Meeting Title: Brainforge x LMNT: Omni Demo Date: 2026-01-08 Meeting participants: Awaish Kumar, Shivani Amar, Demilade Agboola, Jason Wu, Uttam Kumaran
WEBVTT
1 00:03:12.210 ⇒ 00:03:13.410 Demilade Agboola: Hi, Shivani.
2 00:03:13.750 ⇒ 00:03:14.580 Shivani Amar: Hi!
3 00:03:16.000 ⇒ 00:03:18.709 Demilade Agboola: Do you know how long Jason will take to get here?
4 00:03:19.840 ⇒ 00:03:22.509 Shivani Amar: Let’s see…
5 00:03:23.330 ⇒ 00:03:37.470 Shivani Amar: I actually don’t know, but I think we can get started and then record the demo, and so then if he comes late, or if anybody else comes late… Sorry, if he comes late, or if anybody else wants to see the demo, we can… we can link it.
6 00:03:40.270 ⇒ 00:03:41.449 Shivani Amar: Does that work?
7 00:03:43.270 ⇒ 00:03:44.970 Demilade Agboola: Yes, it does, so we can start…
8 00:03:44.970 ⇒ 00:03:45.570 Shivani Amar: Okay.
9 00:03:46.360 ⇒ 00:03:47.320 Demilade Agboola: Okay.
10 00:03:48.030 ⇒ 00:03:55.890 Shivani Amar: Okay, them was saying maybe he was gonna, like, do a Loom or some recording? Like, is that… or do you want to just record the Zoom? How do you want to go about it?
11 00:03:56.200 ⇒ 00:03:59.789 Awaish Kumar: Yeah, we are already recording this meeting, so we can share.
12 00:04:00.990 ⇒ 00:04:08.549 Shivani Amar: Can you cut this whole part about me saying, let’s record the meeting for anybody else who wants to watch the studio? Which…
13 00:04:09.030 ⇒ 00:04:10.230 Shivani Amar: Should we restart?
14 00:04:11.290 ⇒ 00:04:21.980 Demilade Agboola: Yeah, so usually our Zoom meetings auto-record. It helps us to be able to go back and help see context on different calls and how we can assess different things.
15 00:04:22.510 ⇒ 00:04:23.290 Shivani Amar: Perfect.
16 00:04:23.630 ⇒ 00:04:27.820 Demilade Agboola: Alright, alright, so let me share my screen…
17 00:04:35.310 ⇒ 00:04:36.700 Demilade Agboola: Okay, so can you see my screen?
18 00:04:37.460 ⇒ 00:04:38.140 Shivani Amar: Yep.
19 00:04:40.110 ⇒ 00:04:43.450 Demilade Agboola: So Omni is basically a…
20 00:04:43.880 ⇒ 00:04:47.669 Demilade Agboola: Data exploration, as well as a data visualization tool.
21 00:04:50.140 ⇒ 00:04:58.179 Demilade Agboola: And this is our demo account with, like, demo data that we can use to play around and kind of show the concept of Omni.
22 00:04:58.440 ⇒ 00:05:02.970 Demilade Agboola: So, when you get in here, you kind of… you would have your homepage.
23 00:05:03.500 ⇒ 00:05:09.429 Demilade Agboola: And the concept of this is you would have connected it to your data warehouse, where your data will leave.
24 00:05:09.660 ⇒ 00:05:15.190 Demilade Agboola: And you’re now using the data from there to be able to, create new
25 00:05:15.560 ⇒ 00:05:19.689 Demilade Agboola: Topics, so topics in… Hi, Jason.
26 00:05:20.660 ⇒ 00:05:23.049 Jason Wu: Hey there, sorry I was a little bit late here.
27 00:05:23.350 ⇒ 00:05:26.010 Demilade Agboola: Oh, no, it’s good, we just got started, so you haven’t missed much.
28 00:05:27.510 ⇒ 00:05:31.009 Demilade Agboola: So, topics in Omni are…
29 00:05:31.210 ⇒ 00:05:35.289 Demilade Agboola: Curated data sets, so we can start to…
30 00:05:35.680 ⇒ 00:05:42.520 Demilade Agboola: have different topics by business concept. So we can have a finance topic, we can have a marketing topic.
31 00:05:42.690 ⇒ 00:05:47.979 Demilade Agboola: We can have an inventory topic, depending on, like, what pertains to your business, and…
32 00:05:48.170 ⇒ 00:05:55.659 Demilade Agboola: What group of users will be utilizing different topics to find out different answers for their business needs.
33 00:05:55.890 ⇒ 00:06:01.930 Demilade Agboola: So, that’s kind of, like, a high-level explanation of Omni and the topics within Omni.
34 00:06:02.240 ⇒ 00:06:08.140 Demilade Agboola: And so, if you were to come into Omni, you will see a couple of things here on the left.
35 00:06:08.300 ⇒ 00:06:16.929 Demilade Agboola: So, you have your homepage, which is where we are. You have the AI Assistant, so the AI Assistant allows you to be able to chat to different topics.
36 00:06:17.490 ⇒ 00:06:23.990 Demilade Agboola: And I could make… Inquiries across different things in the data available to it.
37 00:06:24.710 ⇒ 00:06:34.589 Demilade Agboola: You have things like your My Documents. So, favorites are where you will select favorite concepts, or like dashboards or explorations that you have in progress.
38 00:06:35.070 ⇒ 00:06:44.789 Demilade Agboola: My documents is where you are currently working on, like, dashboards and where it’s saved, so it’s only visible to you, and other people cannot see unless you explicitly share with them.
39 00:06:45.290 ⇒ 00:06:50.290 Demilade Agboola: Hub is where everyone is… puts in their,
40 00:06:51.950 ⇒ 00:06:55.549 Demilade Agboola: Their dashboards and business exploration in one spot.
41 00:06:56.120 ⇒ 00:07:03.799 Demilade Agboola: And shared with me would be things like my documents that were shared directly with you from another person, so that allows you to visualize it.
42 00:07:04.320 ⇒ 00:07:06.809 Demilade Agboola: So if we go back to the home.
43 00:07:07.420 ⇒ 00:07:10.240 Demilade Agboola: And we want to start playing around with stuff.
44 00:07:11.360 ⇒ 00:07:13.330 Demilade Agboola: We will start with clicking on New.
45 00:07:13.770 ⇒ 00:07:25.599 Demilade Agboola: And the idea is you would have only one, ideally one, warehouse where there’s data living. We have multiple, because this is our instance, and we play around with different things.
46 00:07:25.710 ⇒ 00:07:28.790 Demilade Agboola: So here, if we go into the econ demo.
47 00:07:29.190 ⇒ 00:07:35.320 Demilade Agboola: We have topics that have already been created, so we have things like the marketing data, the orders fulfilled.
48 00:07:35.600 ⇒ 00:07:47.460 Demilade Agboola: And if you click in, you have, like, other subtopics beneath. So if we want to explore, like, order transactions of this demo data, and kind of see what’s happening with orders in here.
49 00:07:47.930 ⇒ 00:07:49.289 Demilade Agboola: So right off the bat.
50 00:07:49.840 ⇒ 00:07:58.080 Demilade Agboola: we can see that we have different things in here. We have the inventory items data, we have the order items data.
51 00:07:59.390 ⇒ 00:08:04.439 Demilade Agboola: We have the product’s data, and we have the user’s data.
52 00:08:04.720 ⇒ 00:08:08.490 Demilade Agboola: So behind the scenes, we have a relationship,
53 00:08:09.520 ⇒ 00:08:20.359 Demilade Agboola: workbook, where we’re tying things together. So that’s part of the modeling here. But that’s a bit, like, on the technical side, where we’re putting, like, the join keys and how things are related to each other.
54 00:08:20.760 ⇒ 00:08:29.859 Demilade Agboola: But on the high level, what you have is this, where you can come in here and start to make inquiries into your data, for instance. So let’s look at our dummy data.
55 00:08:30.220 ⇒ 00:08:37.010 Demilade Agboola: If we wanted to know, for instance, what are our highest selling, say, brand categories.
56 00:08:37.190 ⇒ 00:08:41.500 Demilade Agboola: Or highest, yeah, category. So, we will click on categories.
57 00:08:41.740 ⇒ 00:08:49.000 Demilade Agboola: It gives us the different categories, so there are 26 different categories within our Demo data.
58 00:08:49.590 ⇒ 00:08:52.630 Demilade Agboola: We can then go and say, hey, let’s try and see the…
59 00:08:54.110 ⇒ 00:08:58.400 Demilade Agboola: sales, right? The sale… the total sale price sum.
60 00:08:58.650 ⇒ 00:09:08.860 Demilade Agboola: Of all of this. So now we can kind of see it’s working, and we can see that the highest selling right now, because you can see it’s in descending order, we have jeans, accessories.
61 00:09:09.030 ⇒ 00:09:12.899 Demilade Agboola: Outwear and coats and all of that, and we can see socks.
62 00:09:13.120 ⇒ 00:09:17.980 Demilade Agboola: are the least selling, and maternity… maternity and socks are the least selling things.
63 00:09:18.690 ⇒ 00:09:21.759 Demilade Agboola: But again, we might want to go a bit further. What…
64 00:09:23.140 ⇒ 00:09:31.569 Demilade Agboola: This might feel like too much, like, we moniker for the top 10 things, and every other thing outside the top 10, it just… it’s just bloatware.
65 00:09:31.860 ⇒ 00:09:36.209 Demilade Agboola: So here, we can start to set limits, and we can say, hey, I only care for the 10.
66 00:09:37.430 ⇒ 00:09:43.580 Demilade Agboola: And so now we only care… the top 10 best sellers are right here, in terms of categories.
67 00:09:44.080 ⇒ 00:09:55.600 Demilade Agboola: We can then say, hey, but I might want to, like, look at it by different time periods, because what might be best-selling all time might not be best-selling over the last 3 months, for instance.
68 00:09:55.750 ⇒ 00:09:58.590 Demilade Agboola: So, we can come back in here…
69 00:09:59.800 ⇒ 00:10:13.440 Demilade Agboola: click on, like, the created ad date, and now you can see that Omni automatically starts to clean up the data. So it gets a created ad date, but it starts to extract different portions. So you can view it by the date.
70 00:10:13.550 ⇒ 00:10:16.980 Demilade Agboola: By the week, by the month, by the quarter, by the year.
71 00:10:17.250 ⇒ 00:10:23.519 Demilade Agboola: So, we can say, hey, I only care about quarters, for instance, so let’s add this as a time filter.
72 00:10:25.370 ⇒ 00:10:33.060 Demilade Agboola: And… We can see here… Let’s look at quarters… And in the pasts.
73 00:10:33.530 ⇒ 00:10:34.740 Demilade Agboola: One quarter.
74 00:10:35.610 ⇒ 00:10:39.540 Demilade Agboola: Alright, so this is one criteria, so… Right now, the best…
75 00:10:39.830 ⇒ 00:10:45.500 Demilade Agboola: Sellers, top 10 sellers, are now… Listed here.
76 00:10:45.760 ⇒ 00:10:53.400 Demilade Agboola: We can also decide to change this, we can say between, so, like, between different time dates, on or before, and all of that.
77 00:10:54.380 ⇒ 00:10:58.910 Demilade Agboola: So now, we can look at quarters, we can say, hey, over the last quarter, so…
78 00:10:59.950 ⇒ 00:11:05.399 Demilade Agboola: complete quarters, actually. Over the last quarter, These are the best sellers.
79 00:11:06.370 ⇒ 00:11:12.750 Demilade Agboola: And… Depending on your interest, you can decide to view it as a chart instead.
80 00:11:13.900 ⇒ 00:11:31.440 Shivani Amar: Question for you. Created at, specifically, that title is throwing me off a little bit, versus just, like, yeah, I want to see revenue, quarterly buy product for my top 10 products, or something like that. Like, if I were to say I want to see
81 00:11:31.580 ⇒ 00:11:37.969 Shivani Amar: How jeans, accessories, and outerwear have performed quarterly over the last 10 quarters.
82 00:11:37.970 ⇒ 00:11:44.550 Demilade Agboola: Okay, yeah, in that case, what we’ll need to do is we’ll add… the, quarter…
83 00:11:44.920 ⇒ 00:11:46.489 Shivani Amar: As a domestic in Congress?
84 00:11:46.490 ⇒ 00:11:47.669 Demilade Agboola: dimension, yeah.
85 00:11:47.670 ⇒ 00:11:48.000 Shivani Amar: In here.
86 00:11:48.640 ⇒ 00:11:53.570 Demilade Agboola: So, in that case, we will… Click on quarter…
87 00:11:55.240 ⇒ 00:11:58.049 Demilade Agboola: Alright, so now we have the different quarters.
88 00:11:58.220 ⇒ 00:11:59.310 Demilade Agboola: as well.
89 00:11:59.520 ⇒ 00:12:04.720 Demilade Agboola: But this filter… We’ll need to go, because we don’t necessarily want to see it.
90 00:12:07.140 ⇒ 00:12:12.510 Demilade Agboola: And now… so we have the different quarters, and we can see how well things are going.
91 00:12:13.230 ⇒ 00:12:17.469 Demilade Agboola: So yeah, we can create a chart of this.
92 00:12:18.110 ⇒ 00:12:19.400 Demilade Agboola: And in this case…
93 00:12:19.400 ⇒ 00:12:28.919 Shivani Amar: still just, like, one quarter of data. Like, if you were to say, I want to see how things have evolved over the last X quarters, that’s what I’m trying to play with for… just to.
94 00:12:28.920 ⇒ 00:12:29.240 Demilade Agboola: Perfect.
95 00:12:29.240 ⇒ 00:12:31.160 Shivani Amar: How do you do it in the system.
96 00:12:31.960 ⇒ 00:12:39.389 Demilade Agboola: Okay, so in that case, what we’ll need to do will be… So… the date…
97 00:12:43.120 ⇒ 00:12:51.300 Demilade Agboola: And then, if we come here, we can start to see… What a chart.
98 00:12:55.110 ⇒ 00:12:55.880 Demilade Agboola: Oof.
99 00:12:56.440 ⇒ 00:12:59.409 Shivani Amar: Does this sample dataset have more than one quarter of data?
100 00:13:00.120 ⇒ 00:13:01.160 Demilade Agboola: Yes it is.
101 00:13:01.700 ⇒ 00:13:02.360 Shivani Amar: Okay.
102 00:13:03.930 ⇒ 00:13:07.779 Awaish Kumar: I think it’s maybe because we’ve set the limit on 10 rows.
103 00:13:08.850 ⇒ 00:13:17.929 Shivani Amar: Yeah, I’m not sure… I’m not sure what happened. Sorry, I broke it a little bit, but I thought it would be helpful to just see, like, a time series. Like, a lot of, like.
104 00:13:18.430 ⇒ 00:13:28.650 Shivani Amar: what, like, a lot of folks at Element are kind of like, oh, I have to pull data for the last X time of, you know, the last month, and I want to be able to see things over a longer period of time.
105 00:13:29.150 ⇒ 00:13:30.210 Shivani Amar: And so…
106 00:13:31.260 ⇒ 00:13:35.859 Demilade Agboola: It’s just a function of rearranging
107 00:13:36.950 ⇒ 00:13:39.860 Demilade Agboola: This, so it’s not my date…
108 00:13:47.110 ⇒ 00:13:51.640 Demilade Agboola: And then… Oh.
109 00:13:52.900 ⇒ 00:13:53.830 Demilade Agboola: Interesting.
110 00:13:54.050 ⇒ 00:13:55.470 Demilade Agboola: Gardens.
111 00:13:55.940 ⇒ 00:14:06.340 Shivani Amar: Because it only says 2023 Q1, so that’s why, like, in your data right now, we don’t even have multiple quarters, I’m just trying to figure out how you get the multiple quarters worth of data.
112 00:14:06.950 ⇒ 00:14:09.020 Demilade Agboola: Oh yeah, so because we broke it up by…
113 00:14:09.360 ⇒ 00:14:13.599 Demilade Agboola: By this, so we’ll have to not do the limit.
114 00:14:14.140 ⇒ 00:14:16.929 Demilade Agboola: In this case, we would have all the data.
115 00:14:18.210 ⇒ 00:14:19.750 Demilade Agboola: As best as possible.
116 00:14:19.890 ⇒ 00:14:22.670 Demilade Agboola: And so that will give us all the data over time.
117 00:14:22.670 ⇒ 00:14:26.020 Shivani Amar: But still, it’s just showing, like, Q1, like.
118 00:14:27.120 ⇒ 00:14:30.470 Demilade Agboola: Yeah, because it’s ordered by the… sales.
119 00:14:31.120 ⇒ 00:14:32.290 Demilade Agboola: It’s ordered by the…
120 00:14:32.290 ⇒ 00:14:37.660 Uttam Kumaran: Maybe the data, like, the sample data doesn’t go beyond… Q1? Not sure.
121 00:14:37.780 ⇒ 00:14:38.530 Uttam Kumaran: Oh, hey there.
122 00:14:38.530 ⇒ 00:14:40.749 Shivani Amar: Now, there we go. Okay, okay, okay.
123 00:14:40.750 ⇒ 00:14:41.470 Demilade Agboola: No, it just…
124 00:14:41.470 ⇒ 00:14:41.910 Awaish Kumar: Yeah.
125 00:14:41.910 ⇒ 00:14:43.170 Demilade Agboola: order, basically.
126 00:14:43.460 ⇒ 00:14:44.130 Demilade Agboola: Okay.
127 00:14:45.320 ⇒ 00:14:50.870 Awaish Kumar: I think we are increasing those by adding the date. If you can remove, we will just get the quarters.
128 00:14:51.740 ⇒ 00:14:52.610 Demilade Agboola: Yes.
129 00:14:53.530 ⇒ 00:14:54.930 Demilade Agboola: Let’s do that.
130 00:14:57.550 ⇒ 00:15:06.330 Demilade Agboola: And then… In this case, we just need the… Continuing…
131 00:15:10.910 ⇒ 00:15:15.279 Demilade Agboola: Alright, so here we can kind of see the different categories by…
132 00:15:15.920 ⇒ 00:15:19.540 Demilade Agboola: The quarters, so we can see here…
133 00:15:19.700 ⇒ 00:15:23.519 Demilade Agboola: Jeans had, like, their best quarter in Q4.
134 00:15:23.730 ⇒ 00:15:24.360 Demilade Agboola: Of…
135 00:15:24.360 ⇒ 00:15:33.439 Shivani Amar: Yeah, and it, like, all drops off because we’re not done with this quarter, so that’s a little bit distracting, which is totally fine, that makes sense, but I guess, so it’s just, like, jeans consistently do well.
136 00:15:33.710 ⇒ 00:15:38.860 Shivani Amar: Yeah. And had their best quarter in Q4, that’s what this would tell us, and…
137 00:15:39.170 ⇒ 00:15:48.550 Shivani Amar: everything’s kind of moving up and to the right. Like, a lot of things are moving to the up and the right, except maybe some things down at the bottom are, like, flat, and haven’t really…
138 00:15:48.880 ⇒ 00:15:51.890 Shivani Amar: grown, so it gives you a feel for it. Okay.
139 00:15:53.060 ⇒ 00:15:58.289 Demilade Agboola: Yeah, so for the… this quarter being, like, weird, we can just add this. Perfect. Perfect, yeah.
140 00:15:58.630 ⇒ 00:16:00.059 Demilade Agboola: We have our chart.
141 00:16:01.120 ⇒ 00:16:08.160 Demilade Agboola: And so, in this case, we will be able to now, like Nima or Chatzi, category…
142 00:16:08.480 ⇒ 00:16:11.820 Demilade Agboola: Sales by quarter.
143 00:16:15.780 ⇒ 00:16:17.220 Demilade Agboola: And then we can save that.
144 00:16:18.180 ⇒ 00:16:21.350 Demilade Agboola: And then… so now we have this one chart.
145 00:16:21.480 ⇒ 00:16:27.730 Demilade Agboola: We can then decide, hey, we need We need more.
146 00:16:29.610 ⇒ 00:16:38.539 Demilade Agboola: Alright, so we have this. So we come back to our topics again, and we can click on the same topic in this case, and we might decide, hey, we also want to view
147 00:16:38.960 ⇒ 00:16:42.479 Demilade Agboola: this by brand. So the same thing we did, but by brand.
148 00:16:44.290 ⇒ 00:16:53.110 Demilade Agboola: And so, we could use AI to help us do that, because that’s part of what Omni allows us to do. So you can say, show me…
149 00:16:53.980 ⇒ 00:16:55.070 Demilade Agboola: the…
150 00:16:58.140 ⇒ 00:17:06.400 Demilade Agboola: quarterly… sales… Brent… the qu…
151 00:17:09.900 ⇒ 00:17:11.669 Demilade Agboola: Sales, sun price.
152 00:17:15.490 ⇒ 00:17:26.730 Demilade Agboola: Before… 20, 26… So… Omni starts to generate the query behind the scenes.
153 00:17:27.089 ⇒ 00:17:27.929 Shivani Amar: Hmm.
154 00:17:27.930 ⇒ 00:17:30.749 Demilade Agboola: And then it would use that to create the…
155 00:17:32.520 ⇒ 00:17:33.210 Shivani Amar: Sure.
156 00:17:34.160 ⇒ 00:17:35.520 Demilade Agboola: And so now…
157 00:17:36.660 ⇒ 00:17:43.669 Demilade Agboola: Now, there are way more brands than, categories, as you would expect, so this is a bit noisy.
158 00:17:44.260 ⇒ 00:17:52.889 Demilade Agboola: So what you can decide to do is, potentially… Like, either remove the…
159 00:17:53.220 ⇒ 00:17:58.290 Demilade Agboola: Quarter, because again, we don’t necessarily care for the quarters.
160 00:17:58.530 ⇒ 00:18:09.179 Demilade Agboola: we can start to remove the quarters, so now it’ll just be of all time, and then we can view by quarter, if we want to see every single thing. So, what I mean by that is we can remove this.
161 00:18:11.680 ⇒ 00:18:13.830 Demilade Agboola: And then we can add a filter.
162 00:18:13.950 ⇒ 00:18:16.630 Demilade Agboola: for each quarter.
163 00:18:31.230 ⇒ 00:18:40.610 Demilade Agboola: Quarters… And so now, when we update this, this allows us to see… The last quarter finds us.
164 00:18:41.380 ⇒ 00:18:46.950 Demilade Agboola: Still quite noisy, but at this point, what we can now do is we can limit it to the 10.
165 00:18:47.620 ⇒ 00:18:55.310 Demilade Agboola: And so now, for every quarter that passes, or for every quarter that we care to see, we can cycle through and kind of see
166 00:18:55.440 ⇒ 00:19:00.340 Demilade Agboola: The top 10, best sellers within each quarter.
167 00:19:00.660 ⇒ 00:19:02.219 Demilade Agboola: In terms of the brand.
168 00:19:02.790 ⇒ 00:19:12.750 Demilade Agboola: So in this case, we want it to be… descending… And…
169 00:19:14.970 ⇒ 00:19:24.310 Demilade Agboola: sideways, so we can easily see. So, Levi’s were the best in the past quarter, Ray-Ban, Columbia, and we can continue just kind of seeing that.
170 00:19:24.500 ⇒ 00:19:30.050 Demilade Agboola: And so now, we can rename this… It’s… Bests.
171 00:19:31.780 ⇒ 00:19:37.200 Demilade Agboola: Selling… brand… per quarter.
172 00:19:40.520 ⇒ 00:19:43.460 Demilade Agboola: And so now, if we click on Dashboard…
173 00:19:44.760 ⇒ 00:19:51.770 Demilade Agboola: We would have to come up with the name, so we can say Test Sales… Dashboard?
174 00:19:53.140 ⇒ 00:20:01.579 Demilade Agboola: And so now, you have two options. You can either save this in, like, my documents, which I said… your personal documents, which can be you just exploring things for yourself.
175 00:20:01.690 ⇒ 00:20:09.400 Demilade Agboola: Or you can do hub, so hub is where you’re putting for every other person to see, and then they have access to see the information you are getting.
176 00:20:09.960 ⇒ 00:20:12.359 Demilade Agboola: So, I’m saving this in Hub right now.
177 00:20:13.190 ⇒ 00:20:19.629 Demilade Agboola: And now we can kind of start to get information about, you know, the category sales per quarter.
178 00:20:19.810 ⇒ 00:20:23.729 Demilade Agboola: As well as the brand per quarter.
179 00:20:23.900 ⇒ 00:20:29.990 Demilade Agboola: Now, other things we can do is we can start to add filters, where…
180 00:20:32.730 ⇒ 00:20:35.080 Demilade Agboola: We can add, like, the time.
181 00:20:38.820 ⇒ 00:20:45.259 Demilade Agboola: And, for instance, we can use… Like… a date.
182 00:20:48.840 ⇒ 00:20:54.319 Demilade Agboola: So we can start to say, hey, Okay, I want to see…
183 00:20:54.610 ⇒ 00:20:57.480 Demilade Agboola: For, like, last year, what that looked like.
184 00:21:05.460 ⇒ 00:21:08.669 Demilade Agboola: And so, we can kind of see the flow for last year.
185 00:21:10.350 ⇒ 00:21:13.329 Demilade Agboola: Right, we can see the spread for last year. My bad.
186 00:21:15.760 ⇒ 00:21:18.119 Demilade Agboola: And we can kind of get the feel of, like.
187 00:21:18.600 ⇒ 00:21:22.829 Demilade Agboola: What was going on as well last year for the different, brands as well.
188 00:21:23.740 ⇒ 00:21:29.529 Demilade Agboola: And then if you come back to the workbook, we can also, again, create more information.
189 00:21:33.020 ⇒ 00:21:35.689 Demilade Agboola: Where you can see things like time to ship.
190 00:21:36.650 ⇒ 00:21:41.320 Demilade Agboola: So, if we were really, like, particular about
191 00:21:42.040 ⇒ 00:21:49.340 Demilade Agboola: how long things take to ship. We can also include that as well, so let’s include that finally, and then we might just publish the dashboard.
192 00:21:50.350 ⇒ 00:21:56.419 Demilade Agboola: Alright, so we can start to say, hey, let’s also see what categories take a long time.
193 00:21:56.650 ⇒ 00:21:57.620 Demilade Agboola: to ship.
194 00:22:02.710 ⇒ 00:22:05.679 Demilade Agboola: And now we can get an idea of this.
195 00:22:06.570 ⇒ 00:22:08.229 Demilade Agboola: We can make it an aggregate.
196 00:22:09.490 ⇒ 00:22:11.810 Demilade Agboola: Of the average time to ship.
197 00:22:13.420 ⇒ 00:22:19.219 Demilade Agboola: So we can kind of see how long… what the average time it takes to ship is for each of the different categories.
198 00:22:20.120 ⇒ 00:22:26.459 Demilade Agboola: And we can also do it in descending, so we can start to see what’s going on.
199 00:22:26.670 ⇒ 00:22:39.100 Demilade Agboola: So your maternity things kind of take the longest time to ship, but as you can tell, most things are actually within the same range, so there’s nothing that really jumps off the page in terms of, like, this takes an exceedingly long amount of time to ship.
200 00:22:39.270 ⇒ 00:22:49.500 Demilade Agboola: But you might also just be aware that jeans takes the shortest, so jeans sell more, but they also, like, ship much faster, which probably plays a role into why they ship so quickly.
201 00:22:49.750 ⇒ 00:23:00.459 Demilade Agboola: And again, you might remember that maternity didn’t sell that well, it was one of the lowest selling. So again, it does seem there might be some relationship with how well things sell versus how well they ship.
202 00:23:00.990 ⇒ 00:23:05.980 Demilade Agboola: And so yeah, we can place the chart there, get a feel of that.
203 00:23:08.200 ⇒ 00:23:16.610 Demilade Agboola: Decide to make it this way. And the idea of doing this, for instance, is we can add a filter and make it, say, over the last, say, 7 days.
204 00:23:17.720 ⇒ 00:23:23.440 Demilade Agboola: And why this might be useful for us is we might say, hey, We want to…
205 00:23:23.940 ⇒ 00:23:31.910 Demilade Agboola: only see when, like, how things are going over time. So, like, every time I open my dashboard, I want to know what the shipping was like over the last 7 days.
206 00:23:31.970 ⇒ 00:23:50.250 Demilade Agboola: And that might make it easy for me to catch potential things that are troublesome, for instance, like, oh, all of a sudden, it’s taking a really long time for genes to shape, or it’s taking a really long time for maternity things to shape, or a number has dropped quite drastically from what I’m used to. So that would also be something we could put here.
207 00:23:50.970 ⇒ 00:23:54.240 Demilade Agboola: And then we can see… shipping…
208 00:23:55.800 ⇒ 00:23:59.480 Demilade Agboola: By brand, or time to ship by brand.
209 00:24:06.880 ⇒ 00:24:10.600 Demilade Agboola: And now, we can go back to our dashboard.
210 00:24:12.010 ⇒ 00:24:19.819 Demilade Agboola: At this point, I don’t necessarily think I… Need the filter.
211 00:24:21.850 ⇒ 00:24:23.420 Demilade Agboola: So we can delete it.
212 00:24:26.960 ⇒ 00:24:32.389 Demilade Agboola: And so, yeah, we now have our dashboard that shows us what’s going on with our data.
213 00:24:33.230 ⇒ 00:24:34.900 Demilade Agboola: And… we can publish.
214 00:24:37.500 ⇒ 00:24:42.200 Demilade Agboola: So now we have a dashboard that… We can’t see.
215 00:24:43.050 ⇒ 00:24:45.970 Demilade Agboola: Come here to the hub.
216 00:24:46.730 ⇒ 00:24:57.500 Demilade Agboola: we will see that this dashboard that we just made is here, and we can look at it. You can show it to people in your organization. People can then come and say, hey, I have
217 00:24:58.490 ⇒ 00:25:03.500 Demilade Agboola: A lot of questions about this, or, you know, they can explore a bit further.
218 00:25:03.920 ⇒ 00:25:07.029 Demilade Agboola: Now, the beauty of…
219 00:25:08.230 ⇒ 00:25:16.739 Demilade Agboola: like, something like Omni, beyond just being able to do all of this and create topics for different users in your organization, again, by
220 00:25:17.090 ⇒ 00:25:24.129 Demilade Agboola: business needs. So, again, you can have your marketing topic, you can have your, revenue topic, you can have your…
221 00:25:24.320 ⇒ 00:25:36.689 Demilade Agboola: Time to, like, your inventory topic, so you can kind of see what’s happening across different things, and you can start to create, like, silos where people that need to see specific things only see what they need to see.
222 00:25:36.850 ⇒ 00:25:43.110 Demilade Agboola: So that can help with things like data governance. So you don’t want, for instance, people to see
223 00:25:44.330 ⇒ 00:25:54.659 Demilade Agboola: People who need to see inventory don’t necessarily need to see, like, customers’ details, for instance, or people who need to see, revenue.
224 00:25:54.850 ⇒ 00:25:56.549 Demilade Agboola: Might not need to see…
225 00:25:57.190 ⇒ 00:26:11.320 Demilade Agboola: people need to see revenue might not need to see things like marketing and, like, the click-through rate and all that stuff. So you can kind of segment it and have your different users using different things. So I’ll take a quick pause here before we continue. Does anyone have any questions?
226 00:26:13.920 ⇒ 00:26:15.500 Jason Wu: How do I…
227 00:26:15.700 ⇒ 00:26:20.379 Jason Wu: do an export out of this if I wanted to get the raw data and do other, other…
228 00:26:21.780 ⇒ 00:26:25.149 Jason Wu: I wasn’t… I didn’t see, kind of, where there was an export option.
229 00:26:25.540 ⇒ 00:26:31.390 Demilade Agboola: Okay, so for that, what you will need to do is, if you would come into the dashboard.
230 00:26:31.590 ⇒ 00:26:36.920 Demilade Agboola: You have the opportunity to drill into the dashboard, so you can drill into it.
231 00:26:37.370 ⇒ 00:26:39.489 Demilade Agboola: So, if you have this…
232 00:26:39.790 ⇒ 00:26:46.969 Demilade Agboola: it will give you the opportunity to drill in, and when you drill in, you can see the data that makes up that dashboard. So…
233 00:26:46.970 ⇒ 00:26:47.760 Shivani Amar: Hmm.
234 00:26:47.760 ⇒ 00:26:55.110 Demilade Agboola: If you have the data, on your dash, you can export it as a CSV, Or, you know…
235 00:26:55.480 ⇒ 00:26:57.690 Demilade Agboola: different formats. Let me see…
236 00:27:04.430 ⇒ 00:27:07.260 Demilade Agboola: Yeah, so… You can download.
237 00:27:07.650 ⇒ 00:27:13.450 Demilade Agboola: You can choose your file type, you can say, hey, I want it to be Excel, JSON, or CSV.
238 00:27:14.370 ⇒ 00:27:21.280 Demilade Agboola: And then, at this point, you can either choose, like, a row limit, or you can say, no, I care for all the results, show me everything.
239 00:27:21.570 ⇒ 00:27:28.570 Demilade Agboola: Apply, like, data formatting if you care for it, and then you can name the file, and you can say, time to ship.
240 00:27:32.710 ⇒ 00:27:35.070 Demilade Agboola: Got it. The shape by category.
241 00:27:35.420 ⇒ 00:27:39.569 Jason Wu: Cool. Yeah, you don’t need to export, I just wanted to make sure that that was available, because I didn’t see it.
242 00:27:39.770 ⇒ 00:27:40.500 Jason Wu: Alright. Clearly.
243 00:27:40.500 ⇒ 00:27:49.929 Demilade Agboola: Gotcha. So that is definitely available. Again, before I play around, are there any things you might want to know about, like, your BI tool being able to do?
244 00:27:50.620 ⇒ 00:27:55.800 Jason Wu: So… Definitely for me, I’m curious to know, kind of like.
245 00:27:56.460 ⇒ 00:28:11.689 Jason Wu: where some of the AI capabilities come in at Omni, so what you’ve shown so far is kind of like, here’s how we kind of, like, access the data, configure it, kind of create, like, you know, explicit dashboards, but if we wanted to do more of kind of an explore mode.
246 00:28:11.790 ⇒ 00:28:18.080 Jason Wu: you know, where we’re trying to answer the questions. I’d love to kind of understand, kind of, like, where Omni’s at in terms of kind of, like.
247 00:28:18.290 ⇒ 00:28:23.189 Jason Wu: all their LLM stuff, and just anything as far as, like, being able to kind of slew through the data.
248 00:28:24.160 ⇒ 00:28:25.320 Demilade Agboola: Yeah, so…
249 00:28:25.580 ⇒ 00:28:36.579 Jason Wu: So, for example, like, can you translate… like, we created a dashboard for, like, can… is it already set up where we can do, like, a translation that says, tell me the best categories that have been performing over the past four quarters?
250 00:28:37.500 ⇒ 00:28:38.880 Demilade Agboola: Oh yeah, definitely.
251 00:28:39.110 ⇒ 00:28:51.430 Demilade Agboola: So, in that case, if you came in here and said, Show me… the best performing…
252 00:28:55.180 ⇒ 00:29:01.860 Demilade Agboola: categories… of, the last… Four months?
253 00:29:03.910 ⇒ 00:29:13.379 Demilade Agboola: by… so obviously, we’ll need to just specify what the column or what we’re defining as best performing. So in this case, it will probably be, like, the, sale…
254 00:29:14.540 ⇒ 00:29:16.330 Demilade Agboola: Price. Sum.
255 00:29:17.740 ⇒ 00:29:20.339 Demilade Agboola: So now Omni will write the query for us.
256 00:29:20.340 ⇒ 00:29:27.389 Shivani Amar: So if you just said, show me the best performing best sales, like, you’ve said sales price some…
257 00:29:27.790 ⇒ 00:29:29.169 Demilade Agboola: I mean, we could try sales, too.
258 00:29:29.600 ⇒ 00:29:32.180 Shivani Amar: No, no, it’s okay, I’m just curious, I’m like, how…
259 00:29:32.180 ⇒ 00:29:32.630 Jason Wu: Eric.
260 00:29:32.630 ⇒ 00:29:36.390 Shivani Amar: I’m gonna figure out how much they’ll No.
261 00:29:36.390 ⇒ 00:29:47.199 Demilade Agboola: Okay, sure, we could actually try sales and see how, like, specific we can… or, like, non-specific we can make it. But, like, by sales price norm, it already gives you both the results, in terms of, like, the raw numbers.
262 00:29:47.200 ⇒ 00:29:47.710 Jason Wu: give you.
263 00:29:47.710 ⇒ 00:29:55.330 Demilade Agboola: there for that. But it also gives you the chart, like, it automatically will show you, like, hey, this is what that looks like. It has created the filter for you.
264 00:29:55.670 ⇒ 00:29:58.369 Demilade Agboola: And as I said, oh, in the past 4 months, so, like.
265 00:29:58.470 ⇒ 00:30:04.300 Demilade Agboola: It’s done that for you on your behalf. You can also add more context to it if you care for.
266 00:30:04.980 ⇒ 00:30:06.440 Jason Wu: Can I challenge this a little bit?
267 00:30:06.800 ⇒ 00:30:08.670 Demilade Agboola: Oh, sure, let’s play, let’s play with this.
268 00:30:08.840 ⇒ 00:30:22.860 Jason Wu: Yeah, like… which categories… Are showing a… Like, above, like, well…
269 00:30:22.960 ⇒ 00:30:29.409 Jason Wu: this is the wrong way to say it, right? But, like, this is a tested thing, right? Like, which categories are dropping in sales?
270 00:30:29.520 ⇒ 00:30:31.980 Jason Wu: Are there any categories dropping in sales?
271 00:30:33.420 ⇒ 00:30:35.210 Jason Wu: Over the past 4 months.
272 00:30:51.080 ⇒ 00:30:52.840 Demilade Agboola: So, what do you mean over the platform, do you mean, like.
273 00:30:53.300 ⇒ 00:30:56.100 Demilade Agboola: That’s also a bit vague, to be fair, but let me… let’s see.
274 00:30:56.100 ⇒ 00:31:10.749 Shivani Amar: Like, I think it’ll be interesting to see, right? Like, it’s like, I’m just curious, like, what kind of… if people are like, in the last quarter, what dropped in sales, then it should… like, I’m like, not it should, but is it comparing quarter to quarter, and it’s saying…
275 00:31:10.750 ⇒ 00:31:18.290 Demilade Agboola: Yeah, that’s part of why I was like, that’s a bit vague, because are you doing the previous quarter against the last four… like, the previous four months against the.
276 00:31:18.290 ⇒ 00:31:23.289 Shivani Amar: During the four period… I’ll just see what it does, I’m just curious. Yeah, sure. Yeah.
277 00:31:31.230 ⇒ 00:31:42.590 Shivani Amar: Across the two halves of the four-month period. So it’s, like, telling you what it does, which is interesting, and it’s, like, if you’re like, oh, I wanted to compare four months versus, you know.
278 00:31:42.760 ⇒ 00:31:48.339 Shivani Amar: 8 months ago, or whatever, then you could, like, probably rewrite the query, but… so this is saying…
279 00:31:49.420 ⇒ 00:31:54.740 Demilade Agboola: The current, like, the previous… so this is the 45 days prior to that.
280 00:31:55.110 ⇒ 00:31:56.000 Shivani Amar: Okay.
281 00:31:56.170 ⇒ 00:31:58.430 Demilade Agboola: Okay, so it’s still working, it’s still doing its thing.
282 00:32:02.590 ⇒ 00:32:03.510 Demilade Agboola: Oh, okay.
283 00:32:05.740 ⇒ 00:32:06.639 Demilade Agboola: Alright, so it’s complete.
284 00:32:06.640 ⇒ 00:32:11.379 Shivani Amar: And then it’s saying, based on those, there are 3 categories with declining sales.
285 00:32:11.730 ⇒ 00:32:12.300 Demilade Agboola: Yeah.
286 00:32:12.500 ⇒ 00:32:15.510 Shivani Amar: Plus, socks and suits and sport coats.
287 00:32:15.650 ⇒ 00:32:16.590 Shivani Amar: Cool.
288 00:32:18.550 ⇒ 00:32:19.290 Shivani Amar: Those buttons.
289 00:32:19.290 ⇒ 00:32:19.620 Jason Wu: But, like.
290 00:32:19.620 ⇒ 00:32:20.400 Shivani Amar: Thank you.
291 00:32:20.750 ⇒ 00:32:26.690 Jason Wu: Yeah, and by the way, like, I don’t mean to be putting you on the spot, you know, but it’s, like, definitely, like.
292 00:32:27.070 ⇒ 00:32:29.310 Jason Wu: Thinking about, like, our business.
293 00:32:29.410 ⇒ 00:32:36.079 Jason Wu: audience here, you know, it’s… it’s… I’m trying to understand, kind of, like, the savviness that they’re gonna need.
294 00:32:36.090 ⇒ 00:32:50.540 Jason Wu: to create the prompt themselves, or is… and to just kind of let them know, right? Like, is Omni’s LLM models, like, sophisticated enough to either, one, ask for the clarification.
295 00:32:50.580 ⇒ 00:33:06.260 Jason Wu: You know, or not. So in this case here, right, like, this query said, I’m going to separate the two halves. Like, can I have a follow-up here right now? And again, I just want to see what it does. Can you make… can you compare… can you make the period a month instead of a 45-day period?
296 00:33:07.010 ⇒ 00:33:08.990 Jason Wu: Like, that’s the prompt I want to do, I would say.
297 00:33:09.110 ⇒ 00:33:11.940 Jason Wu: Can you adjust the period so it’s…
298 00:33:12.070 ⇒ 00:33:15.010 Jason Wu: A calendar month versus a 45-day period.
299 00:33:16.450 ⇒ 00:33:24.040 Demilade Agboola: So, would the first half be, like, would it be comparing the last month against the 45-day… the initial 45-day period ahead of it? Or is it…
300 00:33:24.040 ⇒ 00:33:43.579 Jason Wu: Well, no, what I want to do is, okay, that was the first pass, but now what I want to see is month over month. I don’t want to see 45 day over 45 day, right? It was Omni AI that decided to do a 45-day period for me, right? So I want to add to the prompt to say, instead of a 45-day period, do the same comparison month over month.
301 00:33:43.750 ⇒ 00:33:45.380 Jason Wu: For the past 4 months.
302 00:33:49.530 ⇒ 00:33:58.880 Jason Wu: And by the way, this is looking great, you know, but it’s like, yeah, just kind of, like, seeing, like, how good at… how good is it in terms of, like, refining the prompt?
303 00:33:59.210 ⇒ 00:33:59.880 Demilade Agboola: Okay.
304 00:34:00.630 ⇒ 00:34:07.229 Jason Wu: Instead of a 45-day period. Like, instead of the 40-day, 45-day period.
305 00:34:27.219 ⇒ 00:34:28.389 Demilade Agboola: Okay, so it’s true.
306 00:34:29.010 ⇒ 00:34:39.440 Jason Wu: Yeah, like, I don’t have a problem with, like, it saying 45D in the beginning, but what I’m now trying to figure out is, okay, now I want to, like, start to refine, right? So, it’s doing it right now, right? So this is pretty cool.
307 00:34:39.860 ⇒ 00:34:40.440 Demilade Agboola: Yeah.
308 00:34:40.800 ⇒ 00:34:42.829 Demilade Agboola: So you’re showing us the changes…
309 00:34:43.170 ⇒ 00:34:48.449 Demilade Agboola: Across the different, month comparisons, so… and the ability is you can always, like.
310 00:34:49.850 ⇒ 00:34:50.380 Jason Wu: QF.
311 00:34:50.380 ⇒ 00:35:03.550 Demilade Agboola: kind of see what it’s doing behind the scenes. If you don’t care for something, you could drop a certain column, so you could remove… if you’re like, I don’t really care about now versus October, for instance, I only care about, like, back-to-back months versus, like.
312 00:35:03.550 ⇒ 00:35:05.469 Jason Wu: Yeah, exactly, like, November versus October.
313 00:35:05.670 ⇒ 00:35:07.000 Demilade Agboola: Exactly, so…
314 00:35:07.740 ⇒ 00:35:20.240 Demilade Agboola: Yeah, but it does it, and I think one of the beautiful things about Omni, personally, is that it talks. Like, it doesn’t just do things in silence and ask you an answer. You kind of can read and see what it’s done.
315 00:35:20.320 ⇒ 00:35:30.240 Demilade Agboola: And if you’re like, I don’t agree with this logic, or I want the logic changed, you can then talk to it and be like, yo, change this logic to this, and it will do it.
316 00:35:31.010 ⇒ 00:35:34.749 Jason Wu: I know, I know both Utem and, Siobhani have to drop.
317 00:35:34.970 ⇒ 00:35:45.209 Jason Wu: For a hard stop, so I don’t know if it makes sense for me to keep asking questions or not, and I want to respect your time, too. What would be the next steps here? Is this something that we have access to, to kind of play with?
318 00:35:45.470 ⇒ 00:35:49.079 Jason Wu: Is this, like, yeah, like…
319 00:35:49.210 ⇒ 00:35:51.940 Jason Wu: Trying to understand more about, like, how the tool works.
320 00:35:52.120 ⇒ 00:35:53.969 Jason Wu: What’s the next step here?
321 00:35:54.690 ⇒ 00:35:57.270 Jason Wu: To kind of consider, you know, for consideration.
322 00:35:57.520 ⇒ 00:36:00.570 Jason Wu: I don’t know if this is your question, by the way, or if this is more a question for us.
323 00:36:01.170 ⇒ 00:36:05.100 Demilade Agboola: I think it might be more of a question for Awash, because, like, I’m not exactly sure of the details of your…
324 00:36:05.100 ⇒ 00:36:14.030 Awaish Kumar: Yeah, so… yeah, like, if you… if you want to play with it, I can… I can, like, sync with Utam, and maybe we can share.
325 00:36:14.420 ⇒ 00:36:18.729 Awaish Kumar: the access, so you can just try it out. Are there…
326 00:36:18.730 ⇒ 00:36:25.409 Jason Wu: Are there other tools, Awish, that… that you’re planning to kind of show us, or is Omni, like, the recommendation?
327 00:36:26.480 ⇒ 00:36:32.780 Awaish Kumar: Yeah, like… So, Omni, like, we do have different tools which we could…
328 00:36:32.840 ⇒ 00:36:48.619 Awaish Kumar: like, have an assessment for, for example, Tableau or Power BI, but most likely, our recommendation will be the Omni, because, like, it provides us, like, best-in-class AI assistant.
329 00:36:48.770 ⇒ 00:36:49.190 Jason Wu: Yeah.
330 00:36:49.190 ⇒ 00:36:51.899 Awaish Kumar: Like, not any other tool will have that.
331 00:36:51.900 ⇒ 00:36:57.239 Jason Wu: Yeah, I’m less thrilled with Tableau. Personally, yeah. Okay.
332 00:36:57.670 ⇒ 00:37:08.510 Awaish Kumar: So, if we use any other tool, we will have to, like, use a separate AI assistant. And while in Omni, we’ll have inbuilt, inside of it. So…
333 00:37:09.400 ⇒ 00:37:10.400 Jason Wu: Yeah, I…
334 00:37:11.660 ⇒ 00:37:19.399 Jason Wu: Yeah, I think it’s a broader discussion with UTSM as well, but it’s, okay, understanding how this tool works…
335 00:37:19.680 ⇒ 00:37:24.589 Jason Wu: Like, I think there’s a matter of, like, well, let’s play with it, but do we… are we gonna…
336 00:37:24.700 ⇒ 00:37:31.180 Jason Wu: play with it using the demo data that, and I’m sorry, I’m gonna mispronounce your name. Is it Demolade?
337 00:37:32.840 ⇒ 00:37:33.860 Awaish Kumar: Tamilade.
338 00:37:34.230 ⇒ 00:37:50.250 Jason Wu: Like, is it… is it playing with it against that test data, or are we in a position to start a trial with Omni and hooking it up to anything? And if that’s the case, when is that, like, when does that become an option?
339 00:37:51.540 ⇒ 00:37:52.529 Awaish Kumar: So, like…
340 00:37:52.800 ⇒ 00:38:04.490 Awaish Kumar: Right now, we are just demoing tools, like… like Omni, if you… like, it’s more of a question from… for you, like, if you want us to, like, also
341 00:38:04.490 ⇒ 00:38:14.819 Awaish Kumar: try out different other tools and create, like, assessment, similar to what we did for ETLN warehouses. We can go and do that, and…
342 00:38:14.850 ⇒ 00:38:17.670 Awaish Kumar: Demo a few tools, and then you can say, okay.
343 00:38:17.860 ⇒ 00:38:21.210 Awaish Kumar: We are okay with, for example, Omni, and then…
344 00:38:21.210 ⇒ 00:38:21.730 Jason Wu: So…
345 00:38:21.730 ⇒ 00:38:23.610 Awaish Kumar: We can, start a trial.
346 00:38:25.680 ⇒ 00:38:36.579 Jason Wu: Yeah, let’s have this broader discussion with Utem here. What goes through my mind is, obviously, we’ve seen one tool. I think it’s looking good. I don’t know…
347 00:38:37.230 ⇒ 00:38:43.830 Jason Wu: Part of it is going to be based on the timing, on when we would think we are ready to have a tool on top.
348 00:38:44.150 ⇒ 00:38:45.769 Jason Wu: Of our data warehouse.
349 00:38:45.770 ⇒ 00:38:51.909 Awaish Kumar: kind of… we are… we are kind of ready, like, we are already starting building models, so we might not…
350 00:38:53.050 ⇒ 00:39:05.149 Awaish Kumar: Yeah, so we might not be able to build, like, the really complex dashboards, but we will be able to show something, like, we had some questions for retail, that we can answer normally, things like that.
351 00:39:05.150 ⇒ 00:39:21.639 Jason Wu: And that’s where I’m going, right? Is, like, from the last call, I’m sensing an urgency and appetite to start playing with the data and exploring the data. So, the question is, is when does the team feel ready that we have enough
352 00:39:21.860 ⇒ 00:39:37.309 Jason Wu: ingested into Snowflake that we can start to play with this data, against real data now, right? And if the answer is now, then I would say, okay, let’s not spend too much time then exploring different dashboards and get on a trial with something right away, right? Because…
353 00:39:37.310 ⇒ 00:39:43.139 Jason Wu: Like, for me, it’s about making sure that we’ve got, like, essentially kind of the buy-in
354 00:39:43.140 ⇒ 00:39:49.790 Jason Wu: On, like, the power of the data, you know, and if there are questions that… Can’t be answered.
355 00:39:49.790 ⇒ 00:40:11.810 Jason Wu: right, via the Omni tool. Is it because of the tool limitation? Is it because we don’t have the right model set up yet? You know, I think those questions just kind of come out of it, but it’s more of a matter of, like, yeah, like, you know, based on kind of the timeframes that we have here, like, when is that possible? And if it’s possible now, then I would say, okay, you know, you know, I’m happy to say.
356 00:40:11.980 ⇒ 00:40:16.740 Jason Wu: You know, if this is the number one recommendation, and that’s the number one recommendation by far.
357 00:40:16.870 ⇒ 00:40:21.139 Jason Wu: you know, then let’s figure out how do I get a trial for it.
358 00:40:21.390 ⇒ 00:40:33.250 Jason Wu: you know, if it’s… if there’s… if there’s another tool that we should consider based upon the questions that are being asked by the business, right? And that’s where I’m looking to… to you and Uttoman Brainforce
359 00:40:33.720 ⇒ 00:40:42.820 Jason Wu: Yeah, based on the question that the team is looking to get, there’s actually a different tool that we should think about, then that’s kind of the recommendations that we’d be looking for.
360 00:40:43.530 ⇒ 00:40:50.630 Awaish Kumar: Yeah, I don’t, like, I don’t think so there’s any big difference in these BI tools, where…
361 00:40:50.630 ⇒ 00:41:03.919 Awaish Kumar: they can’t provide an answer for a business question. It’s more like of a… like, how comfortable are people using this tool? Like, it’s… it might not be, like, just us using it.
362 00:41:03.920 ⇒ 00:41:21.709 Awaish Kumar: Yeah. Maybe other people will come in and start to use this. It’s like, how many… how different people in the company will be able… will be comfortable with Omni versus Tableau, for example, and the pricing and stuff like that.
363 00:41:21.930 ⇒ 00:41:27.659 Jason Wu: Yeah, and I can say right now, like, unless Tableau’s got a good LLM setup, like, that’s out.
364 00:41:27.660 ⇒ 00:41:41.639 Jason Wu: You know, it’s like, we, you know, I don’t want to be in a situation where we need kind of, like, an analyst to keep building queries, you know, and, like, setting up the dashboards. Setting us in base dashboards is great, but self-service at this point, should really be about, like.
365 00:41:41.640 ⇒ 00:41:49.419 Jason Wu: how does the end user be able to kind of explore the data on their own with some limited training, right? Obviously, the limited training is…
366 00:41:49.420 ⇒ 00:42:02.140 Jason Wu: you know, what kind of data is available now, you know, and, you know, perhaps a couple, like, sample prompts that we’re comfortable with. One thing that we didn’t talk about, Devin Lottie, is… obviously, like, what was interesting is.
367 00:42:02.150 ⇒ 00:42:05.020 Jason Wu: in order to kind of start that AI flow.
368 00:42:05.190 ⇒ 00:42:08.369 Jason Wu: You kind of had to first say, well, let’s start with this.
369 00:42:08.590 ⇒ 00:42:10.990 Jason Wu: Right? Like, it’s building off of, like, this…
370 00:42:11.130 ⇒ 00:42:23.479 Jason Wu: like, frame set, like we said, we’re gonna start with orders. So I’d be… I don’t know how that works, you know, when we start wanting to go in across different things. You know, if it’s like, hey, I’m looking at…
371 00:42:24.800 ⇒ 00:42:41.830 Jason Wu: I don’t know, I… you know, I’m just thinking off the top of my head, because, you know, without understanding kind of what all the models are, right? But if I’m trying to compare different channels, or if I’m trying to compare, you know, like, we’re not there yet, but, like, you know, what does LTV look like for a customer in retail versus a customer in…
372 00:42:42.000 ⇒ 00:42:53.200 Jason Wu: you know, e-commerce, you know, I’m making shit up now at this point, you know, but it’s like, you know, like, does… does the AI tool have that sophistication to be able to go across different…
373 00:42:53.380 ⇒ 00:42:59.949 Jason Wu: you know, models, you know, or do we all… does it only, like, allow us to, like, explore down one set? You know, and is that just a
374 00:43:00.360 ⇒ 00:43:02.339 Jason Wu: I don’t know how many models we create or not.
375 00:43:02.340 ⇒ 00:43:18.099 Awaish Kumar: So, like, Jason, to answer that, we have, like… that’s why we have this layer transformation layer. That’s where we will be creating different models, like, which… for, like, omnichannel, we will have some models which will join the data.
376 00:43:18.100 ⇒ 00:43:34.490 Awaish Kumar: from all different channels, and we will have one table, basically, which have all the data from all channels for all the customers. And using that data, then we use AI on top of it to answer our questions.
377 00:43:35.290 ⇒ 00:43:47.040 Demilade Agboola: I would also say that it’s… I would personally advise that when you are doing such, like, joins and creating topics, because when we’re using Omni right now, we were exploring within a topic.
378 00:43:47.170 ⇒ 00:43:54.279 Demilade Agboola: But Omni also has you… you can also go to Omni and write queries that will join across multiple things.
379 00:43:54.420 ⇒ 00:44:00.999 Jason Wu: Yeah. But I would always advise that when you are doing things like that, especially if you’re not necessarily, like, a data person.
380 00:44:01.000 ⇒ 00:44:04.180 Demilade Agboola: who has experience. These things can be very, like, tricky.
381 00:44:04.310 ⇒ 00:44:13.780 Demilade Agboola: And so you don’t want to have people who are just like, I want to do this to this, but they might not know what the unique key is, for instance. So, if you don’t know.
382 00:44:13.780 ⇒ 00:44:23.080 Jason Wu: Well, so that’s where I go back to the AI, like, is the LLM intelligent enough to say, hey, here are the keys, right? Or, you know, again, it’s more of the, like.
383 00:44:23.420 ⇒ 00:44:33.050 Jason Wu: again, it’s just more of an exploration exercise at this point, right? But it’s just like, you know, what’s the level of discernment that it has versus needing to be explicit in the prompt?
384 00:44:33.710 ⇒ 00:44:36.519 Demilade Agboola: Yeah, so I will say, like, for the most part.
385 00:44:37.450 ⇒ 00:44:43.519 Jason Wu: I would always, like, when using LLMs, I always tend to lean towards being explicit, so that you don’t run into, like.
386 00:44:43.610 ⇒ 00:44:54.999 Demilade Agboola: those frustrations, like, those moments of frustration where he doesn’t really understand what you’re saying. But in those terms of, like, joining the tables, because I know I have used Omni behind the scenes when I was creating a topic to make those joins that I needed.
387 00:44:55.030 ⇒ 00:45:10.049 Demilade Agboola: But it will try. It would always make the assumption of this is the ID to join to this, but sometimes if you have multiple IDs, it might not necessarily get it right, and if it doesn’t do a good join, in that case, what you have is an empty table, and…
388 00:45:10.360 ⇒ 00:45:10.680 Jason Wu: Yeah.
389 00:45:10.680 ⇒ 00:45:22.360 Demilade Agboola: or you might have, like, a bad joint, and it’s very rare. In that case, if you’re not a data person, you’re just going to be confused as to what’s going on, or you might just not even know and give out bad data. So I would always be…
390 00:45:22.770 ⇒ 00:45:27.129 Demilade Agboola: I always suggest that they use, curator tables.
391 00:45:27.270 ⇒ 00:45:33.680 Demilade Agboola: or created datasets, or created topics, as they’re called within Omni, and so what we can help you do is set up
392 00:45:33.980 ⇒ 00:45:36.220 Demilade Agboola: Multiple of these, sort of, topics.
393 00:45:36.220 ⇒ 00:45:36.750 Jason Wu: Yes.
394 00:45:36.750 ⇒ 00:45:46.130 Demilade Agboola: according to business needs. And people can give us feedback and be like, hey, I might need a bit more data, and we can have that present, and then they can explore away within those topics.
395 00:45:46.650 ⇒ 00:45:53.949 Jason Wu: Yeah, 100%, I’m aligned with that. You know, I think there’s, like, two cases that’ll happen with, like, exploration, right? You’re gonna put, like.
396 00:45:54.060 ⇒ 00:46:08.509 Jason Wu: a business user, like, will… like, so he’s like, you know, our chief revenue officer, right? He’ll ask questions, right? And then if the data comes back and says, this doesn’t look right, right, he’s gonna come back to me, right, who’ll then come back to Brainforge and say, how do we do it, right? And is it about…
397 00:46:09.200 ⇒ 00:46:23.319 Jason Wu: the prompt was wrong, and we educate, like, how to do the prompt differently, or is it the fact that it’s like, actually, you’re asking a question that our current models don’t support? Let’s build that out first so we can continue to answer that question, right? So.
398 00:46:23.320 ⇒ 00:46:37.690 Jason Wu: I wouldn’t expect the system to be, like, so intelligent that it figures it out, but it’s more of a, like, what does that limitation look like? And then, like, how do we create the right processes? So, someone like Will can say, oh, okay, this is what the limitation is, and then if there’s more…
399 00:46:37.690 ⇒ 00:46:40.570 Jason Wu: that needs to be done. It’s ask…
400 00:46:40.570 ⇒ 00:47:00.060 Jason Wu: an in-house expert, right? You know, whether or not it’s Brainforge or, you know, we’ve got a couple people that I definitely trust is more savvy, that would look at it and, like, question the data, than others, you know, so it’s just really trying to understand, like, who are going to be, like, our power users, right? Ones that can kind of figure it out, because I’ve definitely been in your boat, too, where…
401 00:47:00.060 ⇒ 00:47:06.120 Jason Wu: you know, I’ll use, you know, Claude, and I’ll upload a bunch of data, and then it’ll spit something out. I’m like, well, that’s totally wrong.
402 00:47:06.120 ⇒ 00:47:12.530 Jason Wu: Right? And then I’ll say, this doesn’t look right, and then Claude will go, hey, you’re right, this is not right.
403 00:47:12.530 ⇒ 00:47:14.460 Demilade Agboola: Let me fix this for you, right? So…
404 00:47:14.460 ⇒ 00:47:24.910 Jason Wu: You know, so I think I understand your point, and it’s certainly valid, but I think it’s more of a matter of saying, okay, well, what are those limitations, and then just, yeah, and it’s just like the…
405 00:47:25.010 ⇒ 00:47:28.880 Jason Wu: the gotchas that that will have. So, you know.
406 00:47:28.880 ⇒ 00:47:29.790 Awaish Kumar: Mostly…
407 00:47:30.330 ⇒ 00:47:31.130 Jason Wu: Yeah.
408 00:47:31.440 ⇒ 00:47:44.589 Awaish Kumar: Yeah, mostly those will be clear with the topics, like, whatever, like, we will have the topics which will basically define, like, what kind of data
409 00:47:44.640 ⇒ 00:47:53.350 Awaish Kumar: it will be used to answer, like, what kind of questions, right? So, that’s the one way of handling it.
410 00:47:53.460 ⇒ 00:48:12.699 Awaish Kumar: Secondly, AI will, at least will know, like, if we say… we, for example, in e-com demo, we have category and brand, but if you say, okay, I need to see the performance by region, and there’s no region, like, then it can tell you also, like, we don’t have region information.
411 00:48:15.130 ⇒ 00:48:15.920 Jason Wu: Got it.
412 00:48:28.950 ⇒ 00:48:37.570 Awaish Kumar: So, for these trivial things, AI will… will clarify, but it’s more like… more, like, business…
413 00:48:37.780 ⇒ 00:48:51.100 Awaish Kumar: Related questions, like, for profit, like, how we are calculating it, like, total price versus minus, COGS minus, taxes, or what, like, the way we are calculating it.
414 00:48:51.100 ⇒ 00:48:58.700 Awaish Kumar: AI might not understand how we are exactly doing it. That’s why we are going to handle those things inside dbt.
415 00:48:58.710 ⇒ 00:49:06.399 Awaish Kumar: And then, yeah, we will leave EI to just use the… Like, the standard… standard data.
416 00:49:07.070 ⇒ 00:49:08.480 Jason Wu: Yeah, that makes sense.
417 00:49:11.880 ⇒ 00:49:13.160 Jason Wu: Okay.
418 00:49:13.160 ⇒ 00:49:13.830 Awaish Kumar: This is helpful.
419 00:49:13.990 ⇒ 00:49:27.359 Jason Wu: First time seeing it. For what it’s worth, the Omni CEO, for some reason, keeps pinging me on LinkedIn, saying, hey, you should consider a tool. I’m like, I haven’t answered them yet, but I’m like, I’m pretty sure I’m seeing a demo of your tool anyway, so…
420 00:49:27.690 ⇒ 00:49:41.529 Jason Wu: This has been helpful. I don’t have any other questions right now. The next question, I think, really is, like, what’s the next step? You know, is, like, are we at a point where we can connect a trial of this to our current Snowflake environment, even though we know that’s in process?
421 00:49:41.560 ⇒ 00:49:54.300 Jason Wu: you know, if that’s still a ways away, is there a way we can get access to the demo that was shown today, so we can kind of play with it more? You know, I think that’s just more of, like, you know, what makes sense first, given the timing of things right now on your side of Wish?
422 00:49:55.380 ⇒ 00:49:57.110 Awaish Kumar: Yeah, so…
423 00:49:57.330 ⇒ 00:50:12.509 Awaish Kumar: like, according to our plan, like, we wanted to have a decision on BI tool by the end of January, but I will go back to Utam and will clarify if we want to start a trial or give you the access to our demo instance.
424 00:50:12.510 ⇒ 00:50:12.900 Jason Wu: Okay.
425 00:50:12.900 ⇒ 00:50:15.580 Awaish Kumar: And… yeah, so…
426 00:50:15.580 ⇒ 00:50:27.240 Jason Wu: Yeah. I just want to know what those… I just want to know what those options are. Obviously, it’s a lot more powerful when you get to use your own data, but I’m also very wary of…
427 00:50:27.260 ⇒ 00:50:41.190 Jason Wu: giving access to tools with its own data, exactly as kind of Demolites had said, there may be issues, you know, where we’re still refining the models themselves, and I don’t want to be misleading as well if people start to play with it. So, we’ll need to kind of understand who should have access to it, you know.
428 00:50:41.190 ⇒ 00:50:42.170 Awaish Kumar: and whatnot.
429 00:50:42.790 ⇒ 00:50:49.999 Awaish Kumar: Yeah, like, we already told Phil that we might give him the access for Snowflake, so if…
430 00:50:50.200 ⇒ 00:51:00.010 Awaish Kumar: like, if he gets access to Snowflake or Omni, it’s the same. Kind of, like, he will be just… he’ll be able to see the tables and explore the data.
431 00:51:00.850 ⇒ 00:51:05.349 Jason Wu: That… yeah, he’s one of the guys I trust, by the way. He’ll…
432 00:51:05.720 ⇒ 00:51:09.509 Jason Wu: He’s smart enough to figure out what’s right and what’s wrong.
433 00:51:11.150 ⇒ 00:51:12.140 Jason Wu: Okay, cool.
434 00:51:12.140 ⇒ 00:51:12.720 Awaish Kumar: Okay.
435 00:51:12.950 ⇒ 00:51:13.720 Awaish Kumar: Okay, I’ll…
436 00:51:13.720 ⇒ 00:51:14.180 Jason Wu: interrupt you.
437 00:51:14.600 ⇒ 00:51:15.120 Awaish Kumar: Yes.
438 00:51:15.120 ⇒ 00:51:17.930 Jason Wu: Sounds good. Thanks for the demo, Devil. Good to meet you.
439 00:51:18.310 ⇒ 00:51:19.289 Demilade Agboola: Good to meet you.
440 00:51:20.330 ⇒ 00:51:21.040 Jason Wu: By now.