Meeting Title: Brainforge Platform Case Study Discussion Date: 2025-07-17 Meeting participants: Hannah Wang, Casie Aviles
WEBVTT
1 00:00:54.510 ⇒ 00:00:55.350 Casie Aviles: Ian.
2 00:00:56.470 ⇒ 00:00:56.940 Hannah Wang: Hey?
3 00:00:57.790 ⇒ 00:00:58.930 Hannah Wang: How’s it going.
4 00:01:00.452 ⇒ 00:01:01.360 Casie Aviles: Yeah. Doing? Good.
5 00:01:02.540 ⇒ 00:01:03.270 Hannah Wang: Thanks.
6 00:01:08.400 ⇒ 00:01:12.649 Hannah Wang: Okay, let me gather my thoughts.
7 00:01:15.890 ⇒ 00:01:22.460 Hannah Wang: Okay, so basically, for any of the internal work that you guys do
8 00:01:23.460 ⇒ 00:01:33.150 Hannah Wang: we just wanna be able to get some case studies out of it so that we can start sending it out to people. So I know that you worked on a zoom.
9 00:01:33.790 ⇒ 00:01:40.229 Hannah Wang: I don’t. I don’t even remember what it was a zoom to Google drive to
10 00:01:40.620 ⇒ 00:01:43.359 Hannah Wang: an agent project. Is that right?
11 00:01:44.045 ⇒ 00:01:50.690 Casie Aviles: Yeah, yeah, that’s pretty much it. We don’t have like a an official name for all of that yet. But yeah, that’s pretty much it.
12 00:01:50.930 ⇒ 00:02:06.070 Hannah Wang: Okay? So if you could share your screen with me just because I wanna be able to grab some screenshots of the project to put in the case. Study. You can just share, I guess, like the work.
13 00:02:06.240 ⇒ 00:02:07.419 Hannah Wang: Hello!
14 00:02:13.430 ⇒ 00:02:13.995 Casie Aviles: Okay.
15 00:02:15.190 ⇒ 00:02:21.959 Hannah Wang: So sorry there’s yeah. Your zoom like things are co blocking some of the.
16 00:02:22.950 ⇒ 00:02:31.824 Hannah Wang: or I think it’s just on my end. But if you can move your tab. Yeah, that’s good. I think that’s good.
17 00:02:32.210 ⇒ 00:02:32.820 Casie Aviles: Okay.
18 00:02:33.140 ⇒ 00:02:38.920 Hannah Wang: Okay. So do you just wanna give me like a high level overview of what this project is.
19 00:02:40.715 ⇒ 00:02:42.020 Casie Aviles: Yeah, sure. Okay.
20 00:02:42.430 ⇒ 00:02:45.507 Casie Aviles: So I guess I I can start with just
21 00:02:46.960 ⇒ 00:02:49.720 Casie Aviles: with the paint, with like, what?
22 00:02:50.520 ⇒ 00:02:53.474 Casie Aviles: What was the the reason why we worked on this
23 00:02:54.540 ⇒ 00:03:01.800 Casie Aviles: so basically, we a lot of the stuff like the the Zoom Meeting transcripts were still being manually
24 00:03:02.250 ⇒ 00:03:04.910 Casie Aviles: sent to chat gpt so we would like
25 00:03:07.360 ⇒ 00:03:09.554 Casie Aviles: go to zoom and we would.
26 00:03:10.370 ⇒ 00:03:16.380 Casie Aviles: I have to sign in. But yeah, we’d go to zoom and we download the transcript. So not a lot of the contacts that we have
27 00:03:16.620 ⇒ 00:03:21.700 Casie Aviles: internally are easily usable by the AI.
28 00:03:21.980 ⇒ 00:03:28.179 Casie Aviles: So that’s kind. That’s kind of one. The reason why we developed this. And also another reason was that
29 00:03:29.112 ⇒ 00:03:31.369 Casie Aviles: you know, we we started to get
30 00:03:31.957 ⇒ 00:03:38.570 Casie Aviles: we like our limits. Our storage limits is also starting to increase. So
31 00:03:40.410 ⇒ 00:03:46.849 Casie Aviles: wait. Sorry. Yeah. So that’s why we also decided to start moving the the recording files
32 00:03:47.060 ⇒ 00:03:48.959 Casie Aviles: to our own storage.
33 00:03:50.115 ⇒ 00:03:51.300 Casie Aviles: Yeah. So that’s.
34 00:03:51.300 ⇒ 00:03:52.339 Hannah Wang: Google, drive.
35 00:03:53.811 ⇒ 00:03:59.420 Casie Aviles: Yeah, Google drive was, initially, yeah, it was the first.st
36 00:04:00.221 ⇒ 00:04:04.658 Hannah Wang: Location that we decided on, but we also moved away from that. Now.
37 00:04:06.095 ⇒ 00:04:09.130 Casie Aviles: Yeah, we’re using amazon.
38 00:04:09.800 ⇒ 00:04:10.350 Hannah Wang: Oh, that’s great!
39 00:04:10.350 ⇒ 00:04:11.110 Casie Aviles: 3, yeah.
40 00:04:11.110 ⇒ 00:04:11.810 Hannah Wang: Okay.
41 00:04:13.330 ⇒ 00:04:18.430 Casie Aviles: Okay? So yeah. And so what we did initially was, we.
42 00:04:19.912 ⇒ 00:04:23.210 Casie Aviles: develop just the the pipeline to move it.
43 00:04:25.536 ⇒ 00:04:33.059 Casie Aviles: to to Google drive and then create like summaries lock summaries as yeah, I can. Sorry I can send
44 00:04:33.630 ⇒ 00:04:34.760 Casie Aviles: some here.
45 00:04:34.910 ⇒ 00:04:36.799 Casie Aviles: I can show some
46 00:04:40.050 ⇒ 00:04:46.240 Casie Aviles: like for the AI team, we get these summaries. So this is where we started at.
47 00:04:47.130 ⇒ 00:04:52.439 Casie Aviles: And then, before this was still on Google Drive, and we’d have the
48 00:04:52.550 ⇒ 00:04:55.609 Casie Aviles: summary over here. As I reply to the thread.
49 00:04:56.930 ⇒ 00:05:03.509 Casie Aviles: And then, right now we’re just, you know, experimenting on this meeting, scoring. But yeah. And then.
50 00:05:03.840 ⇒ 00:05:17.899 Casie Aviles: after that, we decided that we should also create. We thought we should also create like a platform or like this interface, to make it easier for people and to actually be able to use AI and chat with the transcripts.
51 00:05:18.110 ⇒ 00:05:21.730 Casie Aviles: So that’s what that’s like. The the
52 00:05:21.950 ⇒ 00:05:23.818 Casie Aviles: reason why we have this now.
53 00:05:24.680 ⇒ 00:05:29.800 Casie Aviles: yeah. And we so over here, we have this, this dashboard and
54 00:05:30.840 ⇒ 00:05:35.909 Casie Aviles: this pretty much mimics. You know how Zoom displays the meetings.
55 00:05:37.270 ⇒ 00:05:42.319 Casie Aviles: But yeah, we we are also able to look at the participants. We could. We have all these
56 00:05:43.050 ⇒ 00:05:47.489 Casie Aviles: meeting, I think, light analytics. And then we could also do some search
57 00:05:47.740 ⇒ 00:05:52.370 Casie Aviles: like, ABC, yeah. ABC, stand up, and then
58 00:05:52.520 ⇒ 00:05:56.870 Casie Aviles: we could go and click into the meetings like, for example, we want to
59 00:05:58.067 ⇒ 00:06:04.102 Casie Aviles: investigate this meeting, or like we, we want to ask something. So now we have these
60 00:06:05.550 ⇒ 00:06:10.560 Casie Aviles: messages here that we could send. And this is to help
61 00:06:11.917 ⇒ 00:06:17.079 Casie Aviles: the team basically go back to things that they probably missed like.
62 00:06:18.510 ⇒ 00:06:21.129 Casie Aviles: yeah, and makes it just easier to
63 00:06:21.900 ⇒ 00:06:29.189 Casie Aviles: have all the contacts that they need in here, and I think the benefit is also, for you know, for for an Async company like us
64 00:06:29.450 ⇒ 00:06:35.979 Casie Aviles: we can go back and reference, the transcripts meeting transcripts.
65 00:06:36.320 ⇒ 00:06:37.940 Casie Aviles: Should I go through each
66 00:06:38.340 ⇒ 00:06:43.400 Casie Aviles: of what else do I need to go through right? I don’t really have a script to follow, but.
67 00:06:43.630 ⇒ 00:06:49.576 Hannah Wang: No worries. Yeah, I will. Start asking you questions. I just wanted like a high, high, level overview of what?
68 00:06:50.913 ⇒ 00:07:06.820 Hannah Wang: What the project does so is this just like the platform project, like, I know this. This is called the Platform, and you’re building it out. When Utam said zoom to drive to agent project like, was that a specific part of the platform
69 00:07:07.050 ⇒ 00:07:12.939 Hannah Wang: like a specific yeah project within the platform, or when he says.
70 00:07:13.900 ⇒ 00:07:19.059 Hannah Wang: Zoom to drive to agent like. Is that just like this platform as a whole.
71 00:07:19.630 ⇒ 00:07:21.814 Casie Aviles: I think, yeah, I think it’s just this
72 00:07:22.956 ⇒ 00:07:25.323 Casie Aviles: Because at at the back end it’s just, you know.
73 00:07:26.190 ⇒ 00:07:28.000 Casie Aviles: yeah, we’re taking all the zoom.
74 00:07:30.210 ⇒ 00:07:35.479 Casie Aviles: all the Zoom Meetings, I mean the recording recording files, and we’re
75 00:07:35.700 ⇒ 00:07:39.540 Casie Aviles: transferring that to our own storage and our own. Where, where,
76 00:07:40.610 ⇒ 00:07:42.919 Casie Aviles: This is also where the platform is.
77 00:07:43.620 ⇒ 00:07:48.679 Casie Aviles: basically, that was kind of like the foundation of this platform.
78 00:07:51.940 ⇒ 00:07:55.390 Casie Aviles: Yeah, I think I I’m not sure if I’m making sense. But.
79 00:07:55.630 ⇒ 00:08:00.489 Hannah Wang: Or you can just show me also, like the workflow wherever you built
80 00:08:00.670 ⇒ 00:08:03.899 Hannah Wang: the workflow like the back end stuff. And just like.
81 00:08:04.580 ⇒ 00:08:06.070 Hannah Wang: yeah, you can show me that, too.
82 00:08:07.450 ⇒ 00:08:11.242 Casie Aviles: Okay, let me just
83 00:08:11.790 ⇒ 00:08:26.230 Casie Aviles: Okay. But yeah, we have the workflow here on windmill. So windmill is basically lets us run our custom
84 00:08:26.660 ⇒ 00:08:28.989 Casie Aviles: python scripts, and
85 00:08:30.115 ⇒ 00:08:39.549 Casie Aviles: that’s responsible for. So there are like, I have 2 scripts here. So the 1st one is. It’s it’s like, you know, it’s a zoom event subscription. So it’s like a list.
86 00:08:40.900 ⇒ 00:08:47.460 Casie Aviles: So whenever Zoom has an available cloud recording, this will get triggered.
87 00:08:47.970 ⇒ 00:08:49.589 Casie Aviles: So yeah, as you can see.
88 00:08:50.230 ⇒ 00:08:53.839 Casie Aviles: look here, this, this is like the payload that we receive from Zoom.
89 00:08:53.960 ⇒ 00:08:59.480 Casie Aviles: We get the the meeting topic and all of those metadata, and
90 00:08:59.670 ⇒ 00:09:02.750 Casie Aviles: after that it will trigger another script.
91 00:09:04.997 ⇒ 00:09:06.869 Casie Aviles: Which is this one?
92 00:09:07.674 ⇒ 00:09:13.640 Casie Aviles: Zoom to aws. So this is what is actually exporting to our bucket.
93 00:09:14.050 ⇒ 00:09:14.859 Hannah Wang: I see.
94 00:09:18.960 ⇒ 00:09:22.579 Hannah Wang: And then from the bucket. How does it populate the front end.
95 00:09:24.920 ⇒ 00:09:25.280 Casie Aviles: Oh!
96 00:09:25.280 ⇒ 00:09:26.869 Hannah Wang: Or like, How do you? Yeah.
97 00:09:27.610 ⇒ 00:09:30.969 Casie Aviles: There’s also another piece to this, where we
98 00:09:33.230 ⇒ 00:09:36.280 Casie Aviles: where where we store the transcripts to
99 00:09:36.640 ⇒ 00:09:38.780 Casie Aviles: super base. So super base is.
100 00:09:39.720 ⇒ 00:09:45.700 Casie Aviles: That’s where we have our our meeting transcripts, and then.
101 00:09:45.700 ⇒ 00:09:46.020 Hannah Wang: So.
102 00:09:46.020 ⇒ 00:09:46.960 Casie Aviles: Metadata.
103 00:09:47.170 ⇒ 00:09:49.891 Casie Aviles: And that’s what is. That’s where this
104 00:09:50.470 ⇒ 00:09:54.659 Casie Aviles: interface. That’s yeah. That’s what what’s populating all of this.
105 00:09:54.660 ⇒ 00:09:55.450 Hannah Wang: Okay.
106 00:09:58.720 ⇒ 00:10:09.390 Casie Aviles: It’s yeah in in this, we have a table here. And it’s it looks like this, it’s loading.
107 00:10:10.570 ⇒ 00:10:12.020 Casie Aviles: Yeah. So here.
108 00:10:12.633 ⇒ 00:10:18.420 Casie Aviles: yeah, we have. So you can see we have the the paths, the s. 3 path. We have the folder names
109 00:10:18.770 ⇒ 00:10:22.720 Casie Aviles: and the content. Actually, this is the actual raw transcript content.
110 00:10:22.720 ⇒ 00:10:23.480 Hannah Wang: Okay.
111 00:10:24.800 ⇒ 00:10:27.680 Casie Aviles: We have these paths, that
112 00:10:29.220 ⇒ 00:10:30.690 Hannah Wang: Oh, cool! Wow!
113 00:10:31.760 ⇒ 00:10:33.640 Casie Aviles: And all the participants.
114 00:10:33.640 ⇒ 00:10:42.090 Hannah Wang: I see and then for the agent. I I forgot if we already implemented this, but
115 00:10:42.420 ⇒ 00:10:47.340 Hannah Wang: does each with it? The agent within each client like, does that
116 00:10:47.690 ⇒ 00:10:51.440 Hannah Wang: like? For one meeting does like, if you click into the single.
117 00:10:51.570 ⇒ 00:10:56.780 Hannah Wang: if you go back to the platform, yeah. So within this meeting. Does this
118 00:10:56.980 ⇒ 00:11:03.370 Hannah Wang: Chatbot have context of all the other ABC meetings that we had in the past? Or is it just
119 00:11:04.110 ⇒ 00:11:07.700 Hannah Wang: side like just this one meeting, that it has context for.
120 00:11:09.340 ⇒ 00:11:15.009 Casie Aviles: Oh, okay, let me let me just verify this. Oh, I do know that we have.
121 00:11:15.942 ⇒ 00:11:20.900 Casie Aviles: Initially, our implementation was to have just one meeting
122 00:11:21.932 ⇒ 00:11:24.990 Casie Aviles: like the the agent just knows one meeting.
123 00:11:24.990 ⇒ 00:11:25.680 Hannah Wang: Yeah.
124 00:11:26.677 ⇒ 00:11:29.390 Casie Aviles: But right now, we’ve also implemented.
125 00:11:29.390 ⇒ 00:11:30.690 Casie Aviles: Oh, right? Yeah.
126 00:11:30.690 ⇒ 00:11:31.390 Casie Aviles: Hubs.
127 00:11:31.720 ⇒ 00:11:32.165 Hannah Wang: Right?
128 00:11:33.520 ⇒ 00:11:37.259 Hannah Wang: So that should technically have context for everything.
129 00:11:38.525 ⇒ 00:11:42.280 Casie Aviles: Yes, that’s we built it to have context for
130 00:11:42.650 ⇒ 00:11:45.630 Casie Aviles: also, like slack messages and.
131 00:11:45.630 ⇒ 00:11:46.040 Hannah Wang: Okay.
132 00:11:46.040 ⇒ 00:11:46.929 Casie Aviles: And zoom.
133 00:11:48.530 ⇒ 00:11:50.720 Hannah Wang: And does it have context for notion.
134 00:11:52.530 ⇒ 00:11:54.029 Casie Aviles: No, not at the moment.
135 00:11:54.030 ⇒ 00:11:54.380 Hannah Wang: Okay.
136 00:11:54.380 ⇒ 00:11:56.350 Casie Aviles: Just zoom and slack, for now.
137 00:11:56.350 ⇒ 00:12:04.440 Hannah Wang: Okay, okay, cool, okay, so
138 00:12:05.441 ⇒ 00:12:19.149 Hannah Wang: basically, the structure of the case study that we build out is there’s like 5 sections. So it’s context challenge solution results and impact. So I’m just gonna ask you questions to kind of get
139 00:12:19.550 ⇒ 00:12:32.966 Hannah Wang: the answers to each of those sections. So I know you already answered some of these things in the beginning, when you were giving me the overview. But I’m just gonna ask it, anyway. So yeah, what?
140 00:12:36.290 ⇒ 00:12:44.940 Hannah Wang: what was like the working environment before this project started. Like, I know, you mentioned that
141 00:12:45.440 ⇒ 00:12:53.679 Hannah Wang: getting information you just had to like go to Zoom, and then you had to like go to slack. So I guess was the issue that there wasn’t like a centralized
142 00:12:54.726 ⇒ 00:12:57.310 Hannah Wang: hub for all the information.
143 00:12:58.570 ⇒ 00:12:59.610 Casie Aviles: Yes, exactly.
144 00:13:00.250 ⇒ 00:13:06.348 Casie Aviles: People had to like. They have to go through and search things, and we don’t really have
145 00:13:07.020 ⇒ 00:13:11.399 Casie Aviles: the best documentation like, inform, like context can be scattered. So
146 00:13:11.760 ⇒ 00:13:16.929 Casie Aviles: yeah, I I that’s pretty much it like that was the main pain point, like
147 00:13:17.270 ⇒ 00:13:23.239 Casie Aviles: they had to go through a lot of other places just to look for what they needed to know.
148 00:13:24.110 ⇒ 00:13:28.279 Casie Aviles: Was kind of the reason why we built this to have it consolidated.
149 00:13:28.810 ⇒ 00:13:34.530 Hannah Wang: I see, and before you built the platform I guess. What were the other previous
150 00:13:34.730 ⇒ 00:13:41.289 Hannah Wang: efforts that you like try to do to solve the problem, like, I know, before the platform you had the
151 00:13:41.500 ⇒ 00:13:42.480 Hannah Wang: agent.
152 00:13:43.170 ⇒ 00:13:59.660 Hannah Wang: like each client agent within slack only, and you could like chat with them. I knew that was a thing was that like was that in an effort to kind of solve the problem that the platform is solving like was that kind of the previous attempt at solving this issue.
153 00:14:01.025 ⇒ 00:14:05.260 Casie Aviles: Yes, also. Yeah, that that was also one of our attempts
154 00:14:05.430 ⇒ 00:14:12.959 Casie Aviles: at solving the issue, although we ran into a lot of, you know, technical problems there. And
155 00:14:13.670 ⇒ 00:14:17.015 Casie Aviles: additionally, I think that the adoption was not very good.
156 00:14:17.830 ⇒ 00:14:21.040 Casie Aviles: And it wasn’t just slack based. Bots.
157 00:14:22.350 ⇒ 00:14:28.299 Casie Aviles: Yeah. And a lot of yeah, the the implementation wise. There’s a lot of things there that we were still figuring out, so
158 00:14:28.420 ⇒ 00:14:35.760 Casie Aviles: it didn’t produce the best quality. So we decided to just go when yeah, with the platform.
159 00:14:37.090 ⇒ 00:14:44.829 Hannah Wang: Hmm, I see and do you kinda know, like, why, I guess, like.
160 00:14:45.270 ⇒ 00:14:52.310 Hannah Wang: have people? Do you know the what specific challenges were like impacting.
161 00:14:53.000 ⇒ 00:14:57.280 Hannah Wang: Or I’m trying to re, I have like a list of questions I’m trying to like, rephrase it. So it makes sense.
162 00:14:57.280 ⇒ 00:14:57.610 Casie Aviles: That’s.
163 00:14:57.610 ⇒ 00:15:04.409 Hannah Wang: This context? I guess, like, what impact did
164 00:15:05.596 ⇒ 00:15:09.489 Hannah Wang: scattered knowledge and database have on like
165 00:15:10.602 ⇒ 00:15:12.899 Hannah Wang: the business, the business, or like
166 00:15:13.010 ⇒ 00:15:18.710 Hannah Wang: people within Brainforge trying to like search for things like what were some of the challenges that those
167 00:15:19.100 ⇒ 00:15:25.049 Hannah Wang: that arose from having scattered knowledge like, what type of frustrations or
168 00:15:25.658 ⇒ 00:15:28.199 Hannah Wang: problems where people are having.
169 00:15:28.720 ⇒ 00:15:35.500 Casie Aviles: Yeah, I? Well, I think, yeah, one of the problems, or like frustrations is
170 00:15:36.142 ⇒ 00:15:41.619 Casie Aviles: some. I think one of the things is like Uton becomes like a bottleneck to a lot.
171 00:15:41.620 ⇒ 00:15:42.050 Hannah Wang: There is no.
172 00:15:43.810 ⇒ 00:15:48.139 Casie Aviles: So people tend to. Just you know, keep pinging him or ask him.
173 00:15:48.240 ⇒ 00:15:50.020 Casie Aviles: And yeah,
174 00:15:51.550 ⇒ 00:16:01.600 Casie Aviles: yeah. And sometimes he’s not like he doesn’t have everything, doesn’t have all the context, and like it could have been solved if we had better like documentation on
175 00:16:03.003 ⇒ 00:16:06.720 Casie Aviles: you know, the the needed context or information. So
176 00:16:07.030 ⇒ 00:16:13.949 Casie Aviles: I think that was one of the core frustrations. And then, you know, like.
177 00:16:14.650 ⇒ 00:16:18.674 Casie Aviles: for for instance, like we without like the zoom thing
178 00:16:19.950 ⇒ 00:16:26.450 Casie Aviles: like with, I mean, with the zoom thing people could get answers much easily now. And
179 00:16:27.290 ⇒ 00:16:31.099 Casie Aviles: yeah, like, without that, it was just harder to
180 00:16:32.523 ⇒ 00:16:36.210 Casie Aviles: you know, go back? Or, yeah, basically. Yeah, that’s kind of it.
181 00:16:37.900 ⇒ 00:16:40.880 Hannah Wang: So what do you think would have happened if we didn’t build
182 00:16:41.150 ⇒ 00:16:46.979 Hannah Wang: this platform? Like, I, guess utam would still be the bottleneck and
183 00:16:47.170 ⇒ 00:16:52.450 Hannah Wang: work would just be a lot slower, I guess, like excel. We can’t accelerate in our work.
184 00:16:53.910 ⇒ 00:16:58.350 Casie Aviles: Yes, pretty much. Yeah. I think that would be like
185 00:16:58.480 ⇒ 00:17:00.513 Casie Aviles: that would still happen. And
186 00:17:01.820 ⇒ 00:17:07.969 Casie Aviles: It would still be difficult to like. You know, find the the stuff that needs to that people need.
187 00:17:08.250 ⇒ 00:17:08.960 Hannah Wang: Okay.
188 00:17:09.799 ⇒ 00:17:19.789 Hannah Wang: okay. And then the next section is solutions. So you already kind of walked through everything. One question I have is, how is this?
189 00:17:21.280 ⇒ 00:17:26.000 Hannah Wang: Well, I guess you already kind of answered it. But how is this platform different than
190 00:17:26.974 ⇒ 00:17:33.480 Hannah Wang: different than, and better than the slack agent? Bot thing that you built before.
191 00:17:36.590 ⇒ 00:17:41.336 Casie Aviles: Yeah, I think well, in terms of like
192 00:17:42.810 ⇒ 00:17:45.800 Casie Aviles: like its effect on the team
193 00:17:46.300 ⇒ 00:17:51.890 Casie Aviles: we’ve had. I feel like we’ve had, although we don’t really have. We don’t really track those yet, or like
194 00:17:52.280 ⇒ 00:17:55.900 Casie Aviles: have numbers. But people have been
195 00:17:56.330 ⇒ 00:17:59.160 Casie Aviles: I. I feel like adoption has been much better.
196 00:18:00.159 ⇒ 00:18:07.049 Casie Aviles: Compared to like before, and not entirely sure
197 00:18:07.580 ⇒ 00:18:10.429 Casie Aviles: why? But maybe because it’s also
198 00:18:11.080 ⇒ 00:18:13.830 Casie Aviles: it just works much better than before, like
199 00:18:15.177 ⇒ 00:18:20.069 Casie Aviles: we started I guess we did. We implemented it more
200 00:18:20.680 ⇒ 00:18:29.030 Casie Aviles: like we learned a bit from the mistakes that we have from the past implementation and kind of factored them into
201 00:18:29.360 ⇒ 00:18:30.560 Casie Aviles: building this.
202 00:18:32.970 ⇒ 00:18:39.525 Casie Aviles: And yeah, we’ve been getting more feedback. So that’s that’s why I think, adoption has been better. And
203 00:18:40.860 ⇒ 00:18:46.530 Casie Aviles: yeah, I think that’s pretty much what I have at the top of my head.
204 00:18:46.530 ⇒ 00:18:51.900 Hannah Wang: Okay? And who like? Why, I guess.
205 00:18:52.770 ⇒ 00:18:59.000 Hannah Wang: like what sparked the idea to go from that zoom, chat, bot to this platform
206 00:18:59.260 ⇒ 00:19:03.480 Hannah Wang: like, why, why did we pivot to building this platform?
207 00:19:03.690 ⇒ 00:19:19.130 Hannah Wang: Was it because it’s like more visual, like, there’s a u ui to it. There’s like an interface to it. Or yeah, like, what was the reasoning like? Who decided like, Oh, one day, instead of the zoom, chat, bot! Let’s start building out this platform and
208 00:19:19.370 ⇒ 00:19:20.110 Hannah Wang: like.
209 00:19:20.320 ⇒ 00:19:26.789 Hannah Wang: Why did why did they think that it would be better than the Zoom One, the the slack one sorry.
210 00:19:29.320 ⇒ 00:19:32.409 Casie Aviles: Trying to recall. But I think, yeah, this was
211 00:19:32.820 ⇒ 00:19:36.559 Casie Aviles: oh, Tom’s suggestion that maybe we should do this.
212 00:19:37.790 ⇒ 00:19:42.310 Casie Aviles: But yeah, I guess for me. I thought it also made sense, because
213 00:19:42.590 ⇒ 00:19:46.715 Casie Aviles: before, like the slack bots, they they they couldn’t be used
214 00:19:47.380 ⇒ 00:19:53.109 Casie Aviles: like you, you couldn’t. DM those bots. Maybe that’s 1 factor like, I guess people are
215 00:19:53.410 ⇒ 00:19:56.159 Casie Aviles: shy to use it. On.
216 00:19:56.160 ⇒ 00:19:58.510 Casie Aviles: Oh, yes, publicly. Yeah.
217 00:19:58.650 ⇒ 00:20:04.460 Casie Aviles: Yeah. So I guess, having it like here where?
218 00:20:05.007 ⇒ 00:20:08.189 Casie Aviles: It’s kind of, you know, more individual like
219 00:20:09.400 ⇒ 00:20:14.300 Casie Aviles: it’s not, you know. It’s not on the Channel. And it’s like it’s really just replicating
220 00:20:14.560 ⇒ 00:20:21.020 Casie Aviles: kind of replicating Chat Gpt, where they could just chat. So it’s just like the user. And just the AI.
221 00:20:21.560 ⇒ 00:20:26.080 Casie Aviles: So that’s 1 of the, I think one of the factors that I think
222 00:20:26.470 ⇒ 00:20:33.069 Casie Aviles: influenced. Why, I thought this would. This might be a better solution. And yeah, and I think, sorry.
223 00:20:34.380 ⇒ 00:20:36.850 Casie Aviles: Another. Yeah. I just remembered now like
224 00:20:37.130 ⇒ 00:20:41.639 Casie Aviles: have just using slack bots. Had a lot of limitations as well
225 00:20:43.480 ⇒ 00:20:50.389 Casie Aviles: like, for for instance, now with with slack bots we wanted to like, have multiple tools or like
226 00:20:50.680 ⇒ 00:21:04.459 Casie Aviles: other functionalities that we couldn’t just implement there. Like, for example, he, the the linear ticket creation, I mean, technically, we could try to get it working on the slack bots. But it’s just not the best.
227 00:21:04.690 ⇒ 00:21:06.990 Casie Aviles: I guess. Ux or.
228 00:21:07.190 ⇒ 00:21:07.820 Hannah Wang: Yeah.
229 00:21:08.530 ⇒ 00:21:16.560 Casie Aviles: Yeah, so there’s just a lot more flexibility and with with, you know, with having this interface.
230 00:21:16.860 ⇒ 00:21:17.590 Hannah Wang: Hmm!
231 00:21:18.410 ⇒ 00:21:25.580 Hannah Wang: That’s true. I mean, personally, for me, I was one of the shy people that didn’t want to embarrass myself using slack. So
232 00:21:25.990 ⇒ 00:21:36.160 Hannah Wang: I feel like. That’s why I’m using this platform a lot more, just because it’s more private, and I’m less afraid to quote unquote, look silly, or feel.
233 00:21:36.160 ⇒ 00:21:36.770 Casie Aviles: Yeah, like.
234 00:21:36.770 ⇒ 00:21:59.730 Hannah Wang: Kind of silly asking like questions. So, okay, yeah, that’s that’s helpful. And just so I know what tools, or I guess, like the back end tech stack that you use. So it’s obviously, zoom s 3 windmill super base. And Nan, I’m assuming that’s that’s Nan. Right? Yeah, okay.
235 00:21:59.730 ⇒ 00:22:00.450 Casie Aviles: Yes.
236 00:22:00.760 ⇒ 00:22:02.929 Hannah Wang: Was there anything else that you used.
237 00:22:05.100 ⇒ 00:22:09.020 Casie Aviles: Hmm, I we, I mean for the for the code
238 00:22:09.340 ⇒ 00:22:15.730 Casie Aviles: of this we use, we, we have a Github repository, and we create, react, we use react.
239 00:22:16.050 ⇒ 00:22:17.400 Hannah Wang: Okay. Yep.
240 00:22:18.010 ⇒ 00:22:21.390 Casie Aviles: Yeah, I think that’s pretty much all of it.
241 00:22:21.580 ⇒ 00:22:22.250 Hannah Wang: Okay,
242 00:22:25.620 ⇒ 00:22:31.113 Hannah Wang: And you said, I know you mentioned that we weren’t really tracking anything right now.
243 00:22:32.660 ⇒ 00:22:33.450 Casie Aviles: Yeah.
244 00:22:33.450 ⇒ 00:22:42.719 Hannah Wang: Yeah, I guess the results section that’s usually we want numbers. Is there any number you can provide for me like.
245 00:22:43.070 ⇒ 00:22:48.451 Hannah Wang: I know, adoption rate, we don’t really know. Yeah, cause we’re not like tracking much.
246 00:22:50.790 ⇒ 00:22:56.859 Casie Aviles: Yeah, I’m i i don’t think we have any any anything to report like in terms of like the numbers.
247 00:22:56.860 ⇒ 00:22:57.220 Hannah Wang: Okay.
248 00:22:57.220 ⇒ 00:23:01.489 Casie Aviles: I mean, that’s something we want to implement in the future. But.
249 00:23:03.500 ⇒ 00:23:15.959 Hannah Wang: okay, no worries. And then the feedback that you were getting, like, I, I use it and I, I think it’s very helpful. Do you know, if other people are kind of giving feedback about the platform.
250 00:23:17.160 ⇒ 00:23:17.915 Casie Aviles: Yes,
251 00:23:18.920 ⇒ 00:23:24.689 Casie Aviles: So far I’ve gotten. I know that amber Kyle.
252 00:23:25.380 ⇒ 00:23:28.919 Casie Aviles: the milade awaysh have been using this.
253 00:23:30.550 ⇒ 00:23:37.650 Casie Aviles: and yes, some of them have been vocal about how helpful this is, but and other than that, of course they also have, like
254 00:23:37.770 ⇒ 00:23:44.000 Casie Aviles: feature requests or things. They wish the platform supported, or
255 00:23:44.170 ⇒ 00:23:45.850 Casie Aviles: things they wish they could do.
256 00:23:46.000 ⇒ 00:23:54.120 Casie Aviles: And yeah, I think overall, that’s like a positive. I would say.
257 00:23:54.380 ⇒ 00:23:54.900 Hannah Wang: Yeah.
258 00:23:56.180 ⇒ 00:24:00.200 Casie Aviles: But yeah, I I that’s pretty much all I have, all I know from.
259 00:24:00.200 ⇒ 00:24:00.840 Hannah Wang: Okay.
260 00:24:01.330 ⇒ 00:24:01.920 Casie Aviles: Yeah.
261 00:24:02.640 ⇒ 00:24:04.246 Hannah Wang: Okay, cool.
262 00:24:07.070 ⇒ 00:24:18.209 Hannah Wang: I mean, how long did this project take? I know you had, like a previous iteration of like a demo hub. But approximately. How long do you think this project took.
263 00:24:21.240 ⇒ 00:24:27.470 Casie Aviles: I’m trying to recall, because the zoom work I started
264 00:24:27.760 ⇒ 00:24:32.110 Casie Aviles: like I think that was in January or February.
265 00:24:33.690 ⇒ 00:24:37.350 Casie Aviles: And we didn’t have like the platform yet. It was just the summaries.
266 00:24:37.690 ⇒ 00:24:38.390 Hannah Wang: -
267 00:24:39.895 ⇒ 00:24:43.950 Casie Aviles: And that I think that took me around a month.
268 00:24:44.430 ⇒ 00:24:44.770 Hannah Wang: Okay.
269 00:24:44.770 ⇒ 00:24:47.370 Casie Aviles: And it was just me and Otam working on it.
270 00:24:47.490 ⇒ 00:24:48.830 Hannah Wang: And then.
271 00:24:49.310 ⇒ 00:24:57.160 Casie Aviles: For the platform stuff. I think we started this June. I believe.
272 00:24:57.360 ⇒ 00:24:57.960 Hannah Wang: Yeah.
273 00:24:58.990 ⇒ 00:25:09.549 Hannah Wang: So around 3 months, 3 to 4 months of work. Well, I know it was like iterative, like, you worked on certain things before. And then you just built up on it. But yeah, I guess
274 00:25:10.200 ⇒ 00:25:12.880 Hannah Wang: I can say, like 3 to 4 months. Okay,
275 00:25:14.080 ⇒ 00:25:16.020 Hannah Wang: then, we kind of paused on it
276 00:25:16.820 ⇒ 00:25:17.680 Casie Aviles: But yeah.
277 00:25:18.950 ⇒ 00:25:27.149 Hannah Wang: Okay, I’ll just say like, like 6 months or something. I don’t know. It’s like a
278 00:25:27.460 ⇒ 00:25:32.240 Hannah Wang: because I feel like, you know, not everything was linear and.
279 00:25:32.240 ⇒ 00:25:32.870 Casie Aviles: Yeah.
280 00:25:32.870 ⇒ 00:25:48.872 Hannah Wang: It was a lot of like back and forth, pausing and then like restarting. So I feel like, probably you probably technically worked on this for, like the entire year, like of 2025. But like you just like, pause here and there, which is okay.
281 00:25:49.840 ⇒ 00:25:54.780 Hannah Wang: okay. And let’s see what else.
282 00:25:58.180 ⇒ 00:26:00.599 Hannah Wang: Would you say adoption was pretty quick
283 00:26:00.720 ⇒ 00:26:05.720 Hannah Wang: for this like obviously quicker than the slack one, right.
284 00:26:06.690 ⇒ 00:26:22.879 Casie Aviles: Yeah, I would say so. Because before, like, it was really hard to get people using those slack bots before and now I think more people were open to using this, and it was easier for them to just use it. So yeah.
285 00:26:23.200 ⇒ 00:26:38.029 Hannah Wang: Okay? And do you have, like, I know, post how you can use it to track like website, how many people like visit the website. Do we have anything in the back end that’s set up for the platform, like how many times people visited
286 00:26:39.570 ⇒ 00:26:40.990 Hannah Wang: this domain.
287 00:26:41.190 ⇒ 00:26:42.139 Casie Aviles: Not yet, but.
288 00:26:42.140 ⇒ 00:26:42.490 Hannah Wang: Okay.
289 00:26:42.490 ⇒ 00:26:45.610 Casie Aviles: Also part of our backlog.
290 00:26:45.960 ⇒ 00:26:46.710 Hannah Wang: Okay.
291 00:26:48.050 ⇒ 00:27:04.603 Hannah Wang: yeah, that’d be interesting to see. Just like how many I mean, a lot of I don’t know. I visit this website pretty often. So I don’t know. Obviously that’ll skew everything. And you probably need to look at it per user instead of like per session or something. But that’s a later problem.
292 00:27:05.410 ⇒ 00:27:10.700 Hannah Wang: okay, cool. I think that’s everything that I need. And then.
293 00:27:11.213 ⇒ 00:27:19.146 Hannah Wang: I guess later on I’ll like, create a rough draft of the case study, and then I’ll ping you and Utam just to make sure everything’s correct.
294 00:27:19.440 ⇒ 00:27:20.030 Casie Aviles: Okay.
295 00:27:20.440 ⇒ 00:27:33.759 Hannah Wang: That I I didn’t like AI didn’t hallucinate anything. And then, yeah, I’ll go from there. This is super helpful. Thank you so much and any other project that you guys do
296 00:27:34.630 ⇒ 00:27:36.170 Hannah Wang: it’d be helpful to like.
297 00:27:36.460 ⇒ 00:27:55.379 Hannah Wang: let me know, I guess, like, oh, this is a project we did for a client or for our internal team just so that we can make case studies. And I can just like interview you again for 30 min, and then, like, make a case study out of it. So yeah, just let me know. Like, whenever you do a cool project cause I feel like you guys are always
298 00:27:55.540 ⇒ 00:27:58.440 Hannah Wang: doing cool things. So it’s awesome.
299 00:27:59.950 ⇒ 00:28:01.400 Casie Aviles: Okay. Yeah. Sure.
300 00:28:01.400 ⇒ 00:28:05.329 Hannah Wang: Cool. Alright. Thank you so much, Casey. Have a good day.
301 00:28:05.700 ⇒ 00:28:06.639 Casie Aviles: Thank you, Anna.
302 00:28:06.640 ⇒ 00:28:07.250 Hannah Wang: Thank you. Bye.