Meeting Title: Brainforge_Saasholic_Sales Agent Demo Date: 2024-12-09 Meeting participants: Uttam Kumaran, Robert Tseng, Miguel De Veyra, Diego Gomes
WEBVTT
1 00:08:16.430 ⇒ 00:08:18.190 Diego Gomes: Hi! Hello!
2 00:08:23.410 ⇒ 00:08:24.360 Diego Gomes: Hello!
3 00:08:24.360 ⇒ 00:08:25.169 Uttam Kumaran: Can you hear me?
4 00:08:25.710 ⇒ 00:08:27.679 Diego Gomes: Hi! Good to see you again.
5 00:08:27.680 ⇒ 00:08:29.259 Uttam Kumaran: Hey? Good to see you! How are you?
6 00:08:29.700 ⇒ 00:08:36.079 Diego Gomes: Very good, very good. Thanks for taking the time, and sorry for the last minute change.
7 00:08:36.080 ⇒ 00:08:44.959 Uttam Kumaran: No, no, no problem, thank you. And thanks. I know it’s it was holidays here in the States. So it’s always like one week of like so much family stuff that
8 00:08:45.480 ⇒ 00:08:54.349 Uttam Kumaran: it’s like hard to book anything. So yeah, I’m super super excited to kind of share. What we got done let me just
9 00:08:54.690 ⇒ 00:08:56.800 Uttam Kumaran: confirm. We have everybody here.
10 00:08:57.590 ⇒ 00:09:02.100 Diego Gomes: Perfect. Just one question. This meeting is being recorded right.
11 00:09:02.100 ⇒ 00:09:02.890 Uttam Kumaran: Yes.
12 00:09:03.320 ⇒ 00:09:09.299 Diego Gomes: Is it okay to share the video recording afterwards? So I can share to my partners as well.
13 00:09:09.300 ⇒ 00:09:11.200 Uttam Kumaran: 100%. Yeah, yeah. Happy to.
14 00:09:11.600 ⇒ 00:09:12.710 Diego Gomes: Thank you, appreciate it.
15 00:09:13.020 ⇒ 00:09:17.379 Uttam Kumaran: We just recorded for internally, actually mainly just to keep up with notes and everything.
16 00:09:17.380 ⇒ 00:09:19.880 Diego Gomes: We love. It makes life a lot easier.
17 00:09:20.422 ⇒ 00:09:22.550 Uttam Kumaran: Yeah. Okay. Okay. Cool.
18 00:09:22.550 ⇒ 00:09:25.369 Diego Gomes: People do not like we, we love it.
19 00:09:25.560 ⇒ 00:09:30.179 Uttam Kumaran: Yeah, I mean, look, I think part of it. I get the security. But at the same time
20 00:09:30.440 ⇒ 00:09:36.720 Uttam Kumaran: we’re in so many meetings, and to keep track of everything is impossible. What’s the pro and con like.
21 00:09:36.770 ⇒ 00:09:50.459 Uttam Kumaran: you know? And I don’t know I could. I’m happy to pause, or what, but that. That’s the thing. It’s like some stuff, you, the technology so strong that you want to just use everything. So we we take all of our meetings and we have a we have an AI bot that we can ask questions to about like.
22 00:09:50.460 ⇒ 00:10:12.029 Uttam Kumaran: what do we talk in this meeting? Everything. But let me. Let me just jump into what we did, and Miguel on my team is here. He leads sort of our AI engineering and and him and his team. They did a great job kind of building this proof of concept. And then Robert who just joined as well. He’s a partner here at Brain Forge as well, and just kind of jump into things. Everybody here is
23 00:10:12.060 ⇒ 00:10:16.820 Uttam Kumaran: understands kind of like what we’re doing. So let me just share this and I’ll
24 00:10:17.620 ⇒ 00:10:18.940 Uttam Kumaran: a little bit smaller.
25 00:10:20.410 ⇒ 00:10:43.640 Uttam Kumaran: so basically, what we chatted about last time. You know, we we really just tried to accomplish a basic proof of concept. To kind of give everybody a brief overview of what we talked about last time. Basically, Diego and his team, you know, small team of just a few people looking to deploy, you know, capital in Brazil, really, really lean in terms of technology stack like just email. And the Crm. And then, of course.
26 00:10:43.930 ⇒ 00:11:08.919 Uttam Kumaran: your you talking and you pitching but also for actually understanding their portfolio company, the potential prospect portfolio companies. They go through a pretty rigorous research process. Of course, I’m sure a lot of which is is just googling and clicking around and trying to find some esoteric information from multiple different sources. Whether that’s Linkedin, whether that’s Google, whether that’s crunch, base.
27 00:11:08.920 ⇒ 00:11:28.007 Uttam Kumaran: Apollo, everything. So one of the things that we really look to do was to take some of the companies that you sent us and look to create that investment research report, basically mimicking exactly the process. You would do, of course, doing that all. Via AI. So I just want to show you. I’ll walk you through 3 examples.
28 00:11:28.320 ⇒ 00:11:45.989 Uttam Kumaran: I’ll just start with this company, Salvi Brazil. That I that I think was on the list. So really, what we did is we’re using this tool called relevance relevance is our preferred tool for building these sorts of research agents. Something that we we do pretty commonly for folks. So
29 00:11:46.180 ⇒ 00:11:59.810 Uttam Kumaran: basically, a relevance agent has access to different tools. And so we’ve built several tools to not only look and scrape Linkedin, scrape crunch base, go onto the actual website for the company itself.
30 00:12:00.137 ⇒ 00:12:12.444 Uttam Kumaran: And we’ve actually been able to build this agent where you can put in. You know, basically any information that you have, and it’s able to actually output this in the key format. So the company history all the stats
31 00:12:13.060 ⇒ 00:12:19.139 Diego Gomes: Do you mind if I read? This is a company I’m super familiar with. It’s 1 of our portfolio investments.
32 00:12:19.370 ⇒ 00:12:23.310 Diego Gomes: So nation company stats.
33 00:12:35.290 ⇒ 00:12:39.579 Diego Gomes: Got it. Can you go down? I already screened through everything.
34 00:12:41.250 ⇒ 00:12:57.670 Diego Gomes: So until the the until topic 3, you got everything right. The only thing you didn’t got right is the number of phone lines they operate. But you you would not supposed to do that. It’s a confidential. They currently manage 10,000 lines
35 00:12:57.990 ⇒ 00:13:01.719 Diego Gomes: when it comes to the competitive landscape.
36 00:13:01.830 ⇒ 00:13:03.619 Diego Gomes: I think this is the area.
37 00:13:03.620 ⇒ 00:13:08.590 Uttam Kumaran: In a sense of on. It’s on their Y combinator page. They put 5,000 so.
38 00:13:08.770 ⇒ 00:13:22.297 Diego Gomes: I can imagine that. And that’s totally fine. I wasn’t expecting that the competition space and the competition analysis is really off like it’s not. in line.
39 00:13:23.490 ⇒ 00:13:27.109 Diego Gomes: the team is okay. Value prop is okay.
40 00:13:27.510 ⇒ 00:13:32.559 Diego Gomes: assessment for competition, not good product offerings.
41 00:13:35.830 ⇒ 00:13:40.880 Diego Gomes: Product offering is very close pricing. Very good.
42 00:13:41.090 ⇒ 00:13:45.010 Diego Gomes: Market opportunity. Man.
43 00:13:46.900 ⇒ 00:13:48.200 Diego Gomes: Very good.
44 00:13:50.950 ⇒ 00:13:52.460 Diego Gomes: Risks.
45 00:13:53.140 ⇒ 00:13:54.310 Diego Gomes: Okay.
46 00:13:55.980 ⇒ 00:14:07.659 Uttam Kumaran: And of course, like the stuff that’s more so. That isn’t just like, find the answer and return it where it’s giving like an analysis, probably where we we would need to work with you a lot closer on, like determining the prompt.
47 00:14:07.660 ⇒ 00:14:31.569 Diego Gomes: And maybe these are things that we. I was thinking about this as as I was seeing. Maybe these are things that we wouldn’t like to automate because they really re require a deep judgment like who are the competitors. How is the exit optionality for our company? I think that by possibly having this field by AI today, it would.
48 00:14:31.660 ⇒ 00:14:35.680 Diego Gomes: it would likely, make us lazy, you know, to a certain degree.
49 00:14:35.840 ⇒ 00:14:37.010 Uttam Kumaran: That’s so weird.
50 00:14:37.700 ⇒ 00:14:45.529 Uttam Kumaran: Yeah. And and then the other. The other last piece is that all the information from where it took it provides the references at the bottom, which for us.
51 00:14:45.530 ⇒ 00:14:46.120 Diego Gomes: Very good.
52 00:14:46.120 ⇒ 00:15:05.120 Uttam Kumaran: Or because you’re totally right, there’s 2 kinds of information. One is just purely like, find it, summarize it and include it. The second thing is where it’s making like a judgment, or it’s almost like providing that’s totally I would consider those separate and especially in terms of like accuracy. Right. This is purely as you mentioned, it’s gonna be based.
53 00:15:05.120 ⇒ 00:15:27.479 Diego Gomes: But this is pretty impressive, pretty good, and useful. One thing I would be curious. This is a company that I believe there is more public information because it went through. I combinator curious to see maybe a company that does not have such a presence in the Us. But very, very good congrats. Thanks for sharing that.
54 00:15:27.690 ⇒ 00:15:42.740 Uttam Kumaran: Yeah. And to give you another example, we we did this for hub for pay as well. And and again happy to try this for another one. If you have one off hand we can, we can put it in. But again, you kind of get the gist. We also, of course, tested it with.
55 00:15:43.040 ⇒ 00:15:46.400 Uttam Kumaran: I I wanted to test it with Brainforge, which was our company, and like.
56 00:15:46.610 ⇒ 00:16:13.509 Uttam Kumaran: generally like looking at what it said. And again. Yeah, some stuff we don’t have any public information really on like metrics or something like that. So it’s definitely going to kind of pull from industry benchmarks at some point and kind of get a sense of like what it could be. But I would say, for the amount of effort we put in the other. Yeah, it’s pretty good. The other thing that I wanted to share is one this takes about like a few minutes to generate. I don’t know Miguel is even shorters, maybe like one or 2 min
57 00:16:13.906 ⇒ 00:16:14.660 Uttam Kumaran: and it’s.
58 00:16:14.660 ⇒ 00:16:15.750 Miguel de Veyra: 3 to 5.
59 00:16:15.940 ⇒ 00:16:35.279 Uttam Kumaran: Okay, okay, so a few minutes to generate. And then it really goes through each of these tools and gets access. So really, the response is as accurate as these tools that we built. But of course, like, if we want to hook into if you have like French based pro if you have, like Linkedin Sales navigator, we can include that and and get that information. A lot of this is.
60 00:16:35.280 ⇒ 00:16:35.860 Diego Gomes: Shape.
61 00:16:35.860 ⇒ 00:16:40.189 Uttam Kumaran: Available. And we’re also using perplexities. Api, to to grab some stuff as well.
62 00:16:40.617 ⇒ 00:16:46.540 Uttam Kumaran: And then the last piece. Of course, we this isn’t integrated with your Crm, but that’s just something that we would. We would
63 00:16:46.790 ⇒ 00:16:48.470 Uttam Kumaran: pretty much do. Yeah.
64 00:16:48.470 ⇒ 00:17:11.019 Diego Gomes: And that’s a a question I would like to share, maybe 2 or 3 additional companies to take a look at the the results, if that’s possible, and also to to try to understand the, because there’s a ton of value there and a ton of things that could be already included automatically in our Crm.
65 00:17:11.050 ⇒ 00:17:14.649 Diego Gomes: that’s something that you guys could build as well. Is that correct?
66 00:17:15.079 ⇒ 00:17:19.769 Uttam Kumaran: Yeah. Do you mean? Adding it to the Crm or taking it from the Crm.
67 00:17:20.490 ⇒ 00:17:25.129 Diego Gomes: Adding it like stuff such as industry. There are
68 00:17:25.329 ⇒ 00:17:35.319 Diego Gomes: you? I imagine you’re gathering this information from Crunch base, and in case that it doesn’t exist fill it in into affinity. For instance.
69 00:17:35.320 ⇒ 00:17:35.830 Uttam Kumaran: Definitely.
70 00:17:37.460 ⇒ 00:17:38.260 Diego Gomes: Murphy.
71 00:17:38.260 ⇒ 00:17:44.159 Uttam Kumaran: It’ll be, it’ll be the interaction back to writing it back to affinity. But then, also modifying any of the Crm fields.
72 00:17:45.270 ⇒ 00:17:46.650 Diego Gomes: Got it clear.
73 00:17:47.660 ⇒ 00:17:48.189 Uttam Kumaran: Yeah, if you want.
74 00:17:48.750 ⇒ 00:17:53.000 Uttam Kumaran: If I’m happy to try another one like let’s if you want to send me
75 00:17:53.020 ⇒ 00:17:55.210 Uttam Kumaran: even like a a link, or a blurb, or.
76 00:17:55.210 ⇒ 00:18:00.970 Diego Gomes: I just sent 3 links on the on the chat if if you could.
77 00:18:01.200 ⇒ 00:18:04.969 Diego Gomes: If it takes more time, doesn’t need to be. Now, that’s totally fine.
78 00:18:05.280 ⇒ 00:18:08.480 Uttam Kumaran: No, let’s I would love to. I would love to try it.
79 00:18:14.430 ⇒ 00:18:15.380 Uttam Kumaran: one sec.
80 00:18:31.900 ⇒ 00:18:52.319 Uttam Kumaran: And then maybe while this is working I’ll give you a little bit of sense of the tools that we’re using. So we built not only tools to find to to search for the company on Google extract stuff from Linkedin. We also built some of these like investor finder tools. I’ll show you an example of like what one of these looks like
81 00:18:55.150 ⇒ 00:19:00.359 Uttam Kumaran: Correct me from Miguel. Do you remember which source we’re using for the investor, finder, tool.
82 00:19:00.646 ⇒ 00:19:04.650 Miguel de Veyra: It’s on the Llm. If you expand it, I added. Some prompts in there.
83 00:19:06.700 ⇒ 00:19:08.929 Miguel de Veyra: The one below Llm. Step.
84 00:19:09.320 ⇒ 00:19:10.030 Uttam Kumaran: Okay.
85 00:19:13.590 ⇒ 00:19:16.299 Miguel de Veyra: Basically, I instructed it to, you know, use some stuff.
86 00:19:16.300 ⇒ 00:19:43.309 Diego Gomes: Oh, so it does a search for a specific company name plus keywords. This is super interesting, because when I think about this specific case we, because we operate in La 10, these queries, you are using their work. Okay? But we could also have a queries for similar similar structure in Portuguese and Spanish. So in
87 00:19:43.310 ⇒ 00:19:47.299 Diego Gomes: totally decidores, which is the Portuguese for investors.
88 00:19:49.036 ⇒ 00:19:57.320 Diego Gomes: Let’s say, angel investment is investment to angel, etcetera. But I get the idea in it.
89 00:19:57.870 ⇒ 00:20:18.040 Uttam Kumaran: Yeah. And this is where, like, we will work with you 100. And basically, we would probably watch you go through one of this steps, especially for a company that’s easy, but also company that may not have anything, and then really just mimic that and take like a lot of those those nuance and and add it in here. So yeah, let’s so let’s see what it came up with. I just.
90 00:20:18.040 ⇒ 00:20:20.969 Miguel de Veyra: You might have to reload the page.
91 00:20:23.890 ⇒ 00:20:31.630 Miguel de Veyra: Alright, it’s still working, I think one of the things with them you can show them is not. It’s not a American company. Is the auto. Me one
92 00:20:32.390 ⇒ 00:20:34.169 Miguel de Veyra: Ottoman Poland? Yeah.
93 00:20:35.420 ⇒ 00:20:37.809 Miguel de Veyra: So this is more of like a Polish company.
94 00:20:39.950 ⇒ 00:20:46.089 Diego Gomes: Got it. Yeah, harder for me to understand, because it’s not a company I’m super familiar with.
95 00:20:46.300 ⇒ 00:21:00.500 Diego Gomes: I think the way to time tune this, in a sense, is to pick a small set of companies that we know very well, and make sure that we can optimize for those, and then it will build a generalist approach. For the rest.
96 00:21:00.960 ⇒ 00:21:27.466 Uttam Kumaran: Yeah. And also even even to talk about the ways that we were thinking about this. There’s some research reports that are going to be purely just extracting summary. And then there’s probably going to be some more advanced, as you guys trust the data more and are more comfortable with making some sort of judgments or pre screening in some way where there’ll be a little bit more sophistication. That’s what I’m I’m super super excited, you know, to share. Of course we just, you know, it’s probably spent a few days on this. But
97 00:21:27.950 ⇒ 00:21:30.460 Uttam Kumaran: yeah, let’s see, this may be taking a little bit longer.
98 00:21:31.070 ⇒ 00:21:47.909 Diego Gomes: That’s totally fine. So while this is loading would you mind sharing a little bit about how do you guys work? How we could leverage your software. And your services to work together here is to understand how we could possibly collaborate
99 00:21:48.426 ⇒ 00:21:52.799 Diego Gomes: to to bring that back to my partners and devise a potential path.
100 00:21:53.120 ⇒ 00:21:57.229 Uttam Kumaran: Yeah, totally. I don’t know, Robert, do. Would you? Wanna take that.
101 00:22:00.040 ⇒ 00:22:10.309 Robert Tseng: Yeah, sure. So I guess. At a high level, Diego, I think for the way that we were thinking about breaking this down is, we have the implementation fee, obviously, to set this up for you.
102 00:22:10.821 ⇒ 00:22:16.789 Robert Tseng: But then I think we wanted to introduce rather than like a retainer, or whatever kind of like
103 00:22:16.830 ⇒ 00:22:35.370 Robert Tseng: usage or kind of user based pricing, we like to structure our our our pricing around the quality of lead. So what we’ve done is kind of broken out we can send you a proposal as well. But our idea is like to structure it based on the quality of the lead that we bring you.
104 00:22:35.370 ⇒ 00:22:52.410 Robert Tseng: So you know, if if a lead, if if it’s a lead that actually goes through your pipeline and you and you close you close that close that deal. Then we want. We want that to be, you know, price more than just like a regular lead that we do just to hit hit the numbers. So
105 00:22:53.150 ⇒ 00:23:17.970 Robert Tseng: that’s kind of like the model that we had in in mind. Or it’s more usage based pricing. We know that your volumes don’t seem to be that high right now, like we kind of did some napkin math and kind of back into maybe like beating you 200 or 300 leads, but we want to do it in a way where we’re incentivized to give you leads that have a higher likelihood of closing than what you’d be able to do with your current resourcing. So
106 00:23:18.540 ⇒ 00:23:19.790 Robert Tseng: yeah, that’s kind of.
107 00:23:19.790 ⇒ 00:23:24.490 Diego Gomes: I believe, thinking about this on our on our process here.
108 00:23:25.059 ⇒ 00:23:31.630 Diego Gomes: We receive let’s say, 100 200 new leads per month.
109 00:23:31.870 ⇒ 00:23:43.109 Robert Tseng: Yeah, we end up investing 4, maybe 5 companies a year. So it’s it’s maybe close rate is not the the necessarily success we’re looking for.
110 00:23:43.150 ⇒ 00:23:50.650 Diego Gomes: Okay. I I think that elimination is often a good thing like this. Company has
111 00:23:50.780 ⇒ 00:23:57.446 Diego Gomes: 300 people. It’s too large. It doesn’t make any sense, and we don’t know why, is in the pipeline
112 00:23:57.880 ⇒ 00:24:08.060 Diego Gomes: elimin. Eliminating companies and manual data entry is a big value for us helping us find.
113 00:24:08.120 ⇒ 00:24:12.280 Diego Gomes: Let’s pick an example of a company. We looked into Salvi.
114 00:24:12.310 ⇒ 00:24:17.990 Diego Gomes: Let’s say we are investing. We solve, we have. We’re looking into the company we haven’t invested yet.
115 00:24:18.210 ⇒ 00:24:30.109 Diego Gomes: But if you can provide competitive and similar companies to look for, to also makes us improve the quality of our process by a lot
116 00:24:30.730 ⇒ 00:24:32.900 Diego Gomes: that can be definitely.
117 00:24:33.020 ⇒ 00:24:38.050 Diego Gomes: Let’s say we are looking to invest in brain Forge would be extremely helpful to.
118 00:24:38.060 ⇒ 00:24:47.599 Diego Gomes: These are 5 companies that are, we have a similar value. Proposition are doing similar things, and you should also take a look at. I think
119 00:24:47.720 ⇒ 00:24:53.070 Diego Gomes: I can definitely see some some opportunities being created from these.
120 00:24:54.180 ⇒ 00:25:18.869 Robert Tseng: Great. Yeah, no, I think that makes sense. So rather than making our success metric conversion rate make it kind of we could. I don’t know if we want to make it like a lead quality score, whatever ends up being but we could do something that helps you to like filter out leads that don’t fit your profile. So but yeah, some some sort of quality metric that we can work towards. I think that makes a lot of sense to us. Yeah.
121 00:25:18.870 ⇒ 00:25:41.530 Uttam Kumaran: Yeah. And and one other thing there, Diego. So because of, you know, as we saw, we have these different sophistication of reports. One thing that we were thinking is, you know, we have this initial research report that you just submit a lead into you. Get back the report. Of course, we, even we can even go some step beyond which is, as you mentioned, like an industry report or competitive analysis, where that’s like all the people that are competing with them.
122 00:25:41.540 ⇒ 00:25:53.480 Uttam Kumaran: maybe reports on them. There’s also like meeting brief. So let’s say you decide to go forward and you meet with one of the leads like, what is a pre meeting brief that you can have in your inbox? Basically right? And then the last thing.
123 00:25:53.480 ⇒ 00:25:53.820 Diego Gomes: It is.
124 00:25:53.820 ⇒ 00:26:18.859 Uttam Kumaran: Trying to get toward more of the like, more of the qualitative analysis that you probably would do right. I think this will, that this. This sort of stuff will require a lot of heavy work between us to like integrate. And of course the level of quality has to be high like there can be no hallucinations. It has to be adhered to you. It’s not as easy as Googling for everything and and summarizing, but of course the value there is.
125 00:26:18.860 ⇒ 00:26:44.760 Uttam Kumaran: You guys can go from taking on 100 200 leads a month, maybe taking on a thousand without scaling your operations right? And just maybe your disqualification rate stays the same. But then you’re you’re able to get down to the really, really core and have a lot more of those. So that’s how we think about also the quality and the sophistication of the reports. Given the complexity of like how your lead qualification works.
126 00:26:45.480 ⇒ 00:26:58.250 Diego Gomes: And and I think that for a lot of these companies we have a valuable source of data in our hands which are not for everything. But most of the companies after the initial stage, have the pitch deck associated
127 00:26:59.173 ⇒ 00:27:09.260 Diego Gomes: saved on affinity, and that could filter a lot of even to to the ones that we didn’t find a lot of public information we could find.
128 00:27:09.270 ⇒ 00:27:13.480 Diego Gomes: If we can parse these decks to extract those.
129 00:27:13.550 ⇒ 00:27:22.080 Diego Gomes: For instance, most decks have clear revenues, competitors, team size.
130 00:27:22.330 ⇒ 00:27:27.800 Diego Gomes: It’s a lot of gaps that we don’t find on public data we could find there.
131 00:27:28.160 ⇒ 00:27:37.070 Uttam Kumaran: Yeah. So that’s another step, for sure is, if you have that custom knowledge, and we can just scrape and bring that in and then use that to fill in the gaps, and that becomes a reference.
132 00:27:37.220 ⇒ 00:27:38.550 Uttam Kumaran: But I think we have.
133 00:27:38.730 ⇒ 00:28:00.250 Uttam Kumaran: I think, a a general path forward. So for us. I think we’ll put together a proposal one. It’ll be some sort of initial upfront implementation fee, just to make sure we can hook into affinity. And then again, instead of kind of going on this retainer based model for improvements, we want to look to provide you with the more sophisticated reports, and we’ll price sort of based on the success and quality of those
134 00:28:00.350 ⇒ 00:28:04.740 Uttam Kumaran: as those get more sophisticated. We will have like a pricing scheme for that. So.
135 00:28:04.740 ⇒ 00:28:28.080 Diego Gomes: Got it, and when it comes to fine tuning these, I saw your the interface there, for instance, on the the investor search. And I think I could easily do some tweaks there to to improve, based on what I saw is that something that we can collaborate? It’s we can put our hands on and and play around with as well, or it’s.
136 00:28:28.080 ⇒ 00:28:35.840 Uttam Kumaran: Yeah, 100. So the reason why we use the tool like relevance instead of building this all as code is actually because we want
137 00:28:35.900 ⇒ 00:28:48.540 Uttam Kumaran: people who are non technical to interact with the system. If we were to build this all in code and have it all in Github, it’d become really difficult. So the development process is actually a lot lighter. Of course, the more complicated the task.
138 00:28:48.630 ⇒ 00:29:11.029 Uttam Kumaran: the more creativity we will have to have. So for the for this sort of stuff. It’s pretty manageable, depending on how sophisticated we want the system to get. There may be pieces of it that may not be as easy but a 100. We’re actually hoping that we can include all your team, not only to show you we’ll have an Amira aboard the entire architecture. But I should get your feedback there and show you a process where you can iterate
139 00:29:11.030 ⇒ 00:29:19.099 Uttam Kumaran: for us. The quality of this gets better with your feedback. So there’s no, we don’t have any problem with gatekeeping on on that side.
140 00:29:19.100 ⇒ 00:29:30.949 Uttam Kumaran: I actually would love to to show you how how we develop it, and it’ll give you guys ideas for, hey? We should use this other source. Or if I have some other information, how do we include it? It’ll give you the full picture. Totally.
141 00:29:30.950 ⇒ 00:29:31.670 Diego Gomes: Got it.
142 00:29:32.090 ⇒ 00:29:58.239 Diego Gomes: and one area we would be curious as well is to understand what our variable costs, let’s say, and a specific Api or specific X to have a sense of how much we would have. Because, let’s say, we, we close the deal. We will work together. The one of the 1st things we would like to do would be to run our entire database of companies, which is around. Let me check quickly here.
143 00:30:00.110 ⇒ 00:30:07.870 Diego Gomes: Around 7,000 companies to make sure that they are properly enraged and qualified. But we obviously
144 00:30:07.970 ⇒ 00:30:12.710 Diego Gomes: understanding the cost for running something like that would be critical.
145 00:30:13.020 ⇒ 00:30:13.360 Uttam Kumaran: Okay?
146 00:30:15.200 ⇒ 00:30:33.490 Uttam Kumaran: So then, what we can do is we can take on providing you back with a proposal for the implementation fee for this fill of all past leads and then go forward based on this current research report and probably some of the the proposed items. So we can come back with you with that, probably in the next few days. If not sooner.
147 00:30:33.490 ⇒ 00:30:35.649 Diego Gomes: Appreciate it. Looking forward.
148 00:30:35.650 ⇒ 00:30:38.266 Uttam Kumaran: I’ll I’ll also send you this meeting.
149 00:30:38.760 ⇒ 00:30:40.489 Uttam Kumaran: so that you have it as well.
150 00:30:41.100 ⇒ 00:30:45.400 Diego Gomes: Is it? Okay? If I send you a few more companies here? hey?
151 00:30:45.400 ⇒ 00:30:52.040 Diego Gomes: The, let me just take a quick look on the pipeline. Here the offload.
152 00:30:52.040 ⇒ 00:31:01.559 Uttam Kumaran: And additionally, any any other. If you guys beyond just the company links, you can also send it other information like a founder name, or
153 00:31:01.690 ⇒ 00:31:06.330 Uttam Kumaran: you know, similar Comp companies, and it will leverage for notes. So.
154 00:31:06.330 ⇒ 00:31:16.790 Diego Gomes: So I will do what I will do here will be a I’ll select 5 to 10 records in our current pipeline, and with what we already know to share with you, so we can.
155 00:31:16.790 ⇒ 00:31:17.180 Uttam Kumaran: Perfect.
156 00:31:17.180 ⇒ 00:31:17.850 Diego Gomes: Tests.
157 00:31:18.240 ⇒ 00:31:22.809 Uttam Kumaran: And so, yeah, well, and then we’ll I’ll send you all those. I’ll run it through and send you all send you all those.
158 00:31:23.510 ⇒ 00:31:28.179 Diego Gomes: Appreciate it. Thanks a lot. Guys and congrats. That’s really really interesting.
159 00:31:28.480 ⇒ 00:31:30.120 Uttam Kumaran: Cool. Thank you.
160 00:31:30.480 ⇒ 00:31:32.989 Robert Tseng: Cool. Yeah. Alright.
161 00:31:32.990 ⇒ 00:31:33.350 Uttam Kumaran: Awesome.
162 00:31:33.350 ⇒ 00:31:33.969 Miguel de Veyra: Thanks. Everyone.