Meeting Title: Uttam <> Joshua—AI-Automation Date: 2024-08-14 Meeting participants: Miguel, Nicolas Sucari, Uttam Kumaran
WEBVTT
1 00:01:37.720 ⇒ 00:01:38.450 Uttam Kumaran: Hello!
2 00:01:39.290 ⇒ 00:01:40.280 Miguel: Hello! There!
3 00:01:40.860 ⇒ 00:01:42.030 Uttam Kumaran: Hi! How are you?
4 00:01:42.040 ⇒ 00:01:43.730 Miguel: I’m doing good.
5 00:01:43.850 ⇒ 00:01:46.320 Miguel: I’ve been sleeping. It’s almost 3 a.
6 00:01:47.100 ⇒ 00:01:50.490 Uttam Kumaran: Oh, yeah, sorry. I really appreciate you taking the time today.
7 00:01:50.490 ⇒ 00:01:51.980 Miguel: No, no worries, no worries.
8 00:01:54.170 ⇒ 00:01:55.519 Miguel: Pronounce your name. By the way.
9 00:01:55.810 ⇒ 00:01:56.990 Uttam Kumaran: Yeah, it’s uton.
10 00:01:56.990 ⇒ 00:01:58.150 Miguel: Utsam? Okay? Okay?
11 00:01:58.730 ⇒ 00:01:59.810 Uttam Kumaran: Yeah. Yeah.
12 00:02:00.420 ⇒ 00:02:03.419 Uttam Kumaran: And you go by Josh, or you go by. Miguel.
13 00:02:03.420 ⇒ 00:02:07.979 Miguel: My! My full name’s Joshua Miguel, so I could go by either. But usually I just go, Miguel.
14 00:02:08.199 ⇒ 00:02:09.389 Miguel: something that’s.
15 00:02:09.910 ⇒ 00:02:13.348 Uttam Kumaran: Okay. Nice. How’s the week going.
16 00:02:14.264 ⇒ 00:02:18.930 Miguel: Yeah, pretty hectic, because, you know, preparing for holidays next week.
17 00:02:19.230 ⇒ 00:02:25.319 Miguel: Pretty, I think I’ll be. Yeah. I’ll be out of country, I believe 22 0, no. 21 to 26.th
18 00:02:25.940 ⇒ 00:02:27.639 Uttam Kumaran: Where are you? Where are you going.
19 00:02:28.200 ⇒ 00:02:32.259 Miguel: Probably I’m not sure yet. Probably Hong Kong. I, Hong Kong or Korea.
20 00:02:32.910 ⇒ 00:02:34.000 Uttam Kumaran: Oh, nice. Okay.
21 00:02:34.000 ⇒ 00:02:34.880 Miguel: The iter.
22 00:02:36.020 ⇒ 00:02:37.630 Uttam Kumaran: Nice, exciting.
23 00:02:37.660 ⇒ 00:02:38.750 Uttam Kumaran: jealous.
24 00:02:41.855 ⇒ 00:02:53.369 Uttam Kumaran: Cool. I guess I I know I’m message, and I know we just we’ve been messing back and forth. I just wanted to tell you a little bit about my company, but you know I was kind of impressed with some of the stuff
25 00:02:53.480 ⇒ 00:02:59.126 Uttam Kumaran: that you know you’ve been able to do, and really particularly looking for
26 00:03:00.320 ⇒ 00:03:19.406 Uttam Kumaran: automations to help me automate my business. You know everything from automating like how we do outbound sales to like internal integrations with Github to writing code? So we have a kind of a ton of opportunity. So basically, I’m looking to bring on someone to kind of work with us part time on a weekly basis.
27 00:03:20.211 ⇒ 00:03:30.798 Uttam Kumaran: To kind of help automate several different processes. So my company brain forge is a data analytics consultancy. Basically, we help companies,
28 00:03:31.320 ⇒ 00:03:41.890 Uttam Kumaran: do data analytics anything from data engineering to data modeling to analysis. So we’re a team of data engineers, even myself included. And so
29 00:03:42.090 ⇒ 00:03:52.679 Uttam Kumaran: so yeah, they started the company about a year ago, and so have been getting us, you know, certain contracts. You know, really, with the goal of scaling the business up.
30 00:03:53.360 ⇒ 00:04:00.670 Uttam Kumaran: so yeah, I mean again, I saw a couple of things that I saw on your profile that I liked was, you know, you use relevance AI, which was.
31 00:04:00.670 ⇒ 00:04:01.780 Miguel: In a sense.
32 00:04:01.780 ⇒ 00:04:24.648 Uttam Kumaran: Interesting cause. I’ve been looking into that as well. It looks kind of cool for spinning up agents. Like, you know you have. You had your social media automation. You had your Gen. AI for marketing like images and stuff like that. So yeah, just excited to to hear like kind of like what you’ve done and like where what? You’re interested.
33 00:04:25.150 ⇒ 00:04:26.459 Uttam Kumaran: and working.
34 00:04:28.140 ⇒ 00:04:29.250 Miguel: Yeah.
35 00:04:30.407 ⇒ 00:04:35.882 Miguel: my main experience, honestly in automation is on Ghl Zapier.
36 00:04:36.540 ⇒ 00:04:38.869 Miguel: I’m not sure if you’re familiar with airtable.
37 00:04:39.340 ⇒ 00:04:41.430 Uttam Kumaran: Yeah, I am. What’s ghl.
38 00:04:41.773 ⇒ 00:04:44.519 Miguel: Go high level, it’s basically a Crm, yeah.
39 00:04:44.520 ⇒ 00:04:45.669 Uttam Kumaran: Okay, yeah, yeah.
40 00:04:45.670 ⇒ 00:04:55.650 Miguel: Yeah, so there, cause yeah, that’s that’s where my, you know, core experience in automations are because their automations are a bit complex, I would say.
41 00:04:55.910 ⇒ 00:04:58.430 Miguel: and a bit confusing most of the time.
42 00:04:59.434 ⇒ 00:05:10.139 Miguel: But yeah, mo, but most of the time I do double in AI, right? Basically AI agents. That’s why relevance. I actually met with their founders a couple of times already.
43 00:05:12.030 ⇒ 00:05:13.039 Uttam Kumaran: Oh, nice!
44 00:05:13.320 ⇒ 00:05:18.329 Miguel: Yeah. And then, yeah, conversational agents, voice agents. That’s
45 00:05:18.570 ⇒ 00:05:20.929 Miguel: primarily what I do. Of course.
46 00:05:21.160 ⇒ 00:05:24.249 Miguel: it’s pointless. All of that is pointless without automations.
47 00:05:25.100 ⇒ 00:05:25.720 Miguel: But yeah.
48 00:05:28.310 ⇒ 00:05:29.256 Uttam Kumaran: So then,
49 00:05:30.140 ⇒ 00:05:55.709 Uttam Kumaran: like, what? Out of so kind of the way I’m breaking things down in my company is like, we have engineering related automations which can be things like taking a Github issue and actually like proposing the code change. It could also be like doing Pr reviews and things like that. There’s also things on the operation side, like, for example, summarizing all the Github issues that have been closed so that we can provide updates to clients
50 00:05:55.710 ⇒ 00:06:01.100 Uttam Kumaran: that’s helping us, you know, automate, some of our stand ups and some of our project management operations.
51 00:06:01.883 ⇒ 00:06:12.280 Uttam Kumaran: Other things that are interesting. Are helping us on the sales side. So helping. Do you know, creating personalized messaging for outbound emails.
52 00:06:12.280 ⇒ 00:06:13.069 Miguel: Hi! Greedy.
53 00:06:13.070 ⇒ 00:06:21.879 Uttam Kumaran: Personalized messaging for out for content, that to go get posted on like Linkedin Tiktok, things like that. So social media automation
54 00:06:22.382 ⇒ 00:06:38.770 Uttam Kumaran: and then also, basically just, you know, helping basically try to write anything that comes. You know, any lead that comes in automating like research and things like that using AI, and ideally, these, you know, right now, we’ve been doing a lot using cloud projects. But I want to kind of build some of these like small
55 00:06:38.850 ⇒ 00:06:51.260 Uttam Kumaran: like workflows, basically that can leverage our knowledge base, like our slack messages and things like that to kind of help write these. So those are some of, like the ideas that we’ve been thinking about.
56 00:06:52.572 ⇒ 00:06:54.699 Miguel: Yeah. Yeah. Marketing.
57 00:06:54.790 ⇒ 00:07:03.500 Miguel: My, my main experience is in marketing, right? Github stuff. Yeah, developers I’ve worked with doesn’t really wanna automate their stuff.
58 00:07:03.520 ⇒ 00:07:12.110 Miguel: But marketing sales. That’s primarily where I do my automations with AI like I’ve worked in with this company before, where basically, they’re a British company
59 00:07:12.170 ⇒ 00:07:13.369 Miguel: where you know.
60 00:07:14.220 ⇒ 00:07:17.439 Miguel: they’re more on a digital marketing side of things.
61 00:07:18.927 ⇒ 00:07:23.689 Miguel: But yeah, basically, it allowed they allow their users to like input their, you know.
62 00:07:24.650 ⇒ 00:07:25.970 Miguel: Wait! Can I even see it?
63 00:07:27.200 ⇒ 00:07:29.500 Miguel: No, I think I’m bound by Nda. Sorry.
64 00:07:29.640 ⇒ 00:07:34.979 Miguel: But basically some details about their company, and then we we run a prompt.
65 00:07:35.120 ⇒ 00:07:38.620 Miguel: We engineered the prompt, and then it’ll automatically.
66 00:07:39.518 ⇒ 00:07:43.269 Miguel: what do you call this? Great suggestions on where they could improve, you know.
67 00:07:43.460 ⇒ 00:07:45.519 Miguel: And yeah, something like that.
68 00:07:47.700 ⇒ 00:07:48.886 Uttam Kumaran: Okay? And
69 00:07:50.780 ⇒ 00:07:59.989 Uttam Kumaran: so like, out of the kind of the things that I mentioned, you think like the marketing stuff would be more interesting like, do you? Do you have any interest in working on this stuff across
70 00:08:00.130 ⇒ 00:08:04.450 Uttam Kumaran: like operations and like the github stuff, or no.
71 00:08:05.740 ⇒ 00:08:15.350 Miguel: yeah, I mean, yeah, I’m interested in this. Just I don’t have experiences on automating that stuff. Or maybe I don’t. I just don’t see what you mean completely yet.
72 00:08:15.760 ⇒ 00:08:17.370 Miguel: Okay, yeah.
73 00:08:17.780 ⇒ 00:08:18.830 Miguel: yeah. I mean.
74 00:08:19.070 ⇒ 00:08:25.829 Uttam Kumaran: Actually, it’s literally like, you know, how you can go into Claude and just ask it to post to give you code to answer problems like it’s, it’s just that.
75 00:08:26.332 ⇒ 00:08:27.840 Miguel: Yeah. Yeah. Yeah.
76 00:08:30.100 ⇒ 00:08:54.449 Uttam Kumaran: you know. So it’s automating like some of those. But okay, I mean, so kind of like, let’s say, for example, we want to take a project like right now, we don’t do any posting on Linkedin, or Tiktok, or Twitter or Instagram. But I’m trying to. Basically, we do have case studies and blog posts that we’ve written. One of the smaller goals we’re trying to do is basically take those and turn those into Linkedin tick. Tock Twitter posts.
77 00:08:54.460 ⇒ 00:09:02.999 Uttam Kumaran: Is that something that like you’ve done before? That’s something that you know. Seems like it’s possible. Maybe we can even break down that project and see if that’s something we can try.
78 00:09:03.610 ⇒ 00:09:09.210 Miguel: Yeah, for Linkedin. Specifically, I’ve created like this software, for I think they’re based in Florida.
79 00:09:09.220 ⇒ 00:09:11.410 Miguel: I’m not sure where you guys are based in the Us.
80 00:09:11.800 ⇒ 00:09:12.589 Uttam Kumaran: Based on.
81 00:09:13.060 ⇒ 00:09:13.960 Miguel: Sorry, where?
82 00:09:14.130 ⇒ 00:09:15.250 Uttam Kumaran: In Austin, Texas.
83 00:09:15.250 ⇒ 00:09:26.760 Miguel: Oh, Texas, yeah. So they’re based in Florida. Basically, you know. It’s a Linkedin post generator based on whatever day. Input, and then there’s like a prompt that we did. That’s
84 00:09:26.820 ⇒ 00:09:39.330 Miguel: using a formula for writing, because they were before they were. What do you call this hiring 3rd party copywriters to write the posts. And then when we deployed that AI yeah, kind of saved because they’re
85 00:09:39.550 ⇒ 00:09:41.560 Miguel: we had over a hundred clients.
86 00:09:42.530 ⇒ 00:09:48.759 Miguel: and then they were paying around $400 per post, something like that. And then they were only earning around
87 00:09:48.780 ⇒ 00:09:51.060 Miguel: 2030 after the, you know.
88 00:09:51.612 ⇒ 00:09:57.829 Miguel: Yeah. But yeah, we were able to automate post. So I think that’s pretty relevant to what you guys want to do.
89 00:09:58.620 ⇒ 00:10:06.139 Uttam Kumaran: So what and what technologies did you use for that like? How? What’s what was the workflow and and how did you generate all the pieces of content.
90 00:10:06.460 ⇒ 00:10:12.099 Miguel: Oh, that one was purely Openai, and then relevance wasn’t there yet when I did that
91 00:10:12.370 ⇒ 00:10:21.409 Miguel: at least I didn’t know they were pretty new. So I had to do that all encoding. I used my stuff technically and then connected it to
92 00:10:21.810 ⇒ 00:10:23.909 Miguel: completions. Openai.
93 00:10:26.020 ⇒ 00:10:31.130 Uttam Kumaran: So you actually wrote like open a you connected to the like open AI Api to write all that.
94 00:10:31.130 ⇒ 00:10:34.079 Miguel: Yes, yes, and then fine tune some of it, of course.
95 00:10:34.750 ⇒ 00:10:46.180 Uttam Kumaran: And then. So now that there’s relevance, and there’s like flow wise. And some of these other tools like, are you using any of these workflow tools, like Agent Hub, or any of that? Are you still basically connecting directly to the Api.
96 00:10:46.831 ⇒ 00:10:52.030 Miguel: For some projects open AI directly the Openai assistance, the one in web
97 00:10:54.095 ⇒ 00:10:59.469 Miguel: but most of the time I just do relevance because it’s a lot easier. Honestly, the views are there, the tools.
98 00:11:00.040 ⇒ 00:11:01.740 Miguel: And I really like relevance.
99 00:11:02.550 ⇒ 00:11:06.240 Uttam Kumaran: Yeah, so where did like I, I clicked around in relevance, like.
100 00:11:06.320 ⇒ 00:11:15.460 Uttam Kumaran: I wonder if, can you do you have any like example you could show, because I’ve been clicking around relevant. But there’s so much you could do that I’m like, not really sure even like.
101 00:11:15.520 ⇒ 00:11:20.150 Uttam Kumaran: where to start like. Do you have an example that you can maybe show on your screen or something.
102 00:11:20.573 ⇒ 00:11:22.690 Miguel: Yeah. Sure, let me just.
103 00:11:23.860 ⇒ 00:11:35.130 Uttam Kumaran: This is also a friend of mine said to check out rel relevance like a month ago. And I looked at. I’m like there’s just so much here. But I mean, like I, if you could just put everything in relevance that maybe that’s the best thing to do.
104 00:11:36.380 ⇒ 00:11:41.010 Miguel: Relevance is basically for just like, let me show you.
105 00:11:41.370 ⇒ 00:11:43.190 Miguel: let me share my screen
106 00:11:43.700 ⇒ 00:11:45.310 Miguel: with them. Turning this off.
107 00:11:50.100 ⇒ 00:11:53.020 Miguel: Where is relevance. I think it’s this one. Yeah.
108 00:11:53.670 ⇒ 00:11:55.369 Miguel: can you see relevance.
109 00:11:57.220 ⇒ 00:11:58.210 Uttam Kumaran: Yes. Correct.
110 00:11:58.544 ⇒ 00:12:00.549 Miguel: Yeah. So basically, it’s this one.
111 00:12:01.246 ⇒ 00:12:11.639 Miguel: Once you log in right? The core of relevance is agents. You can build agents to either. Like, for example, we create a new one. They have. They already have templates, but on a stage is great from black.
112 00:12:12.030 ⇒ 00:12:16.769 Miguel: So you know, market research, Marco something like that. And then.
113 00:12:17.050 ⇒ 00:12:20.559 Miguel: once you create one. For example, let’s start from
114 00:12:22.132 ⇒ 00:12:26.840 Miguel: which which makes the most sense from probably in your spring summarizer. Now.
115 00:12:27.080 ⇒ 00:12:28.199 Uttam Kumaran: Yeah, let’s do that.
116 00:12:34.130 ⇒ 00:12:41.599 Miguel: And then you go over here. You, this is basically the prompt. Of course, this is pretty bad, right? We have to change that
117 00:12:41.880 ⇒ 00:12:52.573 Miguel: flow. Builder, is the conversation flow like, what’s the 1st thing you should say? If the user chooses to do this, what should it, for example, cancellation reschedule, or
118 00:12:53.070 ⇒ 00:12:55.589 Miguel: what else? Adding a new session.
119 00:12:55.620 ⇒ 00:12:57.560 Miguel: you know it’s gonna be.
120 00:12:58.513 ⇒ 00:13:00.679 Miguel: How it goes from here.
121 00:13:00.910 ⇒ 00:13:09.659 Miguel: 1st step, second step, and then abilities this one we don’t really, especially meets this labeling tasks. This is just for reviewing. This is a new feature.
122 00:13:09.810 ⇒ 00:13:14.789 Miguel: Tools is basically where it’s connected. So, for example, you want to connect it to.
123 00:13:14.950 ⇒ 00:13:18.159 Uttam Kumaran: Github or something. You can have a yeah, so similar to Openai tools.
124 00:13:18.960 ⇒ 00:13:22.289 Miguel: Yeah, but a lot more. Because if you go like here
125 00:13:22.310 ⇒ 00:13:24.729 Miguel: to tools, there’s actually a lot
126 00:13:25.364 ⇒ 00:13:29.010 Miguel: for example, you can even, I think, connect it to.
127 00:13:29.350 ⇒ 00:13:30.160 Uttam Kumaran: Nice.
128 00:13:30.160 ⇒ 00:13:34.529 Miguel: Yeah, there’s a lot you can directly, I believe, connected to Zapier.
129 00:13:34.760 ⇒ 00:13:36.760 Miguel: There’s stuff like this. And then, you know.
130 00:13:37.240 ⇒ 00:13:41.800 Miguel: yeah, something like this. And then I believe there’s Zendesk. I’m not sure if you guys use that.
131 00:13:42.570 ⇒ 00:13:43.729 Uttam Kumaran: We don’t use Zendesk, but.
132 00:13:43.730 ⇒ 00:13:46.929 Miguel: Apollo, I believe you use here. Yeah. So something like this.
133 00:13:47.875 ⇒ 00:13:48.520 Uttam Kumaran: Okay.
134 00:13:48.520 ⇒ 00:13:53.800 Miguel: I’m not sure if instantly is here, but if not, then, you know, can always go through Zapier, and then just.
135 00:13:54.100 ⇒ 00:13:57.329 Uttam Kumaran: You can write, you can write a direct instant. Api. Request.
136 00:13:58.280 ⇒ 00:13:59.410 Miguel: You can. Sorry what.
137 00:14:00.020 ⇒ 00:14:04.089 Uttam Kumaran: But if if it’s not, if it’s there, or if it’s in Zapier, you can write an Api request to.
138 00:14:04.090 ⇒ 00:14:07.739 Miguel: Yeah, yeah, yeah, yeah, yeah, just post it to Zapier. Yeah. So you know.
139 00:14:08.280 ⇒ 00:14:09.980 Miguel: yeah, it’s pretty, you know.
140 00:14:10.020 ⇒ 00:14:15.449 Miguel: it’s the only thing. The only thing I don’t like about is, it’s pretty expensive. I think their base.
141 00:14:16.080 ⇒ 00:14:21.380 Miguel: at least for us, right? Because I’m based in the Philippines. It’s like $200 a month.
142 00:14:21.690 ⇒ 00:14:22.510 Uttam Kumaran: Yeah.
143 00:14:22.510 ⇒ 00:14:26.040 Miguel: Yes, so it’s pretty steep, especially if you’re just experimenting with it.
144 00:14:27.490 ⇒ 00:14:28.520 Miguel: Yeah, it’s like a hundred.
145 00:14:28.520 ⇒ 00:14:34.740 Uttam Kumaran: It’s per day. 4 credits per run, so it looks like you could probably run it like 20 times per day for free.
146 00:14:35.520 ⇒ 00:14:50.960 Miguel: Yeah. But yeah, like, for example, the the biggest thing they have is multi agent systems, like, so, for example, this agent, I can have. For example, this is the main agent. I want to have an agent that’s does, you know? Marketing, and then one does sales.
147 00:14:51.170 ⇒ 00:15:05.370 Miguel: So it’s all connected. Right? If I if the main agent, basically, I call it router agent, receives a task, it categorizes it. Hey? This task is best for this agent that that agent. So yeah, multi agent systems.
148 00:15:06.020 ⇒ 00:15:06.720 Uttam Kumaran: Okay.
149 00:15:07.410 ⇒ 00:15:12.400 Uttam Kumaran: I mean, what do you think? So you would probably start here, and then if it gets expensive, like, rip it out.
150 00:15:13.939 ⇒ 00:15:21.899 Miguel: Yeah, honestly, unless it cause there was a company. It’s reflex Australia. I built a chat bot for them.
151 00:15:22.020 ⇒ 00:15:27.189 Miguel: and we had like 200,000. We actually never went past 20,000 credits.
152 00:15:27.200 ⇒ 00:15:30.080 Miguel: because you can connect your Openai here. So.
153 00:15:30.080 ⇒ 00:15:39.390 Uttam Kumaran: Yeah, I I don’t mind if it’s like even gets there, because by that time I’m sure we could see if like, can we simplify it or move it out. Have you heard of flow wise before.
154 00:15:39.390 ⇒ 00:15:45.358 Miguel: Flow. Wise. Yes, I actually we tried it before, because we were, you know, dabbling on which
155 00:15:47.176 ⇒ 00:16:00.220 Miguel: which ones to make this one is just a bit too complicated for my taste. Right? Because I want, yeah, like, I wanted low code to no code, right? It’s also cause I usually work closely with Ceos and Ctos
156 00:16:00.799 ⇒ 00:16:07.730 Miguel: and operations. It’s a it’s a lot harder to explain to them why we should do this instead of you know if they can just see it like this.
157 00:16:07.890 ⇒ 00:16:14.479 Uttam Kumaran: Totally. I I think the nice thing about us is like, I’m like, I’m an engineer, and our whole company is like most engineers. So.
158 00:16:14.480 ⇒ 00:16:15.220 Miguel: Yeah, yeah.
159 00:16:15.220 ⇒ 00:16:26.190 Uttam Kumaran: I actually am. I’m interested in like I think I I felt the same way like I I did stuff on cloud projects or custom Gpts, and then I did stuff in Agent an agent hub, which.
160 00:16:26.190 ⇒ 00:16:27.440 Miguel: Exactly. Yes.
161 00:16:28.123 ⇒ 00:16:34.450 Uttam Kumaran: I was doing there. But then some stuff it’s like they don’t have the tool or thing like that. So then I was like, Okay, what’s
162 00:16:34.460 ⇒ 00:16:47.389 Uttam Kumaran: what’s the next step? And it was like relevance or flow. Wise, I think, flow wise. The nice thing I like about flow. Wise is you can self host it. So basically, you can probably get it to where you’re only paying for the for the AI calls.
163 00:16:47.390 ⇒ 00:16:47.790 Miguel: Yeah.
164 00:16:47.790 ⇒ 00:17:01.970 Uttam Kumaran: However, you’re right in that. It’s like probably hard to build. So my idea is like, I’m I wanted to start doing it like 1st you just start by like asking Chaturbt directly, then build a project or custom. Gpt, then let’s see if we can do it in
165 00:17:02.210 ⇒ 00:17:07.460 Uttam Kumaran: relevance, and then finally, we’ll move it to flow wise like down the line, or yeah, or crew.ai, yeah.
166 00:17:07.460 ⇒ 00:17:08.970 Miguel: Yeah, yeah, this one.
167 00:17:09.960 ⇒ 00:17:13.600 Uttam Kumaran: So what’s your favorite out of all these? You you think like relevance is like the best.
168 00:17:14.069 ⇒ 00:17:17.739 Miguel: Yeah, relevance. I think ease of use, you know, easy to deploy everything
169 00:17:18.604 ⇒ 00:17:23.629 Miguel: yeah. And then through is, yeah. I tried signing up, but I think they’re. You know
170 00:17:24.019 ⇒ 00:17:25.069 Miguel: they have.
171 00:17:25.069 ⇒ 00:17:30.969 Uttam Kumaran: Zapier is so expensive. Dude. So that’s why I’m like trying to just do direct. Api calls because
172 00:17:31.169 ⇒ 00:17:32.949 Uttam Kumaran: Zapier is also annoying.
173 00:17:33.190 ⇒ 00:17:40.129 Miguel: Yeah, that Zapier is kinda expensive, like I think in my previous company, it’s like a thousand dollars a month in Zapier alone.
174 00:17:41.780 ⇒ 00:17:45.480 Uttam Kumaran: Yeah, that’s what I’m saying. So I mean, I would be down to pay for
175 00:17:45.640 ⇒ 00:18:01.590 Uttam Kumaran: relevance. If you think that’s worth it. And like. But again, I think maybe like, what would you think would be if if I was to say like, Hey, let’s spend like a week or less on like one task, just to see like, if we can work together. What do you think would be good to try to do.
176 00:18:03.057 ⇒ 00:18:05.490 Miguel: If cost is like the biggest issue.
177 00:18:05.570 ⇒ 00:18:09.560 Miguel: it’s probably best to do directly in Openai platforms.
178 00:18:09.720 ⇒ 00:18:12.810 Miguel: right? And then but the thing there is.
179 00:18:12.810 ⇒ 00:18:20.129 Uttam Kumaran: Do cost isn’t like a huge issue. But I like, I want to just make sure that we’re not spending just because
180 00:18:20.290 ⇒ 00:18:30.930 Uttam Kumaran: it’s like way easier. And also I don’t mind again, I don’t mind doing, because doing the $200 a month, and then our plan is to move it off of it later, you know, for me. I want to work first.st
181 00:18:31.180 ⇒ 00:18:31.920 Miguel: Yep, yep.
182 00:18:33.160 ⇒ 00:18:38.180 Miguel: yeah. Yeah. I mean, yeah, we can start with relevance. Or you know, if
183 00:18:38.790 ⇒ 00:18:43.089 Miguel: cause it depends, if, do we? Do, you guys really, do you guys want to maintain another code base
184 00:18:43.110 ⇒ 00:18:48.620 Miguel: cause? That’s the other thing we will have to do if we stick with like, if you want to avoid, you know, platform.
185 00:18:48.620 ⇒ 00:19:10.219 Uttam Kumaran: I’m sold on the way you kind of told me about relevance. I think the biggest thing is like, how what would your process typically be like, would you sit so? Nico also, who’s on this call, is our project manager. So he runs all of the projects that we have so like, would you typically like, sit with one of us and like, kind of go through the flow or like, what’s your process in working with executives?
186 00:19:11.580 ⇒ 00:19:12.630 Nicolas Sucari: I’m Eagle.
187 00:19:12.630 ⇒ 00:19:18.379 Miguel: Hello work basically planning. How do you say it?
188 00:19:22.420 ⇒ 00:19:24.589 Miguel: Because I work closely with the CTO
189 00:19:24.710 ⇒ 00:19:30.729 Miguel: most of the time, right and I manage cause I usually just manage the developers to be honest.
190 00:19:31.870 ⇒ 00:19:32.699 Uttam Kumaran: Okay. Okay.
191 00:19:32.700 ⇒ 00:19:36.560 Miguel: And then the big tasks, like the hard ones that no one else can do. That’s what I do.
192 00:19:37.304 ⇒ 00:19:47.269 Miguel: So yeah, most of the time people do come to me. The only time I go to like executives like our CTO is when you know it’s like an issue. Even I can solve.
193 00:19:47.470 ⇒ 00:19:53.780 Miguel: And then, of course, you know, managing people, that’s Mo, mostly what I talk about with executives
194 00:19:54.080 ⇒ 00:19:55.510 Miguel: and solving, you know.
195 00:19:55.680 ⇒ 00:19:59.690 Miguel: basically automating a lot of stuff in my last company.
196 00:20:00.680 ⇒ 00:20:08.540 Uttam Kumaran: So yeah, I mean in this thing, you’d be mostly working with me and Nico, like I, I run a company, and then Nico is like, second in command, basically. So.
197 00:20:10.472 ⇒ 00:20:26.550 Uttam Kumaran: okay, I mean, what do you think? Like, you think it’s best to see if we can take on like one sample project like, what’s your availability like? And again, I’m trying to bring someone on to kind of work with us longer. Term to kind of begin to automate a lot of stuff
198 00:20:26.550 ⇒ 00:20:44.369 Uttam Kumaran: I really love that you’re using relevance. And you’re familiar with all these tools. I think there’s a lot of people who I’m like. I tell all these tools to, and they just have no idea. And I’m like about these. And I try every single one. And I’ve I’ve tried relevance. I’ve tried flow wise, but of course, like I just, I’m like running the company, so I don’t. I want.
199 00:20:44.370 ⇒ 00:20:44.940 Miguel: Yeah, yeah.
200 00:20:44.940 ⇒ 00:20:45.420 Uttam Kumaran: Money, that.
201 00:20:45.420 ⇒ 00:20:45.980 Miguel: To explore.
202 00:20:45.980 ⇒ 00:20:55.079 Uttam Kumaran: Basically, I have all the ideas, and I have all the ways of doing it. But, like, you know, I don’t have 40 h enough time to spend on relevance.
203 00:20:55.080 ⇒ 00:20:56.909 Miguel: Yeah, yeah, yeah.
204 00:20:56.910 ⇒ 00:20:58.180 Uttam Kumaran: What? Yeah.
205 00:21:00.790 ⇒ 00:21:07.879 Miguel: yeah, we could. I think you could start with something yeah, start with something small and build from there right?
206 00:21:07.880 ⇒ 00:21:08.670 Uttam Kumaran: Okay.
207 00:21:08.670 ⇒ 00:21:15.130 Miguel: Yeah. Cause that’s usually how I do things right tackle one. The problem, one small step at a time.
208 00:21:15.730 ⇒ 00:21:16.450 Uttam Kumaran: Okay.
209 00:21:16.926 ⇒ 00:21:20.480 Miguel: Yeah. But through AI, this thing I want to have access to.
210 00:21:20.550 ⇒ 00:21:23.059 Miguel: I never had access to this, for whatever reason.
211 00:21:23.750 ⇒ 00:21:30.119 Miguel: like their builder like, you have to sign up. Yeah. Yeah. So I think they’re one of the best. But it’s really hard to get in.
212 00:21:30.660 ⇒ 00:21:42.849 Uttam Kumaran: Oh, okay, okay. The nice thing is, I’ve because of my company. We’ve been getting access to everything really fast, like I got, I got multi on access. I got flow wise access. I’m sure I can probably get through AI access.
213 00:21:44.450 ⇒ 00:21:51.480 Miguel: It might be better to move, you know. Eventually, when we decide to move from relevance, it could be better to move here.
214 00:21:52.350 ⇒ 00:21:53.950 Miguel: Then.
215 00:21:55.280 ⇒ 00:21:59.919 Miguel: flow wise also. There’s another. I think I’ve already talked to one of the better shift.
216 00:22:00.930 ⇒ 00:22:04.519 Miguel: They’re just like flow wise, but a bit more complicated.
217 00:22:04.970 ⇒ 00:22:06.540 Uttam Kumaran: Even more complicated.
218 00:22:06.540 ⇒ 00:22:11.860 Miguel: Yeah. Yeah. And the CEO I talked to him is a bit I would say he’s a bit of an ass, to be honest.
219 00:22:11.860 ⇒ 00:22:12.560 Uttam Kumaran: Oh!
220 00:22:13.260 ⇒ 00:22:20.840 Miguel: So I wasn’t an onboarding call with him, and he was like it should. It’s so simple you should understand it. I was like, that’s why we’re here.
221 00:22:21.810 ⇒ 00:22:22.859 Uttam Kumaran: Yeah, yeah, yeah.
222 00:22:24.202 ⇒ 00:22:27.107 Miguel: But yeah, so I don’t really recommend this guy.
223 00:22:27.940 ⇒ 00:22:32.709 Miguel: flow. Wise is okay. This one I want to have access to relevance is the best, I would say.
224 00:22:33.630 ⇒ 00:22:39.510 Uttam Kumaran: Okay. Then I’m I’m gonna just like I’m I’m I think I already have a relevance account. So I might just set it up. But like
225 00:22:39.590 ⇒ 00:22:47.456 Uttam Kumaran: so what do you think like? Do you want to try and do like one small thing for like a week, and then, or like, what’s your, what’s your availability? Like,
226 00:22:49.020 ⇒ 00:22:49.880 Uttam Kumaran: yeah.
227 00:22:50.675 ⇒ 00:22:53.980 Miguel: Yeah, I mean, hopefully, it’s not. Well, let me stop sharing.
228 00:22:54.500 ⇒ 00:22:58.084 Miguel: Well, hopefully, it’s not like this time all the time. Right?
229 00:22:58.410 ⇒ 00:22:59.459 Uttam Kumaran: No, no, no, no! I mean.
230 00:22:59.460 ⇒ 00:23:00.250 Miguel: Yeah.
231 00:23:00.400 ⇒ 00:23:05.269 Uttam Kumaran: I actually don’t like we. We do like Async almost as much as possible.
232 00:23:05.270 ⇒ 00:23:05.850 Nicolas Sucari: Yeah.
233 00:23:06.020 ⇒ 00:23:09.020 Uttam Kumaran: As long as you’re on, like, as long as you’re good with slack.
234 00:23:09.040 ⇒ 00:23:10.210 Uttam Kumaran: Yeah.
235 00:23:10.460 ⇒ 00:23:13.900 Uttam Kumaran: then that’s it. It doesn’t matter. We just basically.
236 00:23:13.980 ⇒ 00:23:32.719 Uttam Kumaran: we have, like a daily stand up on that can ask questions, and it’ll just give updates. Otherwise slack is perfect. And then I can talk like early morning, or even like late night, my time, if we need to talk about anything and all that stuff so as much Async as possible. Like I don’t. I’m not really trying to sit meetings.
237 00:23:32.963 ⇒ 00:23:33.450 Miguel: Okay, okay.
238 00:23:33.450 ⇒ 00:23:47.700 Nicolas Sucari: Yeah, we’re trying to be. We’re trying to be totally Async but obviously we need some time to discuss on slack, maybe, or something like that. So if you are able to respond to the messages for that, that’s perfect, and we have that stand above that make us
239 00:23:47.730 ⇒ 00:23:56.099 Nicolas Sucari: our life a little bit easier, so that we can understand what is everyone doing? But that’s it. I mean, we don’t need to be jumping into meetings all the time.
240 00:23:56.100 ⇒ 00:24:02.719 Miguel: Oh, yeah, yeah, I mean, I’m I’m I’m fine with jumping to meetings all the time, actually, especially with this type of projects. It’s just, you know.
241 00:24:02.810 ⇒ 00:24:05.450 Miguel: 3 Am. Is a bit harsh, you know. But yeah.
242 00:24:05.450 ⇒ 00:24:10.010 Nicolas Sucari: No, yeah. Where are you? Where are you from now, or where are you? Where it’s your time zone.
243 00:24:10.010 ⇒ 00:24:11.210 Miguel: Manila Manila.
244 00:24:12.110 ⇒ 00:24:19.069 Miguel: I like maybe at up until like 1, 2 am. I don’t really mind cause I do my stuff asynchronously, too, right?
245 00:24:19.070 ⇒ 00:24:23.660 Uttam Kumaran: No, I think we’ll. I think morning. Our time will end up being like 10 o’clock or 9 o’clock.
246 00:24:23.660 ⇒ 00:24:24.030 Nicolas Sucari: Yeah.
247 00:24:24.030 ⇒ 00:24:28.880 Uttam Kumaran: Like we’ll we’ll do a call then, because I’m up at 8 Am. My time, so you can just chat.
248 00:24:29.170 ⇒ 00:24:29.460 Nicolas Sucari: Him.
249 00:24:29.460 ⇒ 00:24:32.210 Uttam Kumaran: And names and stuff. Otherwise, like no problem.
250 00:24:32.430 ⇒ 00:24:33.310 Miguel: Oh, yeah. Yeah.
251 00:24:34.440 ⇒ 00:24:50.059 Nicolas Sucari: Hey? One more question before any temporary what would you need if you like, if we decide to go like with one of the projects. What? What are like your initial steps, or what do you need from us so that you can start like doing anything on these tools?
252 00:24:51.074 ⇒ 00:24:54.820 Miguel: Well, I need like the you know. What do you guys want for the initial project?
253 00:24:54.880 ⇒ 00:24:58.490 Miguel: It doesn’t have to be like super detailed. We can discuss it as we go right.
254 00:24:59.720 ⇒ 00:25:00.370 Nicolas Sucari: Great.
255 00:25:01.550 ⇒ 00:25:02.080 Nicolas Sucari: Yeah.
256 00:25:02.080 ⇒ 00:25:03.780 Miguel: And then, of course, a relevance account.
257 00:25:04.510 ⇒ 00:25:18.879 Uttam Kumaran: Okay, cool. So maybe maybe what I’ll do is I’ll I’m just gonna invite you to a slack channel, and then I’m just gonna put together a notion document with one of these tasks. Let’s just try one of these out, Nico. I I want to see whether we could do the get up thing.
258 00:25:19.800 ⇒ 00:25:20.900 Nicolas Sucari: Yeah, perfect.
259 00:25:21.540 ⇒ 00:25:23.420 Uttam Kumaran: Want to try that or or like.
260 00:25:23.820 ⇒ 00:25:30.979 Uttam Kumaran: I think we should try that, and then we can start to work on stuff in sales, maybe. But I do want to see whether we can basically
261 00:25:31.080 ⇒ 00:25:34.540 Uttam Kumaran: get an update of all the Github issues that were completed that week.
262 00:25:36.520 ⇒ 00:25:38.610 Uttam Kumaran: you know, and and try that 1st
263 00:25:39.169 ⇒ 00:25:48.710 Uttam Kumaran: and then Miguel. I guess, like I can slack or email you, and maybe like we could try one thing. And then after that. Let’s talk about like rate and stuff like that.
264 00:25:49.320 ⇒ 00:25:50.469 Miguel: Okay. Yeah. Sure. Sure. No worries.
265 00:25:50.470 ⇒ 00:26:11.147 Uttam Kumaran: I don’t want to take up like I don’t want to take up like 2 weeks of your time while you were just figuring this out. So let’s just try to pair it down to something that’s like going to take like a week or less. And then I mean, I’m I’m pretty confident that you have what like, what we need anyways. But that way you, you get a sense of what working with us is like, and then you also. You can fire us if you don’t like us, and then
266 00:26:11.915 ⇒ 00:26:21.050 Uttam Kumaran: and then that way, ideally again, like my goal. And what I told the whole company is like, try to automate like 50% of the company or more.
267 00:26:21.050 ⇒ 00:26:21.430 Miguel: If they.
268 00:26:21.430 ⇒ 00:26:28.556 Uttam Kumaran: So I like kind of the way you mentioned, which is like you sit with all the people and basically try to figure out like what the
269 00:26:29.510 ⇒ 00:26:47.839 Uttam Kumaran: what the process are, and slowly automate everything, and that you’re also technical. So that helps because I have like a list already of, like most of the processes that we do and the way I broke it down is it’s between engineering operations and sales. Is is everything. The rest of the stuff on finance and
270 00:26:47.980 ⇒ 00:27:01.559 Uttam Kumaran: everything else is like is really manual, not not. And it’s also like, not worth automating. Yeah, every engineering. And again, I’m going even further. It’s like, I want to be able to write code given like, really, really good requirements.
271 00:27:02.160 ⇒ 00:27:07.289 Uttam Kumaran: And basically, the goal is to scale this agency up without scaling up, you know people.
272 00:27:09.370 ⇒ 00:27:15.009 Uttam Kumaran: So I think. And I’m I’m willing to use whatever tool and like, try out all these tools.
273 00:27:15.020 ⇒ 00:27:17.729 Uttam Kumaran: So yeah, let’s make it happen.
274 00:27:17.730 ⇒ 00:27:19.610 Miguel: Okay. Yeah, sure, I haven’t.
275 00:27:20.240 ⇒ 00:27:27.440 Uttam Kumaran: Okay, cool. So after this I’ll invite you to slack and then on slack. I’ll also send you a notion that we can use as like the running document, and then let’s go from there.
276 00:27:27.997 ⇒ 00:27:31.392 Miguel: Yeah, yeah, just utumn. I have like
277 00:27:32.610 ⇒ 00:27:38.709 Miguel: question, right? Cause I will be out of country. As I mentioned earlier next week. So how should we proceed? Is that fine.
278 00:27:38.910 ⇒ 00:27:42.240 Uttam Kumaran: What what days are you gonna be out? You’re gonna be out all week next week.
279 00:27:42.240 ⇒ 00:27:43.990 Miguel: No, probably starting Wednesday
280 00:27:44.100 ⇒ 00:27:44.630 Miguel: to Monday.
281 00:27:44.630 ⇒ 00:27:48.660 Uttam Kumaran: Yeah, maybe we could just spend time between now and Wednesday. Like, if that’s fine with you.
282 00:27:49.174 ⇒ 00:27:50.430 Miguel: Okay, okay, sure, sure.
283 00:27:50.430 ⇒ 00:28:08.969 Uttam Kumaran: Let’s just talk. Let’s just talk, at least on Tuesday, before you head out. But again we’ll be in slack. I don’t think I’m I don’t think I want to take on anything that’s more than that without like getting a contract in place and everything. So let’s just let’s just take on one small thing for a few days. See how it goes. Sign up for relevance, and then let’s let’s see how it goes.
284 00:28:08.970 ⇒ 00:28:10.249 Miguel: Okay. Okay. Sure. Sure.
285 00:28:11.360 ⇒ 00:28:13.990 Uttam Kumaran: Okay, so great conversation. Yeah. Thank you so much.
286 00:28:13.990 ⇒ 00:28:15.570 Miguel: Okay, thanks. Guys, have a good day.
287 00:28:15.930 ⇒ 00:28:16.540 Uttam Kumaran: Okay. Thanks.
288 00:28:16.540 ⇒ 00:28:17.170 Nicolas Sucari: Thank you.
289 00:28:17.170 ⇒ 00:28:17.710 Miguel: See you all.
290 00:28:17.710 ⇒ 00:28:18.359 Nicolas Sucari: Bye, bye.
291 00:28:18.720 ⇒ 00:28:19.700 Nicolas Sucari: size.