Meeting Title: Brainforge AI Team Introduction and Sync Date: 2025-11-14 Meeting participants: Casie Aviles, Joseph Good
WEBVTT
1 00:00:12.450 ⇒ 00:00:14.199 Joseph Good: Hey, Casey, how’s it going?
2 00:00:14.870 ⇒ 00:00:16.210 Casie Aviles: Hey, hey, Joseph.
3 00:00:16.680 ⇒ 00:00:18.389 Casie Aviles: Doing good. How about you?
4 00:00:19.220 ⇒ 00:00:23.530 Joseph Good: Good, good. You’re, calling from the Philippines, or where are you?
5 00:00:24.210 ⇒ 00:00:26.060 Casie Aviles: Yeah, Philippines.
6 00:00:27.010 ⇒ 00:00:31.029 Joseph Good: Nice. How’s your… are you staying safe out there? I know you said there was a typhoon going on.
7 00:00:31.530 ⇒ 00:00:36.039 Casie Aviles: Oh, yeah, yeah, fortunately, we’re safe,
8 00:00:36.140 ⇒ 00:00:40.600 Casie Aviles: But it was a pretty strong typhoon. It was a super typhoon, I believe.
9 00:00:40.810 ⇒ 00:00:47.059 Casie Aviles: And yet, we, we didn’t have any power for, for, like, 4 days, so…
10 00:00:47.330 ⇒ 00:00:51.499 Casie Aviles: But now we’re… it’s got… it’s back up, so… yeah, that’s great.
11 00:00:51.620 ⇒ 00:00:52.280 Casie Aviles: Yeah.
12 00:00:52.280 ⇒ 00:00:53.060 Joseph Good: Okay.
13 00:00:53.540 ⇒ 00:00:56.290 Joseph Good: Well, glad to hear you’re safe.
14 00:00:56.770 ⇒ 00:01:01.070 Joseph Good: And yeah, good to meet you. I just joined the team…
15 00:01:01.540 ⇒ 00:01:07.550 Joseph Good: A couple weeks ago, kind of helping out on the go-to-market side of things with
16 00:01:07.980 ⇒ 00:01:20.660 Joseph Good: Robert, and then I’ve met Utam a few times as well, and I’ve chatted with Mustafa earlier this week, and hopefully Sam next week. So, just kind of getting up to speed on
17 00:01:20.900 ⇒ 00:01:24.570 Joseph Good: Everything that you and the team are doing.
18 00:01:25.130 ⇒ 00:01:30.520 Joseph Good: I do handle more of the sales side, but, I’m definitely interested in just…
19 00:01:30.580 ⇒ 00:01:45.360 Joseph Good: learning more about the technical side, and I mean, hopefully eventually helping out with some stuff on that front. So, kind of wanted to just get a sense of, what your day-to-day looks like, the types of projects that you work on,
20 00:01:45.510 ⇒ 00:01:47.880 Joseph Good: And yeah, kind of, like, what,
21 00:01:48.150 ⇒ 00:01:50.989 Joseph Good: what tools you use, and all that good stuff, so…
22 00:01:52.190 ⇒ 00:01:56.120 Casie Aviles: Sure, yeah, so primarily I…
23 00:01:56.900 ⇒ 00:02:02.019 Casie Aviles: I’m part of… yeah, I’m a part of the AI team, so what I usually do is…
24 00:02:02.630 ⇒ 00:02:07.290 Casie Aviles: I work on, like, clients and also the internal work that we have.
25 00:02:08.720 ⇒ 00:02:11.920 Casie Aviles: So, for the client work, we…
26 00:02:12.300 ⇒ 00:02:19.530 Casie Aviles: have, like, ABC and also Insomnia, at least from my end. Those are the clients I’m working on.
27 00:02:20.030 ⇒ 00:02:28.040 Casie Aviles: And the solutions we’re building for them are… for ABC, we have, like, an AI chatbot for them.
28 00:02:28.220 ⇒ 00:02:34.340 Casie Aviles: that… Draw, like, gets from, like, a custom knowledge base.
29 00:02:36.640 ⇒ 00:02:39.580 Casie Aviles: Yeah, and then you’re just really helping them
30 00:02:39.700 ⇒ 00:02:45.740 Casie Aviles: Helping their customer service representatives in order to, like, you know.
31 00:02:46.060 ⇒ 00:02:55.150 Casie Aviles: answer, like, the questions from customers. They would usually be on the call, and they would use the chatbot to help them answer.
32 00:02:55.400 ⇒ 00:03:01.279 Casie Aviles: And, yeah, yeah. And then, I guess for Insomnia, what I’m just doing there is more on, like.
33 00:03:01.740 ⇒ 00:03:04.530 Casie Aviles: Automations, so less of AI work.
34 00:03:07.050 ⇒ 00:03:13.099 Casie Aviles: Yeah, it’s more of just getting, like, data from certain… sources, like.
35 00:03:13.390 ⇒ 00:03:18.150 Casie Aviles: They have marketing data that they need to report on, so I’m helping them
36 00:03:18.990 ⇒ 00:03:22.589 Casie Aviles: Get all of that into one spreadsheet.
37 00:03:24.830 ⇒ 00:03:32.499 Casie Aviles: And yeah, so… I guess internally, we’re also building a lot of stuff that we have, like automations that we have. I’ve helped
38 00:03:33.160 ⇒ 00:03:38.220 Casie Aviles: work on, like, I think later on in the demo, we’ll be early in the meeting.
39 00:03:38.510 ⇒ 00:03:40.070 Casie Aviles: We’ll be demoing…
40 00:03:40.470 ⇒ 00:03:46.020 Casie Aviles: A platform… the internal platform that we use, like a case study assistant for the marketing team.
41 00:03:48.820 ⇒ 00:03:56.259 Casie Aviles: Yeah, so I think that’s about it. And also, like, I’m not sure if you had the chance to use, like, the platform
42 00:03:56.760 ⇒ 00:04:03.290 Casie Aviles: for Brainforge, where we… we kind of consolidated, like, all of the Zoom
43 00:04:03.980 ⇒ 00:04:11.450 Casie Aviles: Meetings and transcripts that we have there, and… Yeah, we’re continuously… expanding the…
44 00:04:11.820 ⇒ 00:04:15.799 Casie Aviles: Features that that platform has in order to support, like.
45 00:04:16.839 ⇒ 00:04:21.089 Casie Aviles: The team, so we have, like, marketing-related
46 00:04:23.140 ⇒ 00:04:27.820 Casie Aviles: Tools, and then project management tools, so we’re trying to…
47 00:04:28.470 ⇒ 00:04:33.180 Casie Aviles: Make them more efficient and get people up to speed much faster, so…
48 00:04:33.510 ⇒ 00:04:35.939 Casie Aviles: Yeah, I think that’s pretty much,
49 00:04:36.690 ⇒ 00:04:40.749 Casie Aviles: what I’ve been working on for Brainforge,
50 00:04:40.960 ⇒ 00:04:43.609 Casie Aviles: So I’ve been… I’ve been with Brainforge, I think.
51 00:04:44.180 ⇒ 00:04:46.760 Casie Aviles: Last… November.
52 00:04:47.160 ⇒ 00:04:48.559 Casie Aviles: Yeah, last year.
53 00:04:48.800 ⇒ 00:04:51.070 Casie Aviles: So, I… I guess, yeah.
54 00:04:51.460 ⇒ 00:04:52.649 Casie Aviles: That’s about it.
55 00:04:53.720 ⇒ 00:04:56.089 Joseph Good: Yeah, that makes sense, that’s great.
56 00:04:56.730 ⇒ 00:04:59.499 Joseph Good: Okay, and what… what was the…
57 00:05:00.110 ⇒ 00:05:08.630 Joseph Good: data, I think, automation stuff, you said. I know Mustafa showed me some stuff in NAN and Subabase,
58 00:05:08.730 ⇒ 00:05:11.829 Joseph Good: But, yeah, what does the automation kind of look like for you?
59 00:05:13.310 ⇒ 00:05:18.319 Casie Aviles: Is it, like… For client work, or for, like, internal work?
60 00:05:19.330 ⇒ 00:05:32.480 Joseph Good: For the client at work, mainly, I think you said you were, like… I think you mentioned that you were helping to get one client, maybe for Insomniac, get some data, like, into Sheets or something like that. Is that… is that right?
61 00:05:32.680 ⇒ 00:05:37.789 Casie Aviles: Hmm, yeah, yeah, okay. I can, I can share a little bit about What that looks like.
62 00:05:38.220 ⇒ 00:05:38.770 Joseph Good: Yeah.
63 00:05:40.890 ⇒ 00:05:41.820 Casie Aviles: Okay.
64 00:05:42.020 ⇒ 00:05:43.010 Casie Aviles: Jess…
65 00:05:48.210 ⇒ 00:05:48.960 Casie Aviles: Alright.
66 00:05:49.100 ⇒ 00:05:51.360 Casie Aviles: You can see my screen right now.
67 00:05:52.060 ⇒ 00:05:52.620 Joseph Good: Yeah.
68 00:05:54.720 ⇒ 00:06:03.999 Casie Aviles: Yeah, so, it’s not… it’s not, there’s not a lot to… it’s not a very fancy work that we have here, not a lot of… it’s just a lot of data.
69 00:06:04.730 ⇒ 00:06:11.010 Casie Aviles: But, basically, we’re just… Automating. We have, like, it’s just a bunch of spreadsheets, and…
70 00:06:12.250 ⇒ 00:06:16.690 Casie Aviles: Where we’re, like, getting the data from… For example, Braze.
71 00:06:16.990 ⇒ 00:06:23.429 Casie Aviles: And so that’s, like, where they… where we get the data for their marketing… owned marketing, so…
72 00:06:23.870 ⇒ 00:06:30.250 Casie Aviles: We get the revenue of how much, like, their email campaign made, how many messages, so for example, this is…
73 00:06:31.030 ⇒ 00:06:39.719 Casie Aviles: what we’re seeing right now, this is, like, the result or the output of, like, the automated poll that we do every day, and…
74 00:06:40.940 ⇒ 00:06:45.980 Casie Aviles: They… they… we generate, like, we have… they have, like, this scorecard that Kind of.
75 00:06:46.270 ⇒ 00:06:51.630 Casie Aviles: Sums up their revenue, and then… Also compares, like.
76 00:06:52.190 ⇒ 00:06:55.720 Casie Aviles: The la- this year versus last year, so…
77 00:06:56.990 ⇒ 00:07:02.619 Casie Aviles: I think that’s pretty much what we’re trying to do with them, and then we also have, like, other pieces, like DoorDash.
78 00:07:03.250 ⇒ 00:07:07.550 Casie Aviles: And then Uber Eats, we’re also, scraping data from those.
79 00:07:08.710 ⇒ 00:07:10.130 Casie Aviles: platforms.
80 00:07:10.750 ⇒ 00:07:15.729 Casie Aviles: So, yeah, it’s just a lot of numbers, but there’s, like, an automation that’s happening.
81 00:07:17.820 ⇒ 00:07:24.100 Casie Aviles: Behind everything… behind all of this. So, it’s just code, really, that gets scheduled, like, every day.
82 00:07:25.490 ⇒ 00:07:31.490 Joseph Good: Okay, and what’s the… is the automation added in, or is that, like, a custom script, or what is… what does that look like?
83 00:07:32.680 ⇒ 00:07:39.370 Casie Aviles: Yeah, so for this particular client, we are using, like, custom Python scripts.
84 00:07:41.830 ⇒ 00:07:51.159 Casie Aviles: Yeah, we have them… we use, like, this… Orchestration… tool. I’ll dig stir on.
85 00:07:52.320 ⇒ 00:07:56.470 Casie Aviles: I’m not logged in right now, but basically, this is where we,
86 00:07:58.070 ⇒ 00:08:03.110 Casie Aviles: Where we schedule… this is how we schedule, like, the Python scripts that we have.
87 00:08:03.920 ⇒ 00:08:10.839 Casie Aviles: And Yeah, that’s pretty much where it lives, although… We also use, like, N8N…
88 00:08:10.940 ⇒ 00:08:16.910 Casie Aviles: for, like, more simple automations. The reason why we’re using, like, custom code here is because
89 00:08:17.780 ⇒ 00:08:23.870 Casie Aviles: we… there’s not, like, really native integrations with NHN. It’s not easy to, like, get
90 00:08:24.570 ⇒ 00:08:32.370 Casie Aviles: the data from these platforms into N8N, so we had to use, like, something more custom.
91 00:08:34.620 ⇒ 00:08:42.260 Joseph Good: And how does the… okay, cool. I don’t really know how Dagstra works. Can you kind of explain how you’re doing stuff in Dagstra?
92 00:08:43.289 ⇒ 00:08:48.499 Casie Aviles: Yeah, so, in essence, it just lets us schedule, you know, and
93 00:08:49.079 ⇒ 00:08:52.539 Casie Aviles: create this pipeline, so it’s just Python underneath.
94 00:08:53.559 ⇒ 00:08:56.739 Casie Aviles: It’s just letting us, allows us to schedule
95 00:08:57.139 ⇒ 00:08:59.839 Casie Aviles: Like, for example, we have all of these,
96 00:09:01.359 ⇒ 00:09:08.279 Casie Aviles: jobs that, we can run, like, daily, so that’s pretty much it. It’s, like, just scheduling
97 00:09:10.119 ⇒ 00:09:11.779 Casie Aviles: Scheduling for us.
98 00:09:12.099 ⇒ 00:09:15.689 Casie Aviles: That’s pretty much it. And then also deploying the scripts.
99 00:09:17.229 ⇒ 00:09:22.509 Casie Aviles: So instead of, like, running them locally on our own computers or machines.
100 00:09:22.789 ⇒ 00:09:26.149 Casie Aviles: it’s going to be on Dagster, so it will run…
101 00:09:28.430 ⇒ 00:09:38.180 Joseph Good: Okay. Sorry, this is all new to me, but I’m definitely interested in learning more about it, so, excuse me if I ask kind of silly questions, but…
102 00:09:38.180 ⇒ 00:09:39.830 Casie Aviles: Yeah, yeah. Is this…
103 00:09:40.020 ⇒ 00:09:41.100 Joseph Good: So…
104 00:09:41.240 ⇒ 00:09:51.670 Joseph Good: the data is going from Braze, and then is it going into Dagster, or like… and then it’s going back to the sheet, or what’s the flow of information here?
105 00:09:52.050 ⇒ 00:09:56.129 Casie Aviles: Yeah, yeah, that’s… Yeah, that’s pretty much how it works, like…
106 00:09:56.900 ⇒ 00:10:00.390 Casie Aviles: To kind of help you visualize…
107 00:10:02.150 ⇒ 00:10:07.709 Casie Aviles: It’s just really, like, the source would just be, like, Dagster, and then…
108 00:10:08.550 ⇒ 00:10:18.439 Casie Aviles: Or, sorry, sorry, I mean Braze, and then it will go through Dagster, and then it’ll end up in a spreadsheet. That’s pretty much how it works, yeah, it’s just, like, a three-step process.
109 00:10:19.670 ⇒ 00:10:25.099 Casie Aviles: Daxter helps us… Kind of get, yeah, it’s like a pipeline for getting the data.
110 00:10:27.280 ⇒ 00:10:34.949 Joseph Good: Okay, is that actually… so do you guys visually map out the data flow for every project?
111 00:10:36.170 ⇒ 00:10:37.220 Joseph Good: In Figma.
112 00:10:39.360 ⇒ 00:10:50.769 Casie Aviles: Yeah, we tried to… we tried to… I mean, for Insomnia, we didn’t do it at first, but then we had to, like, do a review, an architecture review, so we… eventually, we had to create
113 00:10:51.110 ⇒ 00:10:55.549 Casie Aviles: like, a diagram on Figma, so… Yeah, kind of.
114 00:10:55.840 ⇒ 00:11:04.199 Casie Aviles: This is a little… still… there’s a lot going on still, but as you can see, like, for example, we have Braze, and it’s just gonna get…
115 00:11:05.330 ⇒ 00:11:09.869 Casie Aviles: I’m just gonna go to… Dagster over here.
116 00:11:10.170 ⇒ 00:11:19.099 Casie Aviles: And then it’s gonna send it to, like, this spreadsheet that we have. There are other spreadsheets as well that end up… that are on the client side, but…
117 00:11:19.350 ⇒ 00:11:25.950 Casie Aviles: this is, like, the main spreadsheet for now that we’re routing all the data, so we also have, like, DoorDash.
118 00:11:27.210 ⇒ 00:11:32.110 Casie Aviles: Uber… We also have Google Ads, so that’s kind of what we’re doing.
119 00:11:32.710 ⇒ 00:11:34.859 Casie Aviles: for, Insomniac or Peace.
120 00:11:36.100 ⇒ 00:11:42.059 Joseph Good: Got it, okay. Would you actually mind sending me this Figma? This would be super interesting to look at.
121 00:11:42.330 ⇒ 00:11:43.710 Joseph Good: Okay, sure. If you don’t mind.
122 00:11:45.610 ⇒ 00:11:46.679 Casie Aviles: Let me see if…
123 00:11:48.830 ⇒ 00:11:49.520 Joseph Good: Yeah.
124 00:11:50.420 ⇒ 00:11:56.760 Joseph Good: Because I’m just trying to get a better sense of the architecture here, and the sort of system design. I’m not too familiar with that.
125 00:11:58.070 ⇒ 00:11:58.670 Casie Aviles: Okay.
126 00:12:00.710 ⇒ 00:12:05.960 Joseph Good: Okay, that might… oh, thanks for assuming that. Okay, that makes sense.
127 00:12:06.650 ⇒ 00:12:11.490 Joseph Good: Got it.
128 00:12:11.900 ⇒ 00:12:18.970 Joseph Good: So, Dagster, Naden, Superbase, and then… what are the other, kind of, tools that you use in your…
129 00:12:19.430 ⇒ 00:12:22.840 Joseph Good: workflow. Are there other tools that you guys are using?
130 00:12:23.630 ⇒ 00:12:27.440 Casie Aviles: Yeah, we use cursor a lot, so…
131 00:12:28.190 ⇒ 00:12:35.509 Casie Aviles: we’re actually building, like I mentioned, we were working on, like, an additional assistant.
132 00:12:35.660 ⇒ 00:12:38.739 Casie Aviles: On the… our Brainforge platform, so…
133 00:12:39.010 ⇒ 00:12:46.090 Casie Aviles: We’re using Courser a lot. Basically, it’s like a… it’s like a code editor that has AI built into it.
134 00:12:46.450 ⇒ 00:12:52.950 Casie Aviles: So, like, we basically just, you know, chat with Over here, like, we chat.
135 00:12:53.100 ⇒ 00:12:59.340 Casie Aviles: with the AI, and it’s gonna help us, like, make changes to the code, and it’s just,
136 00:12:59.660 ⇒ 00:13:03.649 Casie Aviles: Basically, makes it faster for us to ship stuff.
137 00:13:03.990 ⇒ 00:13:06.150 Casie Aviles: So we use this a lot.
138 00:13:06.250 ⇒ 00:13:12.110 Casie Aviles: For, like… Yeah, for… in general, for, like, even, like, for the…
139 00:13:12.320 ⇒ 00:13:17.499 Casie Aviles: Dagster stuff that I worked on. I also use Cursor a lot, so the team uses this a lot.
140 00:13:18.260 ⇒ 00:13:23.250 Joseph Good: Yeah, no, I’ve heard of Chrysler, I’ve heard it’s great. Is the…
141 00:13:24.240 ⇒ 00:13:31.300 Joseph Good: Is the repository for the internal BrainForge platform on GitHub, or where is that stored?
142 00:13:31.530 ⇒ 00:13:33.430 Casie Aviles: Yes, it’s on GitHub.
143 00:13:34.980 ⇒ 00:13:41.049 Casie Aviles: Yeah, as you can see here, this is, like, the repository that we have for the Brainforge platform.
144 00:13:42.160 ⇒ 00:13:43.850 Joseph Good: Okay.
145 00:13:44.720 ⇒ 00:13:47.919 Casie Aviles: And then we have a bunch of other repos here as well.
146 00:13:48.670 ⇒ 00:13:54.960 Joseph Good: Got it, okay, I see. Is that… I would love to, just because I’m trying to…
147 00:13:55.760 ⇒ 00:13:59.370 Joseph Good: frankly, learn how to code better.
148 00:13:59.550 ⇒ 00:14:05.030 Joseph Good: Is that GitHub something I can see, or… I would just love to poke around and see what you guys have?
149 00:14:07.050 ⇒ 00:14:09.960 Casie Aviles: I’m not sure if you’re added here.
150 00:14:10.610 ⇒ 00:14:14.139 Casie Aviles: I think… I can’t add people either, but…
151 00:14:15.190 ⇒ 00:14:15.950 Joseph Good: That’s okay, I’ll…
152 00:14:15.950 ⇒ 00:14:16.580 Casie Aviles: Austin.
153 00:14:17.400 ⇒ 00:14:22.450 Joseph Good: Yeah, or whatever you can send me, and then maybe I can ask Robert, or UTam, or whatnot.
154 00:14:24.480 ⇒ 00:14:26.129 Joseph Good: Okay, great, yeah.
155 00:14:27.470 ⇒ 00:14:38.630 Joseph Good: See if this is… Yeah, I’ll, I’ll ask, futon or whatnot, but… Okay, that makes sense.
156 00:14:39.870 ⇒ 00:14:41.200 Joseph Good: Cool, well…
157 00:14:46.000 ⇒ 00:14:48.680 Joseph Good: Great, yeah, I think that was kind of most of the questions.
158 00:14:48.780 ⇒ 00:14:54.099 Joseph Good: That I had, but, I mean, did you have any questions for me, or any ways that I can be helpful to you?
159 00:14:57.310 ⇒ 00:14:57.970 Casie Aviles: Yeah,
160 00:14:59.020 ⇒ 00:15:06.930 Casie Aviles: I don’t have any questions at the top of my mind, but… but you did mention that you’re helping us with the GTM side, right?
161 00:15:08.510 ⇒ 00:15:09.129 Casie Aviles: Okay, cool.
162 00:15:09.130 ⇒ 00:15:15.239 Joseph Good: And then, the second thing is just, like, for positioning, and kind of…
163 00:15:15.460 ⇒ 00:15:18.120 Joseph Good: talking about our AI services, and sort of how.
164 00:15:18.120 ⇒ 00:15:18.700 Casie Aviles: We talked about.
165 00:15:18.700 ⇒ 00:15:24.750 Joseph Good: about that in the market, like, I think Robert would like me to help with that as well, so…
166 00:15:25.060 ⇒ 00:15:32.140 Joseph Good: I’m just trying to figure out, you know, a lot of the awesome work that you and Mestapa and Sam are doing, like, how can we best, kind of.
167 00:15:32.480 ⇒ 00:15:40.300 Casie Aviles: Show that work and talk about it to potential clients, so… Okay, yeah, cool.
168 00:15:41.190 ⇒ 00:15:50.519 Casie Aviles: Yeah, I think… I think that’s pretty much it that I have, but if I have any questions, I’m… yeah, I’ll just Slack you, and yeah.
169 00:15:50.700 ⇒ 00:15:51.660 Casie Aviles: That’s it.
170 00:15:52.380 ⇒ 00:15:54.080 Joseph Good: Yeah, definitely.
171 00:15:54.350 ⇒ 00:15:59.910 Joseph Good: Well, great. Well, thanks, Casey. I really appreciate the time. Hopefully we’ll be chatting again soon.
172 00:16:00.850 ⇒ 00:16:03.489 Casie Aviles: Sure. Thank you, Joseph, for the time as well.
173 00:16:04.010 ⇒ 00:16:05.270 Joseph Good: Okay, bye.
174 00:16:05.460 ⇒ 00:16:06.210 Casie Aviles: Bye-bye.