Meeting Title: AI-Internal-Meeting Date: 2024-09-25 Meeting participants: Uttam Kumaran, Nicolas Sucari, Joshuadeveyra
WEBVTT
1 00:02:42.250 ⇒ 00:02:43.090 joshuadeveyra: Hey! Uttam.
2 00:02:46.980 ⇒ 00:02:47.670 Uttam Kumaran: Hey? How’s it going.
3 00:02:51.590 ⇒ 00:02:53.080 joshuadeveyra: With you. Well, hey, you, Tom.
4 00:02:54.670 ⇒ 00:02:56.050 Uttam Kumaran: Alright. Thanks, guys.
5 00:03:02.190 ⇒ 00:03:04.289 Uttam Kumaran: Cool, I guess.
6 00:03:04.864 ⇒ 00:03:11.205 Uttam Kumaran: Miguel, you wanna leave. I kind of want. Since you’re running AI stuff. I kind of want you to leave if you’re able to.
7 00:03:11.890 ⇒ 00:03:12.530 joshuadeveyra: Okay.
8 00:03:12.530 ⇒ 00:03:22.350 Uttam Kumaran: And maybe just talk about clients we’re working on. And then any other stuff. I kind of want this to be basically the meeting where we talk about like any AI client work, any AI internal work. And then
9 00:03:22.780 ⇒ 00:03:26.420 Uttam Kumaran: basically like anything we want to try to plan or prioritize.
10 00:03:26.720 ⇒ 00:03:27.570 joshuadeveyra: Yeah, okay.
11 00:03:27.740 ⇒ 00:03:28.550 Uttam Kumaran: So perfect.
12 00:03:29.800 ⇒ 00:03:37.010 joshuadeveyra: Yeah. So for internal, there’s really, you know, nothing much I’m doing for internal right now, at least temporarily, because.
13 00:03:37.410 ⇒ 00:03:38.559 Nicolas Sucari: Sorry we’ll go ahead.
14 00:03:38.670 ⇒ 00:03:45.349 Nicolas Sucari: Do we need Pat here? He joined before, but we can ping him if we want him.
15 00:03:45.575 ⇒ 00:03:51.220 Uttam Kumaran: Yeah, I’m I’m I’m just gonna ping him to ask you if he’s if he wants to join and listen in but go ahead.
16 00:03:51.750 ⇒ 00:03:52.430 Uttam Kumaran: Yep.
17 00:03:54.140 ⇒ 00:04:01.220 joshuadeveyra: Okay, yeah. So internally, we’re not really doing much right now. In terms of the AI stuff, I think.
18 00:04:01.914 ⇒ 00:04:07.879 joshuadeveyra: the next stuff internally is gonna be the engineering side or the sales hub. Actually, no, it’s the sales hub. But
19 00:04:08.050 ⇒ 00:04:11.350 joshuadeveyra: yeah, we’re prioritizing Hpi, and
20 00:04:12.088 ⇒ 00:04:16.029 joshuadeveyra: via the coco, especially, Vitaco is just giving us like feedback right.
21 00:04:16.970 ⇒ 00:04:17.760 Uttam Kumaran: Yes.
22 00:04:18.430 ⇒ 00:04:19.430 joshuadeveyra: So.
23 00:04:19.660 ⇒ 00:04:25.260 Uttam Kumaran: Yeah, let’s talk about so Hpi, I’m gonna actually go try to meet with Scott today.
24 00:04:25.761 ⇒ 00:04:28.500 Uttam Kumaran: In person. So I’m gonna show him
25 00:04:29.610 ⇒ 00:04:30.340 Uttam Kumaran: your
26 00:04:30.510 ⇒ 00:04:33.739 Uttam Kumaran: zoom demo, but also just kind of walk through the relevant stuff.
27 00:04:34.050 ⇒ 00:04:40.509 Uttam Kumaran: It’s awesome like I love the process. I think that’s probably where we were gonna end up anyways, which is like
28 00:04:40.680 ⇒ 00:04:49.270 Uttam Kumaran: lease like portions of the lease specific agents. And then I like the idea of it. Can you explain, like the idea of the exception agent.
29 00:04:50.310 ⇒ 00:04:53.559 joshuadeveyra: Oh, yeah, the exception agent is, for example, like.
30 00:04:54.407 ⇒ 00:05:16.889 joshuadeveyra: what do you call this? Because there’s like intense instances that. The AI, you know, I was basically talking to it where it says, like, there will be certain scenarios where, for example, they want to lower their deposit, probably by a couple of months and then return, you know, their security deposit, and then they’ll in return. They’ll extend their lease by a 6 month period, or maybe a year.
31 00:05:17.070 ⇒ 00:05:30.400 joshuadeveyra: So in that scenario it’s gonna be rejected by the security deposit agent. But maybe you know an exception agent or an additional agent that tackles that type of stuff could make an exception for that.
32 00:05:30.890 ⇒ 00:05:36.630 Uttam Kumaran: So the exception agent is gonna be like, kind of what we talked about, which is like, oh, this is a strategic company, or like
33 00:05:36.890 ⇒ 00:05:38.180 Uttam Kumaran: this is like
34 00:05:38.480 ⇒ 00:05:40.640 Uttam Kumaran: they have other lease. Okay, okay, cool.
35 00:05:41.050 ⇒ 00:05:48.590 joshuadeveyra: Yeah. But currently, we don’t really know like what the exceptions are, you know. So we kind of have to ask them. But yeah, I can kind of create just basically
36 00:05:48.760 ⇒ 00:05:51.030 joshuadeveyra: demo one for now.
37 00:05:51.150 ⇒ 00:05:54.450 joshuadeveyra: And we can just adjust it based on their, you know, feedback.
38 00:05:55.040 ⇒ 00:06:07.229 Uttam Kumaran: So a nice, a nice, like actual test for us is like, imagine we were to start with the 1st version of the lease. And actually, we imagine we were able to get the emails of like, Hey, can we request this? Or like the
39 00:06:07.340 ⇒ 00:06:13.399 Uttam Kumaran: the actual request that would be, basically we would run our process in parallel with their process, right.
40 00:06:15.510 ⇒ 00:06:20.509 joshuadeveyra: Yeah. Although there’s like a I have like a slight concern there.
41 00:06:20.870 ⇒ 00:06:21.950 joshuadeveyra: because
42 00:06:22.270 ⇒ 00:06:28.030 joshuadeveyra: I was thinking like, how we will demo it with them when the entire cause. I tried uploading a lease.
43 00:06:28.500 ⇒ 00:06:32.079 joshuadeveyra: This, which was a bit complicated in relevance, to be honest
44 00:06:32.395 ⇒ 00:06:39.520 joshuadeveyra: but the issue there was. It took me around 10 to 15 min just to wait. I mean not me the AI to process that.
45 00:06:40.120 ⇒ 00:06:42.930 joshuadeveyra: So I’m not sure how that will look like in the demo right.
46 00:06:43.940 ⇒ 00:06:50.310 Uttam Kumaran: Oh, yeah, so, but that that’s okay. As long as it’s working, we’ll find a way to basically, I mean, I think ideally, we’ll
47 00:06:50.470 ⇒ 00:06:53.290 Uttam Kumaran: we’ll use like some other vector store for that
48 00:06:54.730 ⇒ 00:06:55.410 Uttam Kumaran: or something.
49 00:06:55.410 ⇒ 00:06:58.789 joshuadeveyra: Yeah, yeah. I mean for them. It should be fine for now.
50 00:06:59.060 ⇒ 00:07:18.040 Uttam Kumaran: Yeah for Demo. I’m just glad to see that it’s working. I think the best the best demo we could do is basically we have like version one through 7 of that one specific lease that they’re doing. We could walk through all the comments. And then basically what you what it would have been like, if you’re using the AI, the copilot, basically, you know.
51 00:07:18.040 ⇒ 00:07:18.340 joshuadeveyra: Oh!
52 00:07:19.380 ⇒ 00:07:22.730 Uttam Kumaran: You see what I mean? So our barometer there is like.
53 00:07:23.320 ⇒ 00:07:27.520 Uttam Kumaran: what would the AI have suggested when the when the changes requested came through.
54 00:07:29.090 ⇒ 00:07:35.109 joshuadeveyra: Okay? Yeah. I mean, that was all. That’s already what I was doing for at least security deposit and cause the
55 00:07:35.170 ⇒ 00:07:43.519 joshuadeveyra: document. The lease document. We’re like working on right now. It’s they change like the tenant name, basically to, I think their parent company.
56 00:07:44.370 ⇒ 00:07:45.699 joshuadeveyra: So yeah.
57 00:07:47.080 ⇒ 00:07:50.760 Uttam Kumaran: Okay, cool. So great. So let me, I’m gonna pitch that to Scott today. And then.
58 00:07:50.950 ⇒ 00:07:53.580 Uttam Kumaran: yeah, I guess. Look, let’s let’s pause on
59 00:07:53.750 ⇒ 00:07:55.200 Uttam Kumaran: development there.
60 00:07:55.240 ⇒ 00:07:58.959 Uttam Kumaran: unless there’s any small things that you’re doing, and I’ll I’ll try to get some feedback to that.
61 00:07:59.470 ⇒ 00:08:06.109 joshuadeveyra: Okay, yeah. I guess. For now we focus on Vita cocoa, right? The 68 New York stores.
62 00:08:06.440 ⇒ 00:08:17.539 Uttam Kumaran: Yeah. So I didn’t look at the excel sheet. But it looks like you just sent us those the stores that we were gonna do. So. The initial thing was, you’re just gonna take a subset of the stores right to test.
63 00:08:17.850 ⇒ 00:08:18.560 joshuadeveyra: Yeah.
64 00:08:19.040 ⇒ 00:08:30.710 joshuadeveyra: I already have, like an idea on how to do it. So basically, we’ll just get so we don’t have to actually touch the scraping part anymore. Because that’s working perfectly, I think what we could do instead is, you know, just
65 00:08:30.720 ⇒ 00:08:40.849 joshuadeveyra: you get the store name. So I’ll just store in instead of like. It’s an array of strings. I’ll just make it an array of objects. And then the you know store URL the name of the store and the Zip code.
66 00:08:40.990 ⇒ 00:08:43.080 joshuadeveyra: and then it just gets sent to them.
67 00:08:43.309 ⇒ 00:08:45.289 Uttam Kumaran: Are you running this on your laptop?
68 00:08:46.047 ⇒ 00:08:48.630 joshuadeveyra: Yeah, yeah, for now we haven’t really deployed it.
69 00:08:48.800 ⇒ 00:08:49.940 joshuadeveyra: What are you running? It.
70 00:08:49.940 ⇒ 00:08:52.499 Uttam Kumaran: What are you running with? You’re just running in python.
71 00:08:53.750 ⇒ 00:08:54.470 joshuadeveyra: Express.
72 00:08:54.470 ⇒ 00:08:55.080 Uttam Kumaran: Okay.
73 00:08:56.590 ⇒ 00:08:59.180 joshuadeveyra: But I think it’s turning ronium or something. I forgot.
74 00:08:59.800 ⇒ 00:09:05.499 Uttam Kumaran: That’s fine. Yeah. What we’ll do is I just wanna we’ll we’ll put it into Cloud eventually to run on a cron.
75 00:09:05.540 ⇒ 00:09:08.289 Uttam Kumaran: But that’s great. Okay, perfect.
76 00:09:08.290 ⇒ 00:09:09.030 joshuadeveyra: Playwright.
77 00:09:09.640 ⇒ 00:09:12.890 Uttam Kumaran: Yeah, something like that. Or in windmill?
78 00:09:14.690 ⇒ 00:09:17.710 Uttam Kumaran: okay, perfect. What else.
79 00:09:19.736 ⇒ 00:09:20.370 joshuadeveyra: I think
80 00:09:20.590 ⇒ 00:09:23.570 joshuadeveyra: for now that’s all I’m working on, at least.
81 00:09:23.570 ⇒ 00:09:25.010 Uttam Kumaran: The sales, hub stuff.
82 00:09:27.320 ⇒ 00:09:28.360 joshuadeveyra: Oh, yeah, yeah.
83 00:09:28.920 ⇒ 00:09:33.639 joshuadeveyra: Oh, yeah, that one like, how should we approach that? Because are we gonna build.
84 00:09:33.640 ⇒ 00:09:34.660 Uttam Kumaran: I’ve
85 00:09:35.630 ⇒ 00:09:42.830 Uttam Kumaran: yeah. I just think that doing one knowledge base, one agent, 2 sub agents.
86 00:09:43.140 ⇒ 00:09:44.690 Uttam Kumaran: It’s like a nightmare.
87 00:09:45.227 ⇒ 00:09:48.700 Uttam Kumaran: I just wanna see whether we can maybe have one
88 00:09:51.140 ⇒ 00:09:54.700 Uttam Kumaran: we could maybe just have one agent per
89 00:09:55.652 ⇒ 00:09:57.760 Uttam Kumaran: clients. And then
90 00:09:57.880 ⇒ 00:09:58.933 Uttam Kumaran: we can
91 00:09:59.820 ⇒ 00:10:06.309 Uttam Kumaran: basically just alter which I think this is all going to be rely on if we can connect a super base or an external vector store.
92 00:10:06.640 ⇒ 00:10:11.529 Uttam Kumaran: Because basically, I want to connect the vector store for each client that can pull in
93 00:10:11.660 ⇒ 00:10:13.450 Uttam Kumaran: images. Text all of that.
94 00:10:13.590 ⇒ 00:10:17.559 Uttam Kumaran: So can you look to see what sort of interactions
95 00:10:17.590 ⇒ 00:10:20.230 Uttam Kumaran: relevance has with vector source
96 00:10:21.088 ⇒ 00:10:23.670 Uttam Kumaran: because I don’t mind using relevance as the front end
97 00:10:24.235 ⇒ 00:10:27.819 Uttam Kumaran: because we may want to use a web research tool and other stuff.
98 00:10:28.060 ⇒ 00:10:30.580 Uttam Kumaran: But for the actual document rag.
99 00:10:30.750 ⇒ 00:10:34.040 Uttam Kumaran: I think I want to try to use something else. So can you look into that?
100 00:10:34.260 ⇒ 00:10:34.770 Uttam Kumaran: Yeah.
101 00:10:34.770 ⇒ 00:10:38.360 joshuadeveyra: I mean, if they have like an Api, because basically, the way.
102 00:10:38.360 ⇒ 00:10:40.160 Uttam Kumaran: We don’t have an Api. Yeah, exactly.
103 00:10:40.160 ⇒ 00:10:41.149 joshuadeveyra: Yeah, we can just do that.
104 00:10:41.150 ⇒ 00:10:42.969 Uttam Kumaran: Just pass in the yeah, exactly.
105 00:10:45.289 ⇒ 00:10:49.029 joshuadeveyra: Have you sent me the invite, for how do you pronounce it? N. 8 M.
106 00:10:49.787 ⇒ 00:10:51.800 Uttam Kumaran: I have not set it up yet.
107 00:10:52.270 ⇒ 00:10:53.420 joshuadeveyra: Okay. Okay. Okay.
108 00:10:54.680 ⇒ 00:10:55.069 Uttam Kumaran: But I know.
109 00:10:55.070 ⇒ 00:10:55.360 joshuadeveyra: You know.
110 00:10:55.360 ⇒ 00:10:57.329 Uttam Kumaran: Looks decent, too.
111 00:11:00.123 ⇒ 00:11:06.670 joshuadeveyra: Sorry. What’s the idea is like? There’s 1 agent for a client, and then one kb, for everything.
112 00:11:08.460 ⇒ 00:11:11.980 Uttam Kumaran: Yeah, well, no. One. Kb, per client. Basically.
113 00:11:12.830 ⇒ 00:11:14.780 joshuadeveyra: But isn’t that what we’re doing already.
114 00:11:15.520 ⇒ 00:11:19.531 Uttam Kumaran: Yeah. But oh, not one, no, but not you also have
115 00:11:20.530 ⇒ 00:11:21.399 Nicolas Sucari: You have one.
116 00:11:21.400 ⇒ 00:11:21.660 Uttam Kumaran: About you.
117 00:11:21.660 ⇒ 00:11:23.190 Nicolas Sucari: Cool. True. True. Yeah.
118 00:11:23.190 ⇒ 00:11:24.250 Uttam Kumaran: Yeah.
119 00:11:25.590 ⇒ 00:11:26.630 Uttam Kumaran: Agent, yeah.
120 00:11:26.630 ⇒ 00:11:28.530 Nicolas Sucari: Exactly 2 tools per agent.
121 00:11:29.010 ⇒ 00:11:29.879 joshuadeveyra: Oh, yeah, so.
122 00:11:29.880 ⇒ 00:11:36.220 Uttam Kumaran: Remember, we had challenges like cause. I wanted to be like, Hey, throw everything easily because
123 00:11:36.280 ⇒ 00:11:42.300 Uttam Kumaran: the process of going to the Kb updating it is just not gonna work like it’s just not gonna scale
124 00:11:43.030 ⇒ 00:11:43.970 joshuadeveyra: We have to put.
125 00:11:43.970 ⇒ 00:11:47.629 Uttam Kumaran: You know, because you have to. Someone has to go and click and do it manually. It’s just never like I’m
126 00:11:48.010 ⇒ 00:11:55.609 Uttam Kumaran: as the person who’s on the sales side. There’s no way I do that. So I think the easiest thing for me is like, if I could toss into a Google drive
127 00:11:55.910 ⇒ 00:12:01.030 Uttam Kumaran: and those and based on new files with Google drive they get, they put into a vector store.
128 00:12:01.360 ⇒ 00:12:04.130 Uttam Kumaran: That’s ideal like that would be
129 00:12:04.420 ⇒ 00:12:05.250 Uttam Kumaran: perfect.
130 00:12:05.600 ⇒ 00:12:11.060 Uttam Kumaran: Right? Cause. I can do a whole host of different ways to get stuff into there. And there’s Gonna be Pdfs videos.
131 00:12:11.230 ⇒ 00:12:12.380 Uttam Kumaran: all that.
132 00:12:12.460 ⇒ 00:12:14.689 Uttam Kumaran: And then basically what you have is you have
133 00:12:14.930 ⇒ 00:12:19.660 Uttam Kumaran: I? I don’t even know whether you need ad. You probably just need a pool
134 00:12:19.940 ⇒ 00:12:24.320 Uttam Kumaran: for every client and an agent for every client, or maybe like
135 00:12:24.910 ⇒ 00:12:27.290 Uttam Kumaran: a tool for every single.
136 00:12:27.860 ⇒ 00:12:32.860 Uttam Kumaran: a tool that can just select. You just take an input client and then select the right knowledge base.
137 00:12:36.050 ⇒ 00:12:44.770 joshuadeveyra: Yeah, yeah, basically, it’s still gonna be the same. You know, architecture where one agent, one kb, one tool. And then
138 00:12:44.940 ⇒ 00:12:49.139 joshuadeveyra: the only thing we’re changing is like how we update the knowledge base.
139 00:12:54.550 ⇒ 00:12:55.810 Uttam Kumaran: So she
140 00:13:02.790 ⇒ 00:13:03.649 Uttam Kumaran: trying to think.
141 00:13:03.972 ⇒ 00:13:12.369 joshuadeveyra: You did mention, I think, yeah, yeah, you did mention yesterday that there’s a way in Google drive like you can use a Google drive as.
142 00:13:12.370 ⇒ 00:13:15.729 Uttam Kumaran: Yeah, I went to this workshop. But like,
143 00:13:16.660 ⇒ 00:13:23.079 Uttam Kumaran: yeah, the vert like all the vertex stuff and they’re doing is a little bit doesn’t work really? Well, it’s really complicated.
144 00:13:23.420 ⇒ 00:13:23.990 Uttam Kumaran: Yeah, very.
145 00:13:25.470 ⇒ 00:13:33.540 Uttam Kumaran: yeah. So even the guy I talked to was like, Yeah, it’s not that good. And I was like, you work for Google, like that’s kind of basically was like.
146 00:13:33.590 ⇒ 00:13:37.140 Uttam Kumaran: I basically asked them, like, you’re really smart like, why do you still work for this company?
147 00:13:37.410 ⇒ 00:13:41.100 Uttam Kumaran: He said, he’s on visa, so he doesn’t. A choice but
148 00:13:42.274 ⇒ 00:13:44.506 Uttam Kumaran: fair answer. But
149 00:13:45.820 ⇒ 00:13:51.170 Uttam Kumaran: yeah, I I think maybe down the line we can consider it. But at this point I don’t really care.
150 00:13:51.390 ⇒ 00:13:54.210 Uttam Kumaran: I think we should try to do something like
151 00:13:54.799 ⇒ 00:13:58.639 Uttam Kumaran: this, which I’ll send in the automation channel, which is like
152 00:13:58.990 ⇒ 00:14:03.629 Uttam Kumaran: hook up relevance to super base. Somehow, I don’t know. There’s a probably
153 00:14:03.640 ⇒ 00:14:06.570 Uttam Kumaran: I think the relevance has an Api.
154 00:14:06.680 ⇒ 00:14:10.401 Uttam Kumaran: Yeah, maybe you can ask the relevance guys in slack, too.
155 00:14:12.140 ⇒ 00:14:16.800 Uttam Kumaran: I just basically want to use an external vector, store, because that’s the only way we’re going to be able to put in.
156 00:14:17.270 ⇒ 00:14:18.050 joshuadeveyra: Yeah.
157 00:14:18.050 ⇒ 00:14:20.990 Uttam Kumaran: Any sort of documents connected to Google drive so
158 00:14:21.870 ⇒ 00:14:23.689 Uttam Kumaran: but poke around and let me know.
159 00:14:24.220 ⇒ 00:14:25.210 joshuadeveyra: Okay. Yeah.
160 00:14:27.320 ⇒ 00:14:30.150 joshuadeveyra: But we prioritize still. Vitaco. Right do we have like
161 00:14:30.580 ⇒ 00:14:31.200 joshuadeveyra: shut up.
162 00:14:32.640 ⇒ 00:14:33.170 joshuadeveyra: Not today.
163 00:14:33.170 ⇒ 00:14:35.020 Uttam Kumaran: Pretty. We’re pretty far ahead.
164 00:14:35.700 ⇒ 00:14:36.550 joshuadeveyra: Oh, yeah.
165 00:14:36.910 ⇒ 00:14:40.419 Uttam Kumaran: And these are all small updates. So hopefully, that doesn’t take like so long.
166 00:14:42.260 ⇒ 00:14:45.810 Nicolas Sucari: We, we should aim to share something on Monday’s meeting. Yeah.
167 00:14:45.810 ⇒ 00:14:51.009 Uttam Kumaran: Yeah. So I know, yeah, let’s probably aim to just share what we updated, at least by next meeting.
168 00:14:53.940 ⇒ 00:15:03.390 Nicolas Sucari: I mean, if we if we can. Yeah, if we can get the new list of source and some new Google sheet with all the information there. That will be fine for next meeting. I think.
169 00:15:04.230 ⇒ 00:15:05.730 joshuadeveyra: I think we already have no.
170 00:15:06.270 ⇒ 00:15:07.445 Uttam Kumaran: You just need to really
171 00:15:08.300 ⇒ 00:15:10.940 Uttam Kumaran: yeah, you just need to add those new columns. And then
172 00:15:11.050 ⇒ 00:15:16.419 Uttam Kumaran: I also wanna test that we can. I wanna I wanna see, like, if we can get the
173 00:15:16.590 ⇒ 00:15:21.260 Uttam Kumaran: like if you run it for the 60 I want to know, kind of like, how much it’s gonna cost, basically. So like, if we can.
174 00:15:21.260 ⇒ 00:15:22.030 joshuadeveyra: Okay.
175 00:15:22.170 ⇒ 00:15:24.080 joshuadeveyra: Okay. Okay. Yeah. Yeah. Sure. I’ll.
176 00:15:24.080 ⇒ 00:15:28.089 Uttam Kumaran: So run it for the 16. Give me a sense of like what sort of timing, and then
177 00:15:28.150 ⇒ 00:15:30.019 Uttam Kumaran: also like the price.
178 00:15:30.507 ⇒ 00:15:35.060 Uttam Kumaran: And that way we can bring that up in that call, because there’ll be they’ll be covering it as well.
179 00:15:36.530 ⇒ 00:15:37.600 joshuadeveyra: Okay, okay, yeah.
180 00:15:38.040 ⇒ 00:15:39.670 joshuadeveyra: Yeah. So no. The.
181 00:15:39.670 ⇒ 00:15:41.160 Uttam Kumaran: Listing.
182 00:15:41.450 ⇒ 00:15:42.910 Uttam Kumaran: Yeah, I think the only.
183 00:15:42.910 ⇒ 00:15:44.500 joshuadeveyra: To search the store one by one.
184 00:15:45.810 ⇒ 00:15:46.135 Uttam Kumaran: Yeah.
185 00:15:47.120 ⇒ 00:15:51.029 Uttam Kumaran: I think the only thing otherwise is tomorrow we have a meeting
186 00:15:51.550 ⇒ 00:15:53.395 Uttam Kumaran: on the AI
187 00:15:55.920 ⇒ 00:15:57.869 Uttam Kumaran: the Pr review automation.
188 00:15:58.415 ⇒ 00:16:00.609 Uttam Kumaran: That Ryan’s gonna be presenting on
189 00:16:02.000 ⇒ 00:16:08.930 Uttam Kumaran: So that’s gonna be our kind of like our 1st major thing we’ve done with like automating on the engineering side. So that’s why, Miguel, I’m excited for you to see that. Because.
190 00:16:10.720 ⇒ 00:16:14.889 Uttam Kumaran: it’s definitely like it’s it’s definitely using github more heavily in workflows. But
191 00:16:15.090 ⇒ 00:16:25.790 Uttam Kumaran: you know, pretty similar, like we have. We have prompts and stuff like that. So I kind of want to fold those efforts into any of the efforts we’re doing on the AI side. Internally, I don’t want to kind of treat that as like so separate.
192 00:16:25.800 ⇒ 00:16:32.039 Uttam Kumaran: but also that falls under like, you know, stuff. I know that Patrick’s working on and like the whole developer experience. So
193 00:16:32.070 ⇒ 00:16:34.219 Uttam Kumaran: I think I’m excited for that. And then.
194 00:16:34.230 ⇒ 00:16:38.847 Uttam Kumaran: yeah, I know we have another meeting Friday. If we don’t have any updates by Friday we can cancel
195 00:16:39.390 ⇒ 00:16:41.469 Uttam Kumaran: But let’s just leave it for now.
196 00:16:42.141 ⇒ 00:16:46.119 Uttam Kumaran: I think the only other thing, Nico, you were gonna intro me
197 00:16:46.390 ⇒ 00:16:50.400 Uttam Kumaran: to folks on your side. So you have. My! My calendar is like
198 00:16:50.830 ⇒ 00:16:55.210 Uttam Kumaran: pretty busy tomorrow, but I’m fine tomorrow or Friday, whatever you want to do
199 00:16:58.260 ⇒ 00:17:02.259 Uttam Kumaran: The only other thing is on the sales side.
200 00:17:02.860 ⇒ 00:17:12.389 Uttam Kumaran: Erickson actually is doing a lot of he’s basically at all of our lead scoring. He’s actually started to leverage. And it’s working really, really. Well, we haven’t sent out any emails yet. But
201 00:17:13.180 ⇒ 00:17:19.760 Uttam Kumaran: he has implemented like the technology scoring the title, scoring the industry and the revenue scoring.
202 00:17:20.266 ⇒ 00:17:27.130 Uttam Kumaran: Which has been. It’s been awesome to see that that’s happening in clay, and then it’s getting stuff is getting enriched.
203 00:17:28.260 ⇒ 00:17:32.140 Nicolas Sucari: Do? Do we have access to play to see all of that things.
204 00:17:32.440 ⇒ 00:17:35.816 Uttam Kumaran: Yeah, I can invite everybody to play
205 00:17:37.761 ⇒ 00:17:45.339 Uttam Kumaran: mainly. I’m just trying to push them as fast as I can to get the emails out. So we can start to basically have feedback. But
206 00:17:45.540 ⇒ 00:17:53.439 Uttam Kumaran: all the work we did on the lead list side and then putting together this stuff for the 1st industry worked. The only things that I re I realize that they can do now is actually
207 00:17:53.460 ⇒ 00:18:01.789 Uttam Kumaran: he’s able to take like, for example, one thing I’m gonna say, is like, Hey, we work with 4 parts to go basically write a campaign that targets all of similar companies to them.
208 00:18:02.000 ⇒ 00:18:05.380 Uttam Kumaran: So he can actually do lookalike audiences really easy.
209 00:18:05.773 ⇒ 00:18:13.630 Uttam Kumaran: So that’s gonna be, basically, we need to have a theme for every campaign. But then I’m writing down kind of what are the components of a campaign that he needs?
210 00:18:14.207 ⇒ 00:18:18.082 Uttam Kumaran: These are gonna be things like, what titles, what technologies?
211 00:18:18.640 ⇒ 00:18:21.970 Uttam Kumaran: what what industry, what revenue size
212 00:18:22.090 ⇒ 00:18:23.958 Uttam Kumaran: to kind of go after. But
213 00:18:24.810 ⇒ 00:18:28.529 Uttam Kumaran: cool he’ll be working on that. So yeah, it’s all.
214 00:18:29.010 ⇒ 00:18:32.070 Uttam Kumaran: It’s all in motion. So okay, that’s all I had.
215 00:18:32.300 ⇒ 00:18:38.039 Nicolas Sucari: Do we have? Yeah, do we have like a weekly meeting with them? Or.
216 00:18:40.920 ⇒ 00:18:42.000 Uttam Kumaran: We don’t
217 00:18:42.345 ⇒ 00:18:48.719 Uttam Kumaran: we usually we should have one. We we did have one earlier this week, but in that in the Channel he’s been sending Wayne.
218 00:18:49.335 ⇒ 00:18:49.840 Uttam Kumaran: So
219 00:18:49.880 ⇒ 00:18:54.352 Uttam Kumaran: yeah, I’m just like the nice thing is, I think,
220 00:18:54.920 ⇒ 00:19:10.250 Uttam Kumaran: he’s been able to send a ton of Updates. So I’ve been able to kind of handle a bunch of stuff, Async, which is the nice but we’ll probably end up before we before we send any emails. I’m gonna do a meeting so we can review the copy. And I wanna invite like everybody here.
221 00:19:10.250 ⇒ 00:19:10.520 Nicolas Sucari: I am.
222 00:19:10.520 ⇒ 00:19:23.729 Uttam Kumaran: As well to kind of get to get a sense of like what that process is just for like educational reasons, but also just to like double spot check emails and stuff like that. But then, yeah, that’ll probably be on Tuesday or Monday.
223 00:19:24.430 ⇒ 00:19:25.220 Nicolas Sucari: Excellent.
224 00:19:25.350 ⇒ 00:19:26.140 Nicolas Sucari: great.
225 00:19:28.980 ⇒ 00:19:29.590 Uttam Kumaran: Cool.
226 00:19:29.910 ⇒ 00:19:31.819 Uttam Kumaran: Okay. Anything. Else.
227 00:19:33.450 ⇒ 00:19:35.360 Nicolas Sucari: Don’t think so, for now no
228 00:19:35.882 ⇒ 00:19:43.659 Nicolas Sucari: anything new from the meeting from with Ben. Sorry we’ve done on pull parts. Anything regarding AI.
229 00:19:43.790 ⇒ 00:19:56.622 Uttam Kumaran: Yeah, I’m writing down. Yeah, none of nothing is related to AI on what he called me about but they’re doing some sort of like skew related analysis. So I’m just writing down what he told me.
230 00:19:56.960 ⇒ 00:19:58.069 Nicolas Sucari: I’ll send it. Commit.
231 00:19:58.070 ⇒ 00:19:59.513 Uttam Kumaran: I’ll send another channel with the notes.
232 00:20:01.030 ⇒ 00:20:01.990 Nicolas Sucari: Excellent. Okay.
233 00:20:02.780 ⇒ 00:20:03.630 Nicolas Sucari: Great.
234 00:20:04.510 ⇒ 00:20:05.100 Uttam Kumaran: Okay.
235 00:20:05.240 ⇒ 00:20:05.990 Uttam Kumaran: Cool.
236 00:20:06.880 ⇒ 00:20:07.730 Uttam Kumaran: Alright.
237 00:20:07.890 ⇒ 00:20:08.980 Uttam Kumaran: thanks. Guys.
238 00:20:09.240 ⇒ 00:20:09.939 Uttam Kumaran: Sounds good. Thank you.
239 00:20:09.940 ⇒ 00:20:10.290 Nicolas Sucari: Guys
240 00:20:11.590 ⇒ 00:20:12.170 Nicolas Sucari: got it?