:48.770 Luke Daque: That’s weird.
405 00:40:49.530 ⇒ 00:40:53.529 Nicolas Sucari: And what what but do do you have the 2 folders there in the.
406 00:40:53.950 ⇒ 00:40:58.270 Luke Daque: No, it’s only one folder, it’s it’s it’s the same thing.
407 00:40:58.270 ⇒ 00:41:03.250 Nicolas Sucari: And what what are the like? The dashboards as we have in the the other one in the forecast.
408 00:41:03.700 ⇒ 00:41:04.460 Nicolas Sucari: That’s for real.
409 00:41:04.750 ⇒ 00:41:11.720 Luke Daque: It’s the same. If we go here. Yeah, we have everything. And even in here the Zendesk tickets is even.
410 00:41:12.880 ⇒ 00:41:13.380 Luke Daque: It’s working.
411 00:41:13.750 ⇒ 00:41:14.859 Luke Daque: Yeah, it’s working.
412 00:41:15.200 ⇒ 00:41:18.719 Luke Daque: But in the other 1, 0, yeah, it looks like it’s already
413 00:41:19.220 ⇒ 00:41:22.550 Luke Daque: as well, cause I already cause I did that refresh
414 00:41:24.360 ⇒ 00:41:28.110 Nicolas Sucari: Check the the last date if you want. Yeah, okay.
415 00:41:30.580 ⇒ 00:41:31.840 Luke Daque: December 18. Yeah.
416 00:41:32.420 ⇒ 00:41:33.620 Luke Daque: Yeah.
417 00:41:34.110 ⇒ 00:41:40.739 Luke Daque: yeah, this is weird. And I can’t delete it, because if I do a pool project real project delete
418 00:41:41.670 ⇒ 00:41:46.229 Luke Daque: it can. It will only delete the the 1st one, which is
419 00:41:46.600 ⇒ 00:41:49.039 Luke Daque: not what we want, because this is what’s
420 00:41:49.940 ⇒ 00:41:53.800 Luke Daque: that’s this is what we are all using like even Kim and.
421 00:41:53.800 ⇒ 00:41:57.089 Nicolas Sucari: Yeah, I’m not seeing the other one. I’m just seeing one.
422 00:41:58.153 ⇒ 00:41:59.180 Luke Daque: That’s weird.
423 00:41:59.530 ⇒ 00:42:01.150 Luke Daque: So maybe this is just.
424 00:42:05.060 ⇒ 00:42:08.500 Nicolas Sucari: I don’t know, because I’m not seeing like the other.
425 00:42:10.480 ⇒ 00:42:16.380 Luke Daque: That’s yeah. That’s weird settings, can I?
426 00:42:16.380 ⇒ 00:42:21.250 Nicolas Sucari: And you are also go to the users there in corporate. Can you go to.
427 00:42:23.550 ⇒ 00:42:26.759 Luke Daque: Yeah, click on, your, on, on your user.
428 00:42:28.880 ⇒ 00:42:29.910 Nicolas Sucari: Go to.
429 00:42:29.910 ⇒ 00:42:30.320 Luke Daque: Air.
430 00:42:30.320 ⇒ 00:42:34.499 Nicolas Sucari: Why, why, it says, enterprise your poor parts to go.
431 00:42:35.480 ⇒ 00:42:38.200 Nicolas Sucari: or you are. You are logged in as.
432 00:42:39.050 ⇒ 00:42:41.019 Luke Daque: I’m logged in as me.
433 00:42:44.760 ⇒ 00:42:47.429 Nicolas Sucari: Oh, because maybe you are that
434 00:42:48.510 ⇒ 00:42:51.820 Nicolas Sucari: everyone from poor to go I don’t know.
435 00:42:52.990 ⇒ 00:42:58.439 Nicolas Sucari: and if you go to the other, go to the other one, the other project, not the dash, real
436 00:42:59.310 ⇒ 00:43:01.449 Nicolas Sucari: and click on the users to see that.
437 00:43:01.780 ⇒ 00:43:03.609 Nicolas Sucari: Yeah. You see, I’m seeing this.
438 00:43:04.360 ⇒ 00:43:05.050 Luke Daque: Hmm.
439 00:43:05.540 ⇒ 00:43:06.330 Nicolas Sucari: Can you?
440 00:43:07.040 ⇒ 00:43:08.220 Nicolas Sucari: Yeah, I don’t know.
441 00:43:08.220 ⇒ 00:43:09.370 Luke Daque: It’s weird.
442 00:43:09.690 ⇒ 00:43:10.760 Nicolas Sucari: Super weird.
443 00:43:11.700 ⇒ 00:43:15.639 Nicolas Sucari: And Payas is using this one, too, like this is the one that we are using right.
444 00:43:16.090 ⇒ 00:43:20.889 Luke Daque: Yeah, can. Can you check just
445 00:43:21.200 ⇒ 00:43:26.069 Luke Daque: if if Zendesk ticket is already fixed in this project.
446 00:43:26.750 ⇒ 00:43:33.320 Nicolas Sucari: Yeah, it’s fixed. I can see it now. But send. This is fixed, but not the Daily Kpi. You need to rewrite right.
447 00:43:33.320 ⇒ 00:43:37.790 Luke Daque: Yeah, yeah, I need to run that. It’s still it should still be running, I believe.
448 00:43:38.220 ⇒ 00:43:41.619 Luke Daque: Yeah. But yeah, there’s an error in Kim’s me to report up.
449 00:43:41.750 ⇒ 00:43:42.440 Luke Daque: Check this.
450 00:43:42.440 ⇒ 00:43:43.080 Nicolas Sucari: Okay.
451 00:43:44.894 ⇒ 00:43:47.080 Luke Daque: Yeah, I’ll continue working on this.
452 00:43:47.510 ⇒ 00:43:53.610 Nicolas Sucari: Okay, let me know if you need anything else. Okay, thanks. Bye, bye.
453 00:43:54.180 ⇒ 00:43:56.580 Luke Daque: No worries, thanks. See you bye, bye.