Meeting Title: [Javvy + Brainforge] Weekly Sync Date: 2025-04-07 Meeting participants: Aakash Tandel, Aman Nagpal, Robert Tseng
WEBVTT
1 00:00:39.410 ⇒ 00:00:40.490 Aakash Tandel: Hey, Robert.
2 00:00:42.020 ⇒ 00:00:42.820 Robert Tseng: Hey? Gosh!
3 00:00:54.350 ⇒ 00:00:59.421 Robert Tseng: I was mainly just gonna run through the stuff that’s pending client feedback with him.
4 00:01:00.360 ⇒ 00:01:03.739 Robert Tseng: I know that you guys have had another thread kind of like
5 00:01:04.830 ⇒ 00:01:07.421 Robert Tseng: about the hours and stuff.
6 00:01:10.550 ⇒ 00:01:16.650 Robert Tseng: yeah, I mean, do you? Do you feel like I mean, I’ll I guess I imagine he’ll probably bring that up. But.
7 00:01:16.650 ⇒ 00:01:19.656 Aakash Tandel: Yeah, we can talk through that if if we get there
8 00:01:20.100 ⇒ 00:01:26.499 Aakash Tandel: Happy to talk to it. I was also gonna have the roadmap stuff in the background. If we need to talk about.
9 00:01:26.750 ⇒ 00:01:31.747 Robert Tseng: Okay, yeah, I’m gonna try not to bring that up unless he really wants to. Cause.
10 00:01:32.350 ⇒ 00:01:33.100 Robert Tseng: Yeah.
11 00:01:33.690 ⇒ 00:01:34.470 Aakash Tandel: Sounds good.
12 00:01:44.090 ⇒ 00:01:45.340 Aman Nagpal: Hey? How’s it going? Guys?
13 00:01:47.210 ⇒ 00:01:47.820 Robert Tseng: Ron.
14 00:01:48.810 ⇒ 00:01:50.009 Aakash Tandel: Hello! How’s it going.
15 00:01:50.790 ⇒ 00:01:51.959 Aman Nagpal: Good. How about you?
16 00:01:53.760 ⇒ 00:01:57.709 Aakash Tandel: Pretty, good, pretty good Robert. Still, in Los Angeles time.
17 00:01:59.180 ⇒ 00:01:59.880 Aman Nagpal: Nice.
18 00:02:06.120 ⇒ 00:02:09.120 Aakash Tandel: Give me one second while I move stuff around.
19 00:02:09.380 ⇒ 00:02:12.589 Robert Tseng: Sure you’re gonna share. You’re gonna share screen. Gosh.
20 00:02:12.590 ⇒ 00:02:16.062 Aakash Tandel: Yeah. Do you want me to share the roadmap screen or the
21 00:02:16.840 ⇒ 00:02:21.299 Robert Tseng: Yeah, let’s do roadmap with all the shoes, and we’ll just hover over the pending client. Feedback column.
22 00:02:21.590 ⇒ 00:02:22.812 Aakash Tandel: Let’s do that
23 00:02:30.320 ⇒ 00:02:32.789 Aakash Tandel: here, this guy.
24 00:02:40.500 ⇒ 00:02:42.200 Aakash Tandel: what’s happening? Why is this blank?
25 00:02:48.540 ⇒ 00:02:51.609 Aakash Tandel: What is happening? Linear, broken is linear down.
26 00:02:52.010 ⇒ 00:02:57.629 Robert Tseng: It was kind of weird for me this morning, too, so I can. I can share a screen if it’s easier.
27 00:02:57.950 ⇒ 00:02:59.279 Robert Tseng: I just send a request.
28 00:02:59.440 ⇒ 00:03:02.589 Aakash Tandel: Yeah, okay, I’ll let you share. That’s weird.
29 00:03:02.770 ⇒ 00:03:03.400 Robert Tseng: Okay.
30 00:03:14.410 ⇒ 00:03:15.370 Robert Tseng: cool.
31 00:03:17.810 ⇒ 00:03:24.719 Robert Tseng: yeah. I mean, just a few things that we need. We wanted to keep moving. I mean the shopify address stuff, I think
32 00:03:25.300 ⇒ 00:03:31.353 Robert Tseng: both Annie ran the ran a script, and then we had Pius add add in the new
33 00:03:31.900 ⇒ 00:03:35.210 Robert Tseng: logic for the Netsuite file, and then we sent that over to you.
34 00:03:36.240 ⇒ 00:03:40.979 Aman Nagpal: That’s the one where Robbie asked for the one where it’s just address, only.
35 00:03:41.790 ⇒ 00:03:42.190 Robert Tseng: Yep!
36 00:03:42.460 ⇒ 00:03:45.669 Aman Nagpal: Gotcha. And so that python script do we have that updated one.
37 00:03:46.170 ⇒ 00:03:47.340 Robert Tseng: Yep, it’s updated.
38 00:03:47.800 ⇒ 00:03:49.600 Aman Nagpal: Where? Where can I find that.
39 00:03:50.483 ⇒ 00:03:51.389 Robert Tseng: There is.
40 00:03:52.060 ⇒ 00:03:53.130 Aakash Tandel: 4 pisses.
41 00:03:53.130 ⇒ 00:03:54.530 Robert Tseng: Tag him in that again. Yeah.
42 00:03:58.650 ⇒ 00:04:01.030 Aman Nagpal: We’ll try to get a ui made for that. Then.
43 00:04:01.310 ⇒ 00:04:01.900 Robert Tseng: Yep
44 00:04:10.030 ⇒ 00:04:20.130 Robert Tseng: cool. I’m just gonna leave in here until you look at it and then see if you need anything else. And then yeah, I mean on the Amazon cog stuff. Seems like we’re still kind of
45 00:04:21.500 ⇒ 00:04:25.700 Robert Tseng: pushing back and forth. I I don’t believe that
46 00:04:26.170 ⇒ 00:04:30.540 Robert Tseng: Blake has filled it filled things out yet. So I think that’s where we’re still waiting on that.
47 00:04:31.540 ⇒ 00:04:44.269 Aman Nagpal: Let me. Sorry also for the other thing. Did once you guys sent back the updated matched sheet with just the address one did Robbie say he was fine with it? Or do you.
48 00:04:44.270 ⇒ 00:04:51.990 Aakash Tandel: He didn’t respond yet. Pi sent it on Saturday
49 00:04:52.360 ⇒ 00:05:04.600 Aakash Tandel: morning. So oh, no, technically, really. Early in the morning it was late Friday Saturday morning. If on Eastern Standard time, so maybe he hasn’t seen. Maybe Robbie hasn’t seen it yet.
50 00:05:06.160 ⇒ 00:05:11.320 Aman Nagpal: Let me just while we’re going in order. I don’t wanna forget it. Let’s just ping him.
51 00:05:12.190 ⇒ 00:05:16.040 Robert Tseng: I mean, Robbie responded. Sorry he reacted to it, but he didn’t respond.
52 00:05:16.300 ⇒ 00:05:17.988 Robert Tseng: Yes, yeah, I can.
53 00:05:20.280 ⇒ 00:05:24.450 Aakash Tandel: Do you want me to add him and ask him, or I’ll let actually, I can let you handle that.
54 00:05:24.898 ⇒ 00:05:31.299 Aman Nagpal: Do. Yeah, whatever I’m having trouble finding the channel. Oh, no, no. I found her. Okay, let me let me tag him.
55 00:05:40.900 ⇒ 00:05:48.099 Aman Nagpal: Okay? And then the other one, the next one. We’re waiting on Blake, right? We. So we got some of the information. I think Aliyah put in some of it. Right? Was that fine.
56 00:05:48.830 ⇒ 00:05:50.969 Robert Tseng: Yeah, it was fine for what she put in. Yeah.
57 00:05:51.650 ⇒ 00:05:56.080 Aman Nagpal: Okay? And then Blake said, 1214. Today, we’re working on it right now. So
58 00:05:56.430 ⇒ 00:05:58.300 Aman Nagpal: hopefully, we get that today.
59 00:05:59.065 ⇒ 00:06:00.670 Aman Nagpal: Cool, what’s next?
60 00:06:01.350 ⇒ 00:06:21.320 Robert Tseng: Cool, and then on the shopify. Dash I mean, this is, there was no work done in the past week on it, but we still kind of after we tested a bunch of orders we send it back to. You don’t know if you’ve looked at it yet. So it was just just a bump on that. Can we call that done? Or if you wanted to, you wanted to see anything else on it?
61 00:06:21.830 ⇒ 00:06:23.310 Aman Nagpal: Per gross margin.
62 00:06:23.310 ⇒ 00:06:23.870 Robert Tseng: Yep.
63 00:06:25.360 ⇒ 00:06:27.359 Aman Nagpal: Let me take a look.
64 00:06:36.980 ⇒ 00:06:39.530 Aman Nagpal: you know which of our channels it was.
65 00:06:43.480 ⇒ 00:06:45.719 Aman Nagpal: If it’s on here. That’s fine on that card.
66 00:06:52.880 ⇒ 00:06:57.069 Robert Tseng: No, it’s it’s yeah. I mean, I just, I just go to Meta Base and just pull this up.
67 00:07:02.610 ⇒ 00:07:05.919 Aakash Tandel: Linear is definitely acting up. I like, can’t load tickets right now.
68 00:07:07.890 ⇒ 00:07:11.329 Robert Tseng: Yeah, it’s there’s it’s a bit off.
69 00:07:11.640 ⇒ 00:07:16.299 Aman Nagpal: But yeah, I’ll I’ll take another look at this but basically on your end.
70 00:07:16.620 ⇒ 00:07:19.679 Aman Nagpal: we’re pretty much at a good place with this. Is there anything else
71 00:07:19.980 ⇒ 00:07:22.889 Aman Nagpal: any changes anticipated? Here? Are we good with this.
72 00:07:23.420 ⇒ 00:07:39.080 Robert Tseng: No, no changes anticipated. I guess we we added recently a like a new field that was like for subscription order. So if we wanted to do 1st subscription order, gross profit, pivot like we could add that. But that wasn’t. That was just kind of like a 1 off.
73 00:07:39.660 ⇒ 00:07:43.149 Robert Tseng: It wasn’t clear if we wanted to add it to the dash or not, so.
74 00:07:44.783 ⇒ 00:07:47.050 Aman Nagpal: Yeah, I mean, we can keep this as
75 00:07:47.390 ⇒ 00:07:52.089 Aman Nagpal: is, for now I just need to check it on my end to make sure the orders are good.
76 00:07:52.090 ⇒ 00:07:52.630 Robert Tseng: Yeah.
77 00:07:56.350 ⇒ 00:08:02.456 Robert Tseng: okay, so those are the things that are pending client feedback. Right? Now.
78 00:08:03.420 ⇒ 00:08:09.509 Robert Tseng: yeah, the cohort retention. I mean, we’re we just finished the modeling on that. So then that’ll probably
79 00:08:09.800 ⇒ 00:08:11.250 Robert Tseng: another day to finish.
80 00:08:13.230 ⇒ 00:08:15.879 Robert Tseng: Alright, yeah, me Akash, I think that’s that’s the
81 00:08:16.220 ⇒ 00:08:18.810 Robert Tseng: that’s the main feedback stuff I wanted to get. You can
82 00:08:18.930 ⇒ 00:08:21.379 Robert Tseng: kind of take it back with your Google sheet if you want.
83 00:08:21.510 ⇒ 00:08:24.650 Aakash Tandel: Sure. Let me go back to here.
84 00:08:25.200 ⇒ 00:08:33.169 Aakash Tandel: Actually, I’m on. If there’s anything you wanted to talk about first, st I was gonna walk through this spreadsheet that we have. But if there’s anything else you want to talk about.
85 00:08:36.230 ⇒ 00:08:43.549 Aman Nagpal: no. The the only thing I would ask is where we’re at with those looms from autumn about the engineering piece.
86 00:08:44.110 ⇒ 00:09:04.029 Aakash Tandel: Yeah, I don’t know. I couldn’t find the thread where we talked about that, or exactly what we needed on that guy. So Utham hasn’t. He was originally going to do some sort of notion documentation. I’m not sure what you guys are looking for in the looms, but we can talk about that, and I also don’t see it here. I know that
87 00:09:05.720 ⇒ 00:09:12.160 Aman Nagpal: I guess the way, so I don’t think we had a threat. I think it was on that call with Annie. But you know.
88 00:09:12.300 ⇒ 00:09:15.269 Aman Nagpal: I think you were saying we can only go so far in the
89 00:09:15.410 ⇒ 00:09:25.779 Aman Nagpal: kind of funnel with Annie. She’s more on this side, and then everything beyond that will be more with them. So yeah, notion, Doc would help. Looms, I think, would help anything from everything
90 00:09:26.010 ⇒ 00:09:37.399 Aman Nagpal: from beginning to end. Right? Like, you know, this is portable is connected in this way. 5 trans. Connect. For this also I’m a little confused. Why, the 5 trend bill is so high. If it’s just Amazon.
91 00:09:37.815 ⇒ 00:09:44.349 Aman Nagpal: Still, couple of 100 bucks. Maybe. I don’t know. Maybe that’s what it’s supposed to be, if you can let me know. But yeah, 5 trans connected for this.
92 00:09:45.640 ⇒ 00:09:48.179 Aman Nagpal: Dbt, you know, we’ve created.
93 00:09:48.600 ⇒ 00:09:51.260 Aman Nagpal: We’ve cleaned up Xyz order
94 00:09:51.570 ⇒ 00:10:00.389 Aman Nagpal: tables and made it into this order table. And this is the clean one, and that’s the old one, and this combines Amazon shopify. And then all the different tables.
95 00:10:01.263 ⇒ 00:10:15.439 Aman Nagpal: And what else is happening in Dbt, you know, is I I’m assuming it sounds like metrics are separate, and the the tables or models are separate. I don’t know if I have that right, just anything and everything. If we can just have that all documented would be great.
96 00:10:15.770 ⇒ 00:10:19.609 Aakash Tandel: Sure. Yeah, and you wanted to walk Vlad through that type of thing, too.
97 00:10:20.060 ⇒ 00:10:22.969 Aman Nagpal: Yeah, it would be Vlad and myself. But also, if we have that
98 00:10:24.930 ⇒ 00:10:28.039 Aman Nagpal: notion also, or whatever it is, that would be super helpful.
99 00:10:28.040 ⇒ 00:10:42.630 Aakash Tandel: Okay, yeah. With them, time is pretty booked. So it’ll probably be a wish walking us through that, because he’s obviously got as similar knowledge set. So we’ll have like something like a wish to walk us through
100 00:10:43.560 ⇒ 00:10:46.540 Aakash Tandel: data sources to Snowflake.
101 00:10:49.020 ⇒ 00:10:53.250 Aakash Tandel: And then maybe the
102 00:10:53.450 ⇒ 00:11:03.959 Aakash Tandel: Robert, how many Dbt models do we have? Is it so many that it’s too much to walk through in a call, or is it within reason to walk through those in a call.
103 00:11:04.820 ⇒ 00:11:10.770 Robert Tseng: And we could do it in a call. But like we could, just, I feel like we should just do it on a loop. We don’t need to schedule anything for stuff.
104 00:11:10.770 ⇒ 00:11:11.549 Aman Nagpal: That works like, yeah.
105 00:11:11.550 ⇒ 00:11:14.009 Robert Tseng: For them to to re-re-reuse.
106 00:11:14.210 ⇒ 00:11:16.460 Aakash Tandel: Yeah. Dbt, models.
107 00:11:18.166 ⇒ 00:11:21.200 Aakash Tandel: Okay? And then I think.
108 00:11:21.200 ⇒ 00:11:25.490 Aman Nagpal: When we say Dvt models, we mean the combined and cleaned up tables right.
109 00:11:26.180 ⇒ 00:11:47.349 Aakash Tandel: Yeah, basically, they use Dvc, just uses SQL to clean up a a table. So it’ll go from like one table to another, and it’ll modify like, I don’t know like it’ll make some sort of calculation and change like raw numbers into a readable output for you guys, for I don’t know for a specific thing. That’s kind of how those that that type of thing works.
110 00:11:47.630 ⇒ 00:11:49.880 Aman Nagpal: And is every single.
111 00:11:50.180 ⇒ 00:11:51.210 Aakash Tandel: Source.
112 00:11:51.340 ⇒ 00:11:55.800 Aman Nagpal: Data source model 3. Dbt, it? Ha! Everyone has a cleanup version.
113 00:11:57.318 ⇒ 00:12:06.250 Aakash Tandel: I don’t know off the top of my head, but everything in order for it to like. Get all the way to Meta base and be clean, that everything there will have been through a model.
114 00:12:06.470 ⇒ 00:12:07.080 Aman Nagpal: Got it.
115 00:12:07.240 ⇒ 00:12:09.210 Aman Nagpal: Well, if it’s not through dbt.
116 00:12:09.720 ⇒ 00:12:12.289 Aman Nagpal: where else would it have been cleaned up.
117 00:12:14.566 ⇒ 00:12:30.269 Aakash Tandel: Like. Oh, so everything that’s cleaned up will have been cleaned up through. Dbt, there is. I guess there could be a situation where there’s like a data source that’s going to snowflake that we just don’t use. And it’s just getting raw data. But I don’t know true or not. I don’t think so right, Robert.
118 00:12:30.400 ⇒ 00:12:40.709 Robert Tseng: Well, it’s like we’re always like, Oh, we should turn this on or test this connector. What we haven’t. We don’t model for those unless there’s a clear like reporting outcome. We just
119 00:12:41.240 ⇒ 00:12:43.020 Robert Tseng: yeah. We just kind of leave them as is.
120 00:12:43.020 ⇒ 00:12:54.310 Aman Nagpal: So if it’s a new source, we don’t know what the hell we’re doing with it. Oh, let’s just connect the raw and see what happens. But if it’s we have a plan, we’re gonna use this source for this alright step one. Let’s clean up the data with Dbt model it, and then
121 00:12:54.650 ⇒ 00:12:57.069 Aman Nagpal: connected to this metabase, or whatever.
122 00:12:57.070 ⇒ 00:12:57.540 Robert Tseng: Yep.
123 00:12:57.540 ⇒ 00:12:58.160 Aakash Tandel: Yeah.
124 00:12:58.500 ⇒ 00:13:06.410 Aman Nagpal: And then separately, there are Dbt quote unquote metrics separate from all of that semantic layer metrics. I don’t know. I’m trying to gauge it all.
125 00:13:08.188 ⇒ 00:13:11.889 Robert Tseng: No, there, it’s part of the I mean, there isn’t like a
126 00:13:12.170 ⇒ 00:13:15.900 Robert Tseng: model that’s just a metrics dump like I feel like.
127 00:13:16.010 ⇒ 00:13:18.349 Aman Nagpal: Yeah, in the Martz layer, like.
128 00:13:20.290 ⇒ 00:13:34.459 Robert Tseng: Yeah, I mean, for I mean, it’s this is where it’s more of an art than a science, I feel. But anytime we’re calculating something multiple times. Then we’ll end up pushing it into the model. But it’s much faster for the analysts to just take
129 00:13:35.180 ⇒ 00:13:35.940 Robert Tseng: what
130 00:13:36.010 ⇒ 00:14:00.210 Robert Tseng: we’re exposing in the model to build metrics and to make changes, and then once it’s solidified and we should push it back into the model. So like the example would be like with cogs like cogs. I didn’t have the team model, because I mean, we were just iterating on it. Oh, what if we added platform fee processing fee? What is fully loaded cogs really mean. There were multiple iterations to like, build out cogs.
131 00:14:00.576 ⇒ 00:14:05.290 Robert Tseng: I mean, assuming that now it’s pretty settled. Then. Yeah, I can. I can.
132 00:14:05.290 ⇒ 00:14:32.610 Robert Tseng: We can. That’s like tech debt to tell the team like, okay, hey, we need to take this. Just move it into the model now, because there’s gonna be other reports like, whatever we’re doing on the gross margin side with shopify the V 2 of the Amazon dash is going to basically be a similar version of the what we have on the shopify side for Amazon. And that means we’re gonna have to reuse a lot of these cogs metrics again, and we should probably just put it in the modeling layer. So it’s easier for
133 00:14:32.610 ⇒ 00:14:38.159 Robert Tseng: for whoever’s building the report to just drag and drop rather than having to calculate in every report.
134 00:14:39.220 ⇒ 00:14:52.050 Aman Nagpal: Got it. Yeah, that that checks out on that note. Also, you mentioned the cog, so that also, if that can be in the looms. But like we’re getting cogs from the Google Sheet versus getting it from shopify, etc, where
135 00:14:52.230 ⇒ 00:14:55.720 Aman Nagpal: pretty much all the little nuances that we wouldn’t know.
136 00:14:55.840 ⇒ 00:14:58.249 Aman Nagpal: That’s kind of what we need. Documentation on.
137 00:14:59.210 ⇒ 00:15:01.440 Robert Tseng: Yeah, I mean, that stuff is
138 00:15:01.550 ⇒ 00:15:12.149 Robert Tseng: kind of, I mean, whenever I in the dashboards, we have like key definitions, assumptions like everything’s kind of spelled out there. But I mean, we have the data platform documentation Akash, that we
139 00:15:12.320 ⇒ 00:15:23.190 Robert Tseng: kind of just throw the same things into so like that. That probably would just be in in that Google sheet that we that we put together for data platform documentation.
140 00:15:24.170 ⇒ 00:15:25.370 Aakash Tandel: Yeah, the ingestion of that data.
141 00:15:25.370 ⇒ 00:15:25.910 Robert Tseng: Yeah.
142 00:15:26.350 ⇒ 00:15:49.909 Aakash Tandel: Yeah, okay, cool. We’ll have a wastewalk like in the loom, or something through that, too, because I think there’s kind of 2 buckets. It’s like the Via tool bucket. So like the things like that’s going through 5 tran portable which are, I think, more understandable because it’s like you’re using a tool to paddling things. And then there’s the spreadsheet integration that’s like pulling data and appending it to the.
143 00:15:50.070 ⇒ 00:15:56.329 Aakash Tandel: to to the data that we have in snowflakes that might be like 2 separate little looms. But yeah, that makes sense.
144 00:15:56.520 ⇒ 00:15:57.110 Robert Tseng: Okay.
145 00:15:58.707 ⇒ 00:16:12.342 Aakash Tandel: Cool. And then, yeah, so I mean, I think that’s the main like 3 things. It’s like data sources. Snowflake itself like, where’s the data like, what like, maybe the main tables that they’re using. And then like, what are the models doing?
146 00:16:12.810 ⇒ 00:16:34.510 Aakash Tandel: And then, yeah, I think that that gets you to the point of like the stuff we talked about with Annie. So like she showed you like the main data in Snowflake, like what the the databases that Meta base is hitting are those like cleaned up ones, and then they get visualized in Meta Base. So I think between the thing Annie walked you through, and then this we should be able to get the whole whole pipeline together.
147 00:16:34.870 ⇒ 00:16:39.279 Aman Nagpal: Yeah. And even just, you know, okay, this, this is shopify data. This is the raw
148 00:16:39.410 ⇒ 00:16:46.869 Aman Nagpal: between raw and intermediate. We did Xyz between intermediate and martyr production, whatever it’s called. We did. You know? Xyz again.
149 00:16:47.020 ⇒ 00:16:56.490 Aman Nagpal: this is then going into Meta Base, and then that for every source, so that we know what exactly are the changes and what we should use. You know.
150 00:16:57.080 ⇒ 00:17:14.890 Aakash Tandel: Yep, that makes sense. Okay, cool. Yeah, I will. Get this broken out by like type for a ways to make zooms zoom looms for and then, yeah, we’ll send those over. I’ll I’ll send those in a piecemeal, probably. Just so that you guys aren’t waiting for one big dump and then, if you have any like
151 00:17:15.000 ⇒ 00:17:17.849 Aakash Tandel: clarifying question or something he can answer in subsequently.
152 00:17:18.270 ⇒ 00:17:19.640 Aman Nagpal: That’d be great. Thank you.
153 00:17:19.869 ⇒ 00:17:23.119 Aakash Tandel: Yeah, no problem. Okay? So that sounds good. So that’s this guy,
154 00:17:25.099 ⇒ 00:17:47.519 Aakash Tandel: I know for. So we’re basically blocked on subscribe and save, I know that there’s that slack thread that Annie kind of put together her like a data analysis inference on a user that could be subscribe and save so I guess we’re just waiting on how you all want to handle that moving forward, because there’s no direct connector to Amazon that says, Hey, this person is definitely subscribe and save.
155 00:17:47.839 ⇒ 00:17:50.409 Aakash Tandel: Let me see if I can even open this.
156 00:17:52.479 ⇒ 00:17:57.749 Aakash Tandel: Yeah, okay, oh, wait. No, it’s a.
157 00:17:58.490 ⇒ 00:18:05.710 Aman Nagpal: The one where you message, saying she flagged a customer as likely subscribe and save if they’ve ever received a 10 or 15% discount.
158 00:18:05.710 ⇒ 00:18:06.590 Aakash Tandel: Yeah, exactly.
159 00:18:06.590 ⇒ 00:18:16.690 Robert Tseng: Yeah. So basically, it’s just, hey, we’re looking for other proxies for subscribe and save. I suggested. Maybe there’s just like a recurring monthly like
160 00:18:17.280 ⇒ 00:18:20.529 Robert Tseng: transaction that happens on the same date every month.
161 00:18:20.640 ⇒ 00:18:39.349 Robert Tseng: But I think we looked into that and doesn’t really seem like that’s the case. People seem to pause, and the order history is actually pretty inconsistent. So then Annie was like, well, what if we looked at discount percentage? Maybe every order gets applied a consistent 10 to 15% discount percentage.
162 00:18:40.030 ⇒ 00:18:45.899 Robert Tseng: That’s like on an order every month, or at least a month apart. And we can assume that those are
163 00:18:46.310 ⇒ 00:18:51.040 Robert Tseng: likely subscribe and save customers. But anyway, it’s just like we’re trying to
164 00:18:52.040 ⇒ 00:18:57.589 Robert Tseng: predict or like, kind of yeah, we’re inferring, like, who who the subscribing save customers are.
165 00:18:57.590 ⇒ 00:19:01.130 Aman Nagpal: So it’s not exactly spot on. Do we know.
166 00:19:01.130 ⇒ 00:19:06.130 Robert Tseng: If it’s that important, it’s like, I think we should probably directionally get some.
167 00:19:06.320 ⇒ 00:19:09.759 Robert Tseng: We we should set a proxy. But it’s not gonna be.
168 00:19:10.270 ⇒ 00:19:12.910 Robert Tseng: It’s not the same as getting the actual label from them.
169 00:19:13.710 ⇒ 00:19:18.359 Aman Nagpal: There would be other discounts that could potentially be, not subscribe and save discounts
170 00:19:18.570 ⇒ 00:19:25.159 Aman Nagpal: that are maybe end up being 10 or 15%. Also right? I wouldn’t know. I’m just curious if you guys have seen that or you’re you don’t know either.
171 00:19:25.550 ⇒ 00:19:33.749 Aman Nagpal: Well, we were just talking anecdotally as subscribe and save customers for other products on Amazon, like, it’s usually like the Amazon, like.
172 00:19:34.070 ⇒ 00:19:37.469 Robert Tseng: Discounting is usually not that clean. It’s.
173 00:19:37.720 ⇒ 00:19:55.110 Robert Tseng: you know, random flash sales that are like minus 8% or like 18% or whatever, whereas a subscribe and save it seems pretty consistent. It’s either 10 or 15 or whatever. So I mean, I don’t know. Okay, it’s probably okay that we just reason through it that way. But there isn’t. I don’t know. We don’t know for sure.
174 00:19:55.280 ⇒ 00:19:57.580 Aakash Tandel: Yeah. And the other thing is like, we can email.
175 00:19:57.770 ⇒ 00:20:19.499 Aakash Tandel: oh, sorry. Oh, really, we we can either over like, Oh, make the net really big. So we definitely catch all the subscribe and saves or make the net really tight and be okay with losing some of the subscribe and save people like that depends on kind of like your use case and what you want to do with that information. But those are kind of like the because it’s a proxy measurement. That’s kind of the sides of the 2 sides of the coin.
176 00:20:19.980 ⇒ 00:20:32.020 Aman Nagpal: Yeah, and I don’t know off top my head, if you know, do we give X discount on the 1st order and then less. This. I I feel like it’s what 10 or 15 1st order, usually 5% on renewal. Something like that that Amazon does. I don’t know for sure.
177 00:20:32.020 ⇒ 00:20:32.440 Aakash Tandel: Yeah.
178 00:20:32.440 ⇒ 00:20:35.930 Aman Nagpal: But I think what we could do is, oh, gosh! If you can send this
179 00:20:36.370 ⇒ 00:20:40.690 Aman Nagpal: in that same thread that we’re working on the Amazon data and ask Blake.
180 00:20:40.860 ⇒ 00:21:01.760 Aman Nagpal: He should have all the answers of Okay, well, you know, the only 10 per 15% discount we ever give, or the only discount we give is subscribe and save. Maybe we don’t do other coupons. Maybe we do. He can list all of that out. He can tell us if 1st and renewal orders have a different discount. He should be able to help us. Kind of figure.
181 00:21:01.760 ⇒ 00:21:04.789 Robert Tseng: He should know everything Amazon pricing as well as
182 00:21:05.020 ⇒ 00:21:07.869 Robert Tseng: the cogs, I guess. Yeah, okay.
183 00:21:08.670 ⇒ 00:21:11.290 Aman Nagpal: So, yeah, maybe that thread is best for this. That group.
184 00:21:11.570 ⇒ 00:21:20.099 Aakash Tandel: Okay, cool. I will take Annie’s information that she put in that thread and then put it to Blake and see what in that one group thread and then see what he says.
185 00:21:20.790 ⇒ 00:21:26.900 Aman Nagpal: Thank you. Yeah. And if you can just arrange it so that he has the context of what we’re looking for, we’re trying to catch the subscribe, and saves.
186 00:21:27.070 ⇒ 00:21:34.250 Aakash Tandel: Yup, that sounds good. Okay, cool. I will. I’m just gonna take a note to send to Blake.
187 00:21:35.100 ⇒ 00:21:36.570 Aakash Tandel: Okay, cool.
188 00:21:37.176 ⇒ 00:21:48.530 Aakash Tandel: This guy’s done. Okay, that’s done. Close tickets. Close. That attentive is on oasis plate. So he he has the metrics. That kind of you all outlined. I guess previously.
189 00:21:48.530 ⇒ 00:22:09.480 Aakash Tandel: He’s going to be working on that this week. Same thing with North Beam. Him and Kyle are splitting up the different components of that amplitude dash to finish or do the modeling around those north beam connectors or north beam pieces of data that we have. So that’s also slated for this week. And that’s I think
190 00:22:10.020 ⇒ 00:22:17.059 Aakash Tandel: Klaviyo is also, I think. Let me just click on this ticket. I don’t remember which one this is.
191 00:22:20.180 ⇒ 00:22:23.869 Robert Tseng: No, we didn’t start this. It’s it. Yeah, we haven’t started this one yet.
192 00:22:23.870 ⇒ 00:22:28.120 Aakash Tandel: Yeah, this one would need to review this requirement, I guess, to
193 00:22:29.830 ⇒ 00:22:32.889 Aakash Tandel: yeah, this one looks like it’s a fair
194 00:22:33.280 ⇒ 00:22:36.859 Aakash Tandel: chunk of things. I need to split out. But yeah, that’s that’s that one.
195 00:22:38.500 ⇒ 00:22:43.080 Aakash Tandel: And then, yeah, I guess this is this is one that I I’m not sure where
196 00:22:44.730 ⇒ 00:22:49.346 Aakash Tandel: what we want, how we want to talk about this one, because I know it’s a data frequency question,
197 00:22:49.850 ⇒ 00:23:00.880 Aakash Tandel: and kind of gave us a a general answer here, but I don’t know what we want to act on that. I know that you were talking about increasing the throughput to via Portable
198 00:23:01.415 ⇒ 00:23:05.949 Aakash Tandel: and it sounded like something that we should probably just do. If you guys want more more of that fresh data.
199 00:23:07.400 ⇒ 00:23:09.639 Aman Nagpal: Yeah. I mean
200 00:23:10.070 ⇒ 00:23:16.470 Aman Nagpal: ideally. Yes, I think, Robert, you said the snowflake cost shouldn’t really go up too much.
201 00:23:17.353 ⇒ 00:23:21.310 Aman Nagpal: And portal is already flat rate, so we can just go ahead and do it.
202 00:23:22.950 ⇒ 00:23:30.889 Robert Tseng: Yeah, I mean, like, storage is not gonna change, whether we refresh more or not. If anything is just additional compute. But it’s if we just bump it up.
203 00:23:31.160 ⇒ 00:23:38.479 Robert Tseng: I mean, we we have to throttle it to kind of see how much it’ll actually cost. But I imagine if we’re just doing like
204 00:23:38.880 ⇒ 00:23:44.220 Robert Tseng: 4 refreshes a day versus versus one like it doesn’t seem like it’d be very much.
205 00:23:44.660 ⇒ 00:23:47.259 Aman Nagpal: Let’s start there. Let’s do 4 refreshes a day.
206 00:23:47.791 ⇒ 00:23:54.620 Aman Nagpal: And kind of see how that goes. But outside of snowflake cost and portable cost, which portable is flat rate.
207 00:23:55.130 ⇒ 00:24:03.359 Aman Nagpal: Will it? Is there anything from a financial or bandwidth, you know.
208 00:24:03.520 ⇒ 00:24:07.589 Aman Nagpal: plate? Do we need to think about? Is there anything else we need to think about or no.
209 00:24:08.550 ⇒ 00:24:11.929 Robert Tseng: No, I mean, it’s just also, I mean, is the increment.
210 00:24:12.270 ⇒ 00:24:21.110 Robert Tseng: Well, we’re we’re already kind of just making this call. But we’re gonna trigger incremental refreshes. So rather than doing a full refresh where you basically like
211 00:24:21.620 ⇒ 00:24:34.100 Robert Tseng: delete and replace every time. You’re just, it’ll scan through existing ones and only add the new any like new values. So that also is like an optimization that we do to like
212 00:24:34.290 ⇒ 00:24:35.880 Robert Tseng: reduce the number of
213 00:24:36.303 ⇒ 00:24:41.969 Robert Tseng: like. So you’re you’re not having to like pay for the same duplication of stuff every every time it refreshes.
214 00:24:41.970 ⇒ 00:24:46.610 Aman Nagpal: That’s on a per source basis, though right? Some sources are all sorts.
215 00:24:46.610 ⇒ 00:24:50.839 Robert Tseng: Some, some do not. Yeah, some some will only allow for refresh. Yeah.
216 00:24:51.190 ⇒ 00:24:52.860 Aman Nagpal: Some will only allow full.
217 00:24:52.860 ⇒ 00:24:53.420 Robert Tseng: Yeah.
218 00:24:53.780 ⇒ 00:24:55.020 Aman Nagpal: Oh, which ones are those.
219 00:24:56.630 ⇒ 00:25:05.710 Robert Tseng: I mean right now. I don’t think we’re doing incremental for any of them. So that’s kind of like we’ll have to. We’ll have to see once we throttle it, to see which ones are are, will support it or not.
220 00:25:05.960 ⇒ 00:25:07.849 Aman Nagpal: So shopify. For example.
221 00:25:07.850 ⇒ 00:25:12.749 Robert Tseng: I’m I mean, shopify must do incremental like I I highly doubt that they would.
222 00:25:12.920 ⇒ 00:25:14.679 Robert Tseng: I would say it’s mainly like.
223 00:25:14.680 ⇒ 00:25:17.030 Aman Nagpal: Catches the changes of old orders.
224 00:25:17.030 ⇒ 00:25:32.250 Robert Tseng: Yeah, I mean, typically, it’s the. It’s the like, the newer tools that don’t really have connectors, or they’re very basic. And only let you do web hooks like they’re less flexible with what you can do with it. But I would say
225 00:25:33.530 ⇒ 00:25:44.219 Robert Tseng: I mean, like Tiktok Shop, I mean, we’re already waiting for that, but, like I don’t know if they would be, they would support it, but shopify must support it. I I had no doubt that would speak that that would be the case.
226 00:25:44.750 ⇒ 00:25:45.749 Aman Nagpal: Got it. So
227 00:25:46.070 ⇒ 00:25:53.239 Aman Nagpal: even shopify is incremental, even though it’s incremental. You know, 6 months old order. If it does a charge back, there’s an update and
228 00:25:53.360 ⇒ 00:25:56.440 Aman Nagpal: portable will catch that and update that information right?
229 00:25:56.440 ⇒ 00:26:15.840 Robert Tseng: Yeah, any any like we kind of we, we would set it up that any field that’s been updated. So assuming that like, obviously, when there’s an update or charge back one of the fields that we’re bringing in has like a value change. It would catch that. And and it would, it would push an incremental change.
230 00:26:16.240 ⇒ 00:26:21.619 Aman Nagpal: Cool that works. And then the other sources where it’s full.
231 00:26:21.760 ⇒ 00:26:23.430 Aman Nagpal: We’re not paying extra.
232 00:26:24.190 ⇒ 00:26:28.749 Aman Nagpal: Are we paying extra anywhere because we’re doing a full refresh versus incremental.
233 00:26:29.780 ⇒ 00:26:47.289 Robert Tseng: No, I mean storage wise. It’s the same. It’s not like we’re increasing like, it’s not like, you’re deleting. Yeah, it’s still the same amount of storage, or like, I mean, you’re paying for the incremental storage. But yeah, I mean you. Your your processing cost is higher. If you’re doing a full refresh. Yeah.
234 00:26:48.370 ⇒ 00:26:53.079 Aman Nagpal: I know you don’t know off top of your head, but like any source that comes to mind, that would be full.
235 00:26:54.030 ⇒ 00:26:57.819 Robert Tseng: I’m like, I said. I’m not, I guess, for other
236 00:26:58.090 ⇒ 00:27:00.801 Robert Tseng: sources that we work with other clients like
237 00:27:02.810 ⇒ 00:27:07.530 Robert Tseng: I don’t know if any tools that, like North Beam, I think, is not
238 00:27:08.000 ⇒ 00:27:12.010 Robert Tseng: incremental. I feel like that’s something we run into with another Ecom client.
239 00:27:12.380 ⇒ 00:27:20.290 Aman Nagpal: That’s a good example. Yeah, okay, we’ll figure that out. I guess if we can have a list of every source, if it’s incremental or full
240 00:27:20.450 ⇒ 00:27:24.790 Aman Nagpal: number of refreshes per day. That would be super helpful. And then we can figure that out from there. Does that work.
241 00:27:24.790 ⇒ 00:27:32.869 Robert Tseng: Yeah. But for now it’s kind of like, try and do 4 for all of them, and we try incremental. And then, whichever ones we don’t like, we’ll we’ll kind of
242 00:27:33.010 ⇒ 00:27:37.460 Robert Tseng: yeah, we’ll. We’ll basically audit it as we’re as we’re turning it on to see which ones don’t allow it.
243 00:27:37.880 ⇒ 00:27:47.620 Aman Nagpal: Let’s do that. And while we’re on the topic that email about recharge so recharge, I don’t know.
244 00:27:48.270 ⇒ 00:27:51.380 Aman Nagpal: Why would we? We would do full versus.
245 00:27:51.780 ⇒ 00:27:56.459 Aman Nagpal: I guess I can see on the the area with all all subscriptions.
246 00:27:56.820 ⇒ 00:28:15.640 Aman Nagpal: Maybe a full makes sense. I don’t know that I feel like that should be incremental events definitely. It’s as they happen like, if an event happened a year ago, whether it’s a subscription change, frequency, change whatever it is that we don’t need a full. We just need whatever events are happening right? So that whole change that they’re doing with the Api and the 90 day thing.
247 00:28:15.880 ⇒ 00:28:16.240 Robert Tseng: Yeah.
248 00:28:16.380 ⇒ 00:28:18.559 Aman Nagpal: I think that only impacts events, right?
249 00:28:19.080 ⇒ 00:28:26.370 Robert Tseng: Yeah, and we should totally move to incremental. I think that was our point. Right? So, like that. With that change like that, I think
250 00:28:27.230 ⇒ 00:28:31.249 Robert Tseng: we yeah, we that that was our. That was my takeaway when I
251 00:28:31.790 ⇒ 00:28:33.900 Robert Tseng: how to always send that message. So.
252 00:28:33.900 ⇒ 00:28:39.110 Aman Nagpal: Is there any reason we did full initially or like once we got the initial data, why keep it full.
253 00:28:41.130 ⇒ 00:28:44.540 Robert Tseng: I think that’s just the default. I don’t. I don’t know if there was any like additional.
254 00:28:44.540 ⇒ 00:28:49.480 Aakash Tandel: Yeah, I was just saying, I think that’s just the standard default way of a lot of these.
255 00:28:49.960 ⇒ 00:28:56.940 Aman Nagpal: So recharge, we changed incremental and any old subscriptions. There’s any changes that happen. It’ll catch it.
256 00:28:57.070 ⇒ 00:29:02.909 Robert Tseng: Yeah, so there are some certain use cases when you would do a full so like.
257 00:29:04.010 ⇒ 00:29:07.669 Robert Tseng: I mean, this is more like for media data. But like
258 00:29:07.860 ⇒ 00:29:16.190 Robert Tseng: like TV ads, like, we look at TV ads company before, like their event stream. A lot of it is. Say, I mean, there’s just like
259 00:29:17.050 ⇒ 00:29:18.290 Robert Tseng: you get
260 00:29:18.450 ⇒ 00:29:29.529 Robert Tseng: a bunch of events in the stream. And then, like, later on, like they, there’s a consolidation, because, like so much happens, it’s more for real time data processing. If
261 00:29:30.020 ⇒ 00:29:31.819 Robert Tseng: yeah, it looks like there’s.
262 00:29:31.860 ⇒ 00:29:42.469 Robert Tseng: let’s say, like, maybe 10 events happened in a single second. On on the 1st time something was triggered because somebody was just like button mashing, or whatever, and then like
263 00:29:42.490 ⇒ 00:30:05.859 Robert Tseng: afterwards. It gets consolidated down into like oh, it was really just one or 2 events in a situation like that. Yeah, instead of doing incremental, you would just do a refresh, because the source itself is already kind of like resetting and doing some consolidation before, like it comes into your warehouse. So like stuff like that, where the data provider is actually just like
264 00:30:05.880 ⇒ 00:30:15.030 Robert Tseng: changing a lot of stuff frequently. I think that’s generally like my understanding of like when we would just leave it as a re as full full refresh.
265 00:30:15.330 ⇒ 00:30:18.649 Aman Nagpal: Let me know if I have this right north beam feels like
266 00:30:18.950 ⇒ 00:30:32.229 Aman Nagpal: the most that’ll get changed is revenue, or spend ad. Spend a couple of days to a week. To Max 2 weeks prior. Everything else pretty much is has already happened and will not change right? So that could be incremental.
267 00:30:32.340 ⇒ 00:30:46.289 Aman Nagpal: But in the situation that we go in and manually add a new type of breakdown. Like we have country, we have landing page, we have whatever breakdowns, we add a new one. At that point we would click and do a full refresh, so that we have the data for that breakdown.
268 00:30:46.570 ⇒ 00:30:48.259 Robert Tseng: Yeah, yeah, that makes sense.
269 00:30:48.580 ⇒ 00:30:53.450 Aman Nagpal: Okay, yeah, let’s move. I mean, we can figure that out. We’ll move everything to incremental. That can be.
270 00:30:54.310 ⇒ 00:31:07.919 Aakash Tandel: Yup, and just like the kind of the the walkthroughs. I’ll do this on a per source basis, so it’ll be so. It’ll be very obvious, like, Oh, North team, we tried. We tried out doing incremental, and we couldn’t because of Xyz or whatever.
271 00:31:08.300 ⇒ 00:31:09.060 Aman Nagpal: Sounds good.
272 00:31:09.580 ⇒ 00:31:10.170 Aakash Tandel: Cool.
273 00:31:12.760 ⇒ 00:31:15.859 Aakash Tandel: Let me take a quick. Okay, so that’s the that’s the.
274 00:31:16.630 ⇒ 00:31:20.799 Aman Nagpal: Speaking of north beam, yeah, where we add with that. Though, that dashboard.
275 00:31:21.000 ⇒ 00:31:22.173 Aakash Tandel: Yeah. So
276 00:31:23.594 ⇒ 00:31:37.070 Aakash Tandel: Kyle and oas are doing those modeling bits right now. So they are gonna be able to. Basically. Well, Annie has was a little blocked on some of the let me go back.
277 00:31:39.280 ⇒ 00:31:40.780 Aakash Tandel: Actually, I think it goes right here.
278 00:31:42.450 ⇒ 00:31:43.200 Aakash Tandel: Yeah.
279 00:31:43.590 ⇒ 00:31:48.214 Aakash Tandel: So this this guy we needed to
280 00:31:49.120 ⇒ 00:31:51.969 Aakash Tandel: get some of these base base pieces of data
281 00:31:52.180 ⇒ 00:32:04.290 Aakash Tandel: modeled or modified and available for metabase. So that’s what we’re waiting on here. This is what a wish. And him and Kyle are basically splitting up these tasks this week. So they’re gonna get this done by the end of the week.
282 00:32:05.870 ⇒ 00:32:11.740 Aman Nagpal: Okay, that works, and that that’ll be the whole dashboard is ready by the 11.th
283 00:32:12.150 ⇒ 00:32:38.930 Aakash Tandel: Yes, unless so, the the caveat there is like, say, and these are probably pretty straightforward. Say, like, you know, orders monthly by customer type, or like something is like, really complicated or really difficult to do. That bit might be left off. And then I’ll put it onto a different ticket for a secondary look. But the bit, all the main piece of data should be available. A lot of stuff that’s kind of out of the box should be directly importable.
284 00:32:40.400 ⇒ 00:32:43.879 Aman Nagpal: Got it. So this stuff the modeling is happening now as well.
285 00:32:44.050 ⇒ 00:32:46.240 Aman Nagpal: We didn’t work on this last week. Then, right.
286 00:32:47.133 ⇒ 00:32:51.860 Aakash Tandel: No, I think they got pulled onto the other work streams last week.
287 00:32:52.364 ⇒ 00:33:01.900 Aakash Tandel: I don’t remember exactly what Kyle did last week. Oh, he was working on that cog stuff. So yeah, I think, that’s the main thing for the beginning of this week is on their plate.
288 00:33:03.080 ⇒ 00:33:04.680 Aakash Tandel: This this northeam stuff.
289 00:33:06.180 ⇒ 00:33:06.870 Aman Nagpal: Got it.
290 00:33:07.940 ⇒ 00:33:08.460 Aakash Tandel: Okay?
291 00:33:10.636 ⇒ 00:33:14.199 Aakash Tandel: But yeah, I think that’s that’s everything on this list.
292 00:33:15.730 ⇒ 00:33:34.609 Aakash Tandel: Yeah, if that’s all I know, we’re at time, too. I don’t know if anyone has to drop free drop. But yeah, we’ll we’ll obviously attentive in North Beam. Those data models to be to be made available for Meta base is kind of a analytics engineering task this week. And then we have some of this incremental stuff
293 00:33:34.720 ⇒ 00:33:38.491 Aakash Tandel: as well to to try to Update
294 00:33:39.360 ⇒ 00:34:02.412 Aakash Tandel: there Annie’s a little late on stuff. So we’re gonna look at the backlog and see if there’s anything that we can work on from the analysis standpoint. Because I think a lot of these are primarily analytics, engineering tasks, and less on her stuff. So if there’s anything you think of like a from a data analysis or like a dashboarding perspective. That’s obviously not like North being more attentive because we’re waiting on those engineering side side of things.
295 00:34:02.690 ⇒ 00:34:10.060 Aakash Tandel: let me know, and I can definitely pull it up. But that was kind of one of the things I wanted to look at today. See if there’s anything she can work on that’s been backlogged.
296 00:34:10.880 ⇒ 00:34:12.770 Aman Nagpal: How about the Klaviyo piece.
297 00:34:13.620 ⇒ 00:34:17.130 Aakash Tandel: Yeah, klaviyo. We need to.
298 00:34:18.035 ⇒ 00:34:24.340 Aakash Tandel: Discovery of what we? Yeah. So this is, I still need to go through and look at what the data
299 00:34:24.850 ⇒ 00:34:34.439 Aakash Tandel: cause. Annie looked at this list, and she said, There’s about 3 days worth of work. So we need to break this into sub tasks. And we need to get these into
300 00:34:34.820 ⇒ 00:34:41.080 Aakash Tandel: basically, we need a wish or Kyle to determine what’s feasible, what’s not feasible. And then
301 00:34:41.888 ⇒ 00:34:46.820 Aakash Tandel: what what the order of operations is, for, like some of these pieces of data.
302 00:34:48.300 ⇒ 00:34:52.079 Aman Nagpal: Okay, got it. So
303 00:34:52.590 ⇒ 00:34:59.309 Aman Nagpal: I guess. And yeah, if you guys need to drop off that, that’s fine. If we have a couple of minutes what
304 00:34:59.460 ⇒ 00:35:05.529 Aman Nagpal: I guess it were the main projects that we worked on last week. And what’s gonna be delivered this week.
305 00:35:06.300 ⇒ 00:35:08.479 Aakash Tandel: Yeah, I can go back to.
306 00:35:09.470 ⇒ 00:35:20.065 Aakash Tandel: So so the main things we did last week was so data ingestion. That was last week for north beam.
307 00:35:20.570 ⇒ 00:35:23.419 Aman Nagpal: Ingestion for Northeam done a couple of weeks back.
308 00:35:27.300 ⇒ 00:35:33.469 Aman Nagpal: I feel like North being we talked about over a month ago the Csv thing and the the weird Api that they had.
309 00:35:34.140 ⇒ 00:35:35.990 Aakash Tandel: Yeah, I don’t know. I think
310 00:35:37.460 ⇒ 00:35:45.380 Aakash Tandel: this I mean, this cloud function stuff happened last week. So that’s part of the North beam ingestion. I don’t know if
311 00:35:45.520 ⇒ 00:35:51.329 Aakash Tandel: I don’t know. Are we getting north beam from a specific connector? Do you know, Robert, is it portable.
312 00:35:52.140 ⇒ 00:36:03.679 Robert Tseng: Yeah, it’s it’s from portable, I mean. But we yeah, I mean, the data has the raw data has been there. We just didn’t model it, and like there was like some scripting that we needed to do to get some like it.
313 00:36:03.870 ⇒ 00:36:09.870 Robert Tseng: like the for some of the segments that were teed up in the amplitude dash like
314 00:36:11.120 ⇒ 00:36:20.253 Robert Tseng: we didn’t have like it came. It came in. I mean, if you look at the raw data, it’s in Snowflake. You can see, like what it was, what it looked like. And
315 00:36:20.750 ⇒ 00:36:28.779 Robert Tseng: I mean whatever we just determined that, like, it needed to have some modeling before we built the reports on it like Annie was not able to use it, as is.
316 00:36:29.640 ⇒ 00:36:34.930 Aman Nagpal: Okay? So outside of north beam, what else did we work on?
317 00:36:36.490 ⇒ 00:36:45.389 Aakash Tandel: Yeah, let’s see. I mean, we we went through some platform fee fields. Consolidation.
318 00:36:45.390 ⇒ 00:36:47.609 Aman Nagpal: That’s 10 or 12 days ago, right?
319 00:36:53.040 ⇒ 00:36:55.113 Aakash Tandel: Yeah, that was done a while back.
320 00:36:55.600 ⇒ 00:37:03.890 Aakash Tandel: is there a way I can sort? By most recent, I think, yeah, I mean, a lot of stuff has been moving. I think it’s just
321 00:37:07.810 ⇒ 00:37:09.779 Aakash Tandel: Yeah, I don’t know. Let me see.
322 00:37:10.170 ⇒ 00:37:21.319 Aakash Tandel: I know I worked on getting you access to things. The light dash setup was created. Kyle’s been working on getting the ingestion from
323 00:37:21.530 ⇒ 00:37:33.260 Aakash Tandel: that spreadsheet updated. Annie worked on subscribe and save information. We went through the Meta base training. I think those are the main things from last week
324 00:37:35.650 ⇒ 00:37:39.409 Aakash Tandel: I know. Away sent you an example of
325 00:37:39.850 ⇒ 00:37:45.760 Aakash Tandel: the Clayview and attentive data which we just need to model. Now.
326 00:37:47.310 ⇒ 00:37:52.300 Aakash Tandel: I think those are the main things. And then, oh, my migrating! This was finished.
327 00:37:52.640 ⇒ 00:37:54.930 Aakash Tandel: Think fully finished last week. Yeah.
328 00:37:55.480 ⇒ 00:37:59.239 Aakash Tandel: So everything should be migrated to metabase. I think those are the main things.
329 00:38:00.050 ⇒ 00:38:09.180 Aman Nagpal: Cool, so what can we expect? Deliverability? Deliverables wise this week? Say one, you said. By the 11th we’ll have north beam right? What else.
330 00:38:09.180 ⇒ 00:38:18.020 Aakash Tandel: Yep, yeah. So it’s right. Now we’re looking at north beam and attentive, which I can move this here
331 00:38:20.090 ⇒ 00:38:24.869 Aakash Tandel: north being attentive. And then we can do. We’ll try to get a waste to
332 00:38:25.140 ⇒ 00:38:37.409 Aakash Tandel: work on these looms, I think. I’ll leave it up to you to determine, like what order of operations you want or what priority you want the training versus some of these engineering pieces, because,
333 00:38:38.230 ⇒ 00:38:51.540 Aakash Tandel: currently, this is not on his plate, because I haven’t written tickets for that yet, but we can do that if you want. The main things. He’s working on our north beam and attentive this week. And then also, we need to set up
334 00:38:51.720 ⇒ 00:38:55.540 Aakash Tandel: time for them to do the incremental refreshes. So those are.
335 00:38:55.700 ⇒ 00:38:58.880 Aakash Tandel: I guess, a prioritization exercise, because I don’t think
336 00:38:59.330 ⇒ 00:39:03.790 Aakash Tandel: they’ll be able to do all of that within the week. But yeah.
337 00:39:04.360 ⇒ 00:39:08.699 Aman Nagpal: Yeah, I mean, ideally, we don’t have too much time left, and I feel like we’ve been stuck on
338 00:39:09.160 ⇒ 00:39:14.829 Aman Nagpal: something like, you know, like I said, North Beam, we’ve been talking about for over a month now. So whatever we can get done
339 00:39:15.620 ⇒ 00:39:21.119 Aman Nagpal: that would be great. And then, yeah, same thing on the training side, right as we need that handoff.
340 00:39:21.370 ⇒ 00:39:26.789 Aman Nagpal: So again, same thing, documentation, side, zoom side, whatever we can do there.
341 00:39:27.870 ⇒ 00:39:33.530 Aman Nagpal: But I’ll expect at least north beam, attentive klaviyo, and then some
342 00:39:34.260 ⇒ 00:39:36.840 Aman Nagpal: training stuff this week or documentation this week.
343 00:39:37.450 ⇒ 00:39:38.319 Aakash Tandel: Yeah,
344 00:39:39.920 ⇒ 00:39:48.160 Aakash Tandel: I need to figure out. So for Klavio, you only are. We only gonna be wanting email sent out revenue and orders placed within the last 5 days.
345 00:39:50.410 ⇒ 00:39:55.830 Aman Nagpal: That was a starter. And then I think we said, whatever else you guys can come up with, that you think would be helpful.
346 00:39:56.600 ⇒ 00:40:00.770 Aakash Tandel: Yeah, so we can, we can probably start with that. I think.
347 00:40:00.970 ⇒ 00:40:05.590 Aakash Tandel: whatever else is gonna be a lot like that’s gonna take us.
348 00:40:06.140 ⇒ 00:40:17.970 Aakash Tandel: we basically need to send that to Annie to see like what would be helpful for like visualization standpoint and data analysis standpoint. And then it has to go back to engineering to model. So there’s a lot of you know.
349 00:40:19.470 ⇒ 00:40:24.029 Aakash Tandel: The thing is these get these things, give you a lot of data. So we basically have to be able to
350 00:40:26.210 ⇒ 00:40:46.009 Aakash Tandel: here. So so basically, here’s the business context understanding how email and SMS sales convert conversions to optimize performance. There’s a lot of questions here, and I’m not sure who created these questions. But like to answer all these, and to develop all the modeling behind this would not be. That’s not a small effort. That’s a lot of
351 00:40:46.120 ⇒ 00:40:46.650 Aakash Tandel: that’s a lot.
352 00:40:46.650 ⇒ 00:40:47.980 Aman Nagpal: This. Yeah, this is a lot.
353 00:40:48.440 ⇒ 00:40:59.009 Aman Nagpal: I mean, I don’t know who made this list. But yeah, these are some ideas outside of the one on the sheet. Whatever of these you think are helpful. Just add whatever you can.
354 00:40:59.250 ⇒ 00:41:01.840 Aman Nagpal: Obviously this is a ton of stuff. So.
355 00:41:01.840 ⇒ 00:41:02.520 Aakash Tandel: Yeah.
356 00:41:03.400 ⇒ 00:41:12.529 Aakash Tandel: yeah, we’ll start with these. And then, maybe. Do you have a specific person who like would be analyzing this data or working with the Klavio data.
357 00:41:14.640 ⇒ 00:41:18.400 Aman Nagpal: There’s the email team and me. But also we don’t want to do things like
358 00:41:18.590 ⇒ 00:41:21.599 Aman Nagpal: open right, for example, that we can see clearly on
359 00:41:22.270 ⇒ 00:41:31.380 Aman Nagpal: Klaviyo right now, like, if it’s attached to another source like actual revenue, etc, that’s helpful, but just showing the exact same won’t be helpful for Klaviyo rettentive
360 00:41:33.030 ⇒ 00:41:34.560 Aman Nagpal: that list you just showed me
361 00:41:34.910 ⇒ 00:41:39.219 Aman Nagpal: those items probably wouldn’t be helpful. Only certain ones from the list.
362 00:41:40.230 ⇒ 00:41:52.780 Aakash Tandel: Okay, yeah. Okay. Well, let’s let me see what what the attentive ticket is.
363 00:41:53.160 ⇒ 00:41:56.329 Aman Nagpal: But they, yeah, that’s I mean, the priority is just the one we have written down.
364 00:41:56.900 ⇒ 00:41:57.570 Aman Nagpal: Yeah.
365 00:41:58.580 ⇒ 00:42:25.440 Aakash Tandel: Okay, yeah, let’s start with those, and then we’ll see what we can. I mean, the the. So the main thing is joining the data across those other things. Right? That’s kind of the idea. We’ll see. If that’s like readily straightforward. Then they could, you know, combine on orders and revenue. And so like that fairly readily. So basically, the data you have currently tied to orders, you’d get that over with emails. That’d be kind of the logic there. Yeah, I can set up.
366 00:42:25.520 ⇒ 00:42:32.379 Aakash Tandel: I can set up tickets for these these 1st pieces and then see what data we get out of that. And then.
367 00:42:32.890 ⇒ 00:42:39.980 Aakash Tandel: yeah, I don’t know what the other questions are, but we can definitely meet with you or the email team. If if there’s follow up questions. After that.
368 00:42:40.930 ⇒ 00:42:41.780 Aman Nagpal: Sounds good.
369 00:42:42.367 ⇒ 00:42:43.590 Aman Nagpal: Also on the
370 00:42:44.190 ⇒ 00:42:50.949 Aman Nagpal: documentation loom piece. If you know, you can kind of explain Api like recharge, you know it does
371 00:42:51.360 ⇒ 00:43:04.600 Aman Nagpal: portable or whatever we’re using, does it just run an Api request for every single page and grab all that data that way? It just runs a ton of requests over and over. Or is there a different way to go about Apis like that?
372 00:43:05.392 ⇒ 00:43:07.750 Aman Nagpal: Yeah. I mean, whatever info we can get there.
373 00:43:09.130 ⇒ 00:43:33.609 Aakash Tandel: Okay, yeah, that’s that’s the type of question. I definitely don’t know the answer to and it’s something that I think. If we have like a list of questions like that that you want to answer, I think that would be helpful, because I feel like she’s probably going to give you an overview of a lot of these things, and might not think about something like that. So if that if you have questions like that definitely like, compile them in the list, and I can get the team to answer this.
374 00:43:33.940 ⇒ 00:43:35.609 Aman Nagpal: Well, and then the light dash
375 00:43:36.501 ⇒ 00:43:59.609 Aman Nagpal: automatic us to a thread with light dash they created an account, I was admin, I added him, and then I guess he mistakenly made another light dash account with my email with a plus in it. I don’t think I can get into that. So if you guys can contact the light dash team, I don’t know if you’re in the the same group with them. Maybe with them can add you. But just tell them, look, we created this account.
376 00:43:59.720 ⇒ 00:44:05.379 Aman Nagpal: then, with the made this account, can you make this one the main one, and make my email the owner of that one.
377 00:44:06.640 ⇒ 00:44:09.090 Aman Nagpal: I think that’s kind of where we where we’re at with that.
378 00:44:09.380 ⇒ 00:44:12.089 Aakash Tandel: Okay, yeah. Let me
379 00:44:20.770 ⇒ 00:44:23.379 Aakash Tandel: added to the main account.
380 00:44:23.380 ⇒ 00:44:28.030 Aman Nagpal: And sorry. Can you do a double V for this one? We’re kind of making that change. Yeah.
381 00:44:28.420 ⇒ 00:44:29.060 Aakash Tandel: Okay.
382 00:44:32.720 ⇒ 00:44:39.130 Aakash Tandel: Okay? Yeah, I will message the Meta base team to to do that and go from there. That sounds good.
383 00:44:39.130 ⇒ 00:44:40.120 Aman Nagpal: By dash, right.
384 00:44:40.720 ⇒ 00:44:42.770 Aakash Tandel: That’s what I meant. Yep, like.
385 00:44:43.080 ⇒ 00:44:48.949 Aman Nagpal: Alright, I think that’s it. Thank you. Guys, and yeah, just just keep me posted like I said, we have
386 00:44:49.160 ⇒ 00:44:54.240 Aman Nagpal: not too much time left. We we need to figure out what we’re gonna do when it comes to.
387 00:44:54.410 ⇒ 00:44:57.390 Aman Nagpal: you know, sign a new contract. But we just really need to get
388 00:44:57.500 ⇒ 00:45:02.180 Aman Nagpal: as much done as possible, so I can go to the team and say, Hey, look! This is what we accomplished.
389 00:45:02.522 ⇒ 00:45:04.950 Aman Nagpal: You know. What else can we do in the future.
390 00:45:05.200 ⇒ 00:45:05.800 Aakash Tandel: Yep.
391 00:45:06.020 ⇒ 00:45:15.210 Aakash Tandel: that makes sense. That sounds good. Yeah. Well, also, if you want to join standups you’re welcome to. But you’ll see these tickets also hopefully moving across board.
392 00:45:15.760 ⇒ 00:45:17.639 Aman Nagpal: Sounds good. Thank you. Guys.
393 00:45:18.160 ⇒ 00:45:19.509 Aakash Tandel: Thanks. Have a good day.
394 00:45:19.510 ⇒ 00:45:20.400 Aman Nagpal: You too, bye.