Meeting Title: DnG Demo Prep Call Date: 2026-03-02 Meeting participants: Uttam Kumaran, Pranav Narahari, Luke Scorziell
WEBVTT
1 00:02:47.460 ⇒ 00:02:48.500 Pranav Narahari: Hey!
2 00:02:49.730 ⇒ 00:02:50.470 Uttam Kumaran: Hey!
3 00:02:54.230 ⇒ 00:02:58.589 Pranav Narahari: I think Luke just messaged me saying he’s gonna join in a little bit. He’s rewatching the meeting.
4 00:02:59.050 ⇒ 00:02:59.839 Uttam Kumaran: Okay, cool.
5 00:03:00.140 ⇒ 00:03:00.720 Pranav Narahari: Yeah.
6 00:03:09.250 ⇒ 00:03:13.839 Pranav Narahari: Yeah, I have a few different things that I was gonna show.
7 00:03:13.970 ⇒ 00:03:14.860 Pranav Narahari: Cool.
8 00:03:17.360 ⇒ 00:03:19.479 Pranav Narahari: I’m not in love with, like.
9 00:03:19.640 ⇒ 00:03:29.680 Pranav Narahari: what Claude, like, spit out on the first run, like, it just… it gets the message… it gets, like, what we’re trying to demo across, but it just looks kind of, like, janky.
10 00:03:30.110 ⇒ 00:03:31.529 Uttam Kumaran: Okay.
11 00:03:32.510 ⇒ 00:03:34.569 Pranav Narahari: Let me just,
12 00:03:35.760 ⇒ 00:03:41.140 Pranav Narahari: I was, like, in the middle of, like, also talking to Clarence about showing, like, vicinity.
13 00:03:41.970 ⇒ 00:03:42.590 Uttam Kumaran: Cool.
14 00:03:46.500 ⇒ 00:03:48.720 Pranav Narahari: And so there’s just, like, some ports issues.
15 00:03:50.990 ⇒ 00:03:51.870 Luke Scorziell: You know.
16 00:03:53.140 ⇒ 00:03:54.279 Pranav Narahari: Hey, Luke.
17 00:03:54.640 ⇒ 00:04:00.040 Luke Scorziell: I was rewatching our… Or a thing on…
18 00:04:00.040 ⇒ 00:04:00.840 Uttam Kumaran: 2X.
19 00:04:01.990 ⇒ 00:04:04.789 Luke Scorziell: That’s helpful, actually. How are you guys doing?
20 00:04:07.150 ⇒ 00:04:10.130 Pranav Narahari: We’re just talking about, like, what,
21 00:04:10.420 ⇒ 00:04:14.240 Pranav Narahari: like, the demo. Let me just spin up, like…
22 00:04:14.780 ⇒ 00:04:26.800 Pranav Narahari: So I have vicinity up, which is, like, what Clarence was talking about, maybe we could show. I don’t… I think it might just be a little bit overkill, and might just confuse the client more than actually help.
23 00:04:28.630 ⇒ 00:04:32.390 Uttam Kumaran: Yeah, I probably wouldn’t share that, because it’s not polished, it’s like…
24 00:04:33.610 ⇒ 00:04:38.110 Uttam Kumaran: Yeah, it’s just, like, not polished at all. I don’t think any of the setup is there for his stuff yet.
25 00:04:38.890 ⇒ 00:04:39.680 Pranav Narahari: Yeah.
26 00:04:39.960 ⇒ 00:04:40.880 Pranav Narahari: Okay.
27 00:04:40.880 ⇒ 00:04:44.550 Luke Scorziell: Yeah, I mean, watching… did you build anything yet, Pranav? Or, not yet?
28 00:04:44.890 ⇒ 00:04:47.329 Pranav Narahari: Oh, I did build something,
29 00:04:47.700 ⇒ 00:04:57.319 Pranav Narahari: I just don’t, like, it doesn’t look great, in my opinion, like, what I was doing, and I’ll show it to you in a sec, I’m just trying to…
30 00:04:58.090 ⇒ 00:04:59.770 Pranav Narahari: Run it right now.
31 00:05:02.910 ⇒ 00:05:05.869 Luke Scorziell: Tom, can I get Starlink? Do you guys… will we do that for…
32 00:05:07.140 ⇒ 00:05:10.250 Uttam Kumaran: Why were you going? Oh, you just want… for your house?
33 00:05:12.180 ⇒ 00:05:14.679 Luke Scorziell: My internet sucks, and there’s no other…
34 00:05:14.680 ⇒ 00:05:15.499 Uttam Kumaran: What do you want?
35 00:05:15.500 ⇒ 00:05:16.440 Luke Scorziell: spectrum.
36 00:05:16.690 ⇒ 00:05:17.760 Luke Scorziell: Spectrum.
37 00:05:18.710 ⇒ 00:05:19.680 Uttam Kumaran: and fiber?
38 00:05:20.240 ⇒ 00:05:24.090 Luke Scorziell: No, it’s like, well, yeah, it’s… I think it’s just there.
39 00:05:25.400 ⇒ 00:05:29.479 Luke Scorziell: And whatever cables, so either that, or I’m gonna have to start working somewhere else.
40 00:05:30.390 ⇒ 00:05:34.840 Uttam Kumaran: What, like, is it… are you renting your place, or, like, what’s the options? Can you switch, or no?
41 00:05:35.190 ⇒ 00:05:37.129 Luke Scorziell: No, this is the only…
42 00:05:37.130 ⇒ 00:05:41.630 Uttam Kumaran: Wi-Fi they have set up here. What is your… what is your up-down, like, speed?
43 00:05:43.040 ⇒ 00:05:44.810 Luke Scorziell: I think it…
44 00:05:44.810 ⇒ 00:05:45.470 Uttam Kumaran: You know?
45 00:05:45.470 ⇒ 00:05:47.960 Luke Scorziell: No, I don’t know off the top of my head.
46 00:05:48.470 ⇒ 00:05:51.290 Uttam Kumaran: Your only option is to pay for, like.
47 00:05:52.700 ⇒ 00:05:57.279 Uttam Kumaran: Can they give you the upgrade, or no? Or, like, what… what… yeah, what are they… what are they able to do?
48 00:05:58.820 ⇒ 00:06:04.410 Luke Scorziell: Like, it’s just because they’ve only set up Spectrum here. So, like, at the…
49 00:06:04.630 ⇒ 00:06:05.330 Uttam Kumaran: Oh, okay.
50 00:06:05.330 ⇒ 00:06:12.880 Luke Scorziell: It’s this thing called Starry, and it was super cheap and super dope, but, yeah, so it’s, I mean…
51 00:06:13.230 ⇒ 00:06:15.200 Luke Scorziell: That’s okay, I’ll just use my hotspot.
52 00:06:21.650 ⇒ 00:06:26.900 Pranav Narahari: Okay, can y’all see… this Nova AI search thing.
53 00:06:28.910 ⇒ 00:06:29.560 Luke Scorziell: Yes.
54 00:06:29.560 ⇒ 00:06:39.319 Pranav Narahari: Okay, cool. Yeah, this is just, like, dummy data that I put in. Probably will change this to, like, D&G, AI search.
55 00:06:39.520 ⇒ 00:06:47.929 Pranav Narahari: But, essentially, what I was trying to do here was… first, I was looking into, like, okay, can we just, like, spin up easily, like, a Teams app?
56 00:06:48.190 ⇒ 00:06:53.539 Pranav Narahari: When I was looking into it, it was clawed and…
57 00:06:54.000 ⇒ 00:07:10.640 Pranav Narahari: Chris, we’re basically just saying, like, the faster approach and, like, easier approach to, like, what you’re trying to demo is just, like, let’s just spin up something that looks similar to Teams to get the same message across, and then you can just have everything locally. So…
58 00:07:11.870 ⇒ 00:07:20.969 Pranav Narahari: this is supposed to be Teams. Think of this as Teams. You know, you have your chat here, you have your admin here, you have, like, your apps hub here, I guess.
59 00:07:21.410 ⇒ 00:07:32.550 Pranav Narahari: And then, so you have this, like, chat interface where there’s, like, a sources button up here, and this is where you can authenticate, like, all of your different, like, data sources and toggle them on and off.
60 00:07:32.730 ⇒ 00:07:36.220 Pranav Narahari: So, these are…
61 00:07:37.980 ⇒ 00:07:46.449 Pranav Narahari: Okay, for some reason it created dummy names for, like, these sources as well, so I can re… rewrite these to, like, actually be…
62 00:07:46.610 ⇒ 00:07:51.480 Pranav Narahari: what they mentioned in that call. So, like, Adobe,
63 00:07:52.720 ⇒ 00:08:01.449 Pranav Narahari: I think ClipNote might have been one, I forget. But essentially, yeah, you should have, like, a way to authenticate here. So you click a button, sign in.
64 00:08:04.650 ⇒ 00:08:08.799 Pranav Narahari: This is just kind of showing the flow, and then you can toggle these things on and off.
65 00:08:09.010 ⇒ 00:08:14.669 Pranav Narahari: And then you’ll also see for, like, some users, like, they’re disabled on, like, a role basis.
66 00:08:15.010 ⇒ 00:08:19.410 Pranav Narahari: And then, you can kind of see that, like, in an admin hub, which, like…
67 00:08:20.430 ⇒ 00:08:27.250 Pranav Narahari: only, like, some people would have access to is individual user permissions, overall, like.
68 00:08:27.560 ⇒ 00:08:39.560 Pranav Narahari: Like, high-level data on, like, okay, where on the knowledge bases, like, the last sync, are they enabled or disabled? And then also different connectors.
69 00:08:40.840 ⇒ 00:08:41.400 Luke Scorziell: Whoa.
70 00:08:42.730 ⇒ 00:08:44.810 Pranav Narahari: this is kind of what I felt like
71 00:08:45.170 ⇒ 00:09:03.549 Pranav Narahari: would be the most useful, to demo, just, like, something that looks familiar to them within Teams. And then something simple, just like MCP servers and knowledge bases. Like, this guy was, like, a lot less in the weeds on things than our, guy on Friday.
72 00:09:03.870 ⇒ 00:09:07.540 Pranav Narahari: So, I think… Less is actually kind of more here.
73 00:09:07.780 ⇒ 00:09:08.859 Pranav Narahari: No.
74 00:09:09.940 ⇒ 00:09:11.849 Pranav Narahari: The one thing, though, that…
75 00:09:12.960 ⇒ 00:09:30.989 Pranav Narahari: I feel like maybe we just need to, like, think a little bit more about, like, what does the product look like for him? Because I think that’s where he’s, like, the most skittish, because he doesn’t want, like, a great product that, like, no one’s gonna adopt. So, it has to be something that’s, like, already built on top of the tools that they’re using.
76 00:09:31.230 ⇒ 00:09:36.260 Pranav Narahari: Teams seems like a great option. We also talked about Copilot,
77 00:09:36.410 ⇒ 00:09:39.500 Pranav Narahari: So I don’t know if there’s something we want to, like, demo there.
78 00:09:40.250 ⇒ 00:09:47.299 Pranav Narahari: But, yeah, this is kind of just, like, to start, like, the brainstorming session, and then we can build a little bit more later today and tomorrow.
79 00:09:48.730 ⇒ 00:09:49.400 Luke Scorziell: Yeah.
80 00:09:49.720 ⇒ 00:09:59.289 Luke Scorziell: I mean, I think this is good. I agree with the… I mean, we already talked last week about, I think, showing him too much is… would just be kind of overwhelming.
81 00:10:01.560 ⇒ 00:10:06.229 Luke Scorziell: So I, like, oven is built out of… out of Teams, and then,
82 00:10:06.960 ⇒ 00:10:20.950 Luke Scorziell: Yeah, just listening to the call again, he mentioned specifically, like, connecting with the Adobe tenant being, like, a big pain point, because they spend, like… it sounds like they spend, like, hours going through that, so I think even just solving that one problem for him would be,
83 00:10:23.060 ⇒ 00:10:29.869 Luke Scorziell: like… I feel like that seems convincing, if you can kind of go on Copilot or teams and…
84 00:10:30.290 ⇒ 00:10:37.260 Luke Scorziell: and at and query. I don’t know exactly what he was talking about with the tenant, but it sounds like they have a lot of documents,
85 00:10:37.750 ⇒ 00:10:39.990 Luke Scorziell: That they have to look through.
86 00:10:40.900 ⇒ 00:10:47.429 Luke Scorziell: pretty regularly… regularly, that it could just be automated with AI. So I think if, like, that’s… that’s a use case where
87 00:10:48.160 ⇒ 00:10:56.250 Luke Scorziell: I think I could see it being pretty, easy to implement, as long as we build it within what they have. And then he… it sounded like he would be impressed.
88 00:10:56.560 ⇒ 00:11:00.670 Luke Scorziell: buy some of that, because my gut is that once we figure out one pain point for him.
89 00:11:01.680 ⇒ 00:11:08.270 Luke Scorziell: Then he’ll kind of open up his eyes to other stuff, so…
90 00:11:08.860 ⇒ 00:11:09.750 Pranav Narahari: Yeah.
91 00:11:09.950 ⇒ 00:11:12.419 Pranav Narahari: So, for the demo, too, like…
92 00:11:13.360 ⇒ 00:11:29.640 Pranav Narahari: do we… I don’t think we want to build that connection, right? Because, I mean, that’s probably not going to be possible, like, just to build an app that’s, like… I mean, how would we even spin up, like, that tenant, like, on our end, right? So I think what we’re trying to show is just, like, what does the product look like?
93 00:11:30.180 ⇒ 00:11:30.810 Luke Scorziell: Yeah.
94 00:11:30.810 ⇒ 00:11:32.760 Pranav Narahari: Because I think that’s where he’s most concerned, like.
95 00:11:33.470 ⇒ 00:11:35.829 Pranav Narahari: And then this should also, of course, like.
96 00:11:36.510 ⇒ 00:11:49.570 Pranav Narahari: we should feel confident in, like, building this for them, like, with actual connections to, all of their different SaaS tools, and then within a Teams app. So…
97 00:11:50.910 ⇒ 00:11:54.370 Pranav Narahari: I don’t know, the, the… what I was…
98 00:11:54.480 ⇒ 00:12:08.429 Pranav Narahari: kind of going for for tomorrow’s demo is just, like, showing them, like, this is what the product looks like, to kind of probably put him a little bit more at ease of, like, okay, yeah, I have… I feel pretty confident that the team is going to adopt something like this.
99 00:12:09.090 ⇒ 00:12:14.240 Luke Scorziell: Yeah, okay, that makes sense. So that’s… and that’s where you’re thinking, that just having, like, a Teams…
100 00:12:15.100 ⇒ 00:12:17.629 Luke Scorziell: like, having this within Teams. It’s good.
101 00:12:18.440 ⇒ 00:12:20.079 Pranav Narahari: I think so, yeah.
102 00:12:20.410 ⇒ 00:12:28.509 Pranav Narahari: and then kind of just having, like, maybe, like, dummy data for, let’s say, like, the Adobe tenant. So, like, if you were to ask, like,
103 00:12:29.520 ⇒ 00:12:36.189 Pranav Narahari: just whatever, any question, like, say this had to do with Adobe, like, it would just pop up stuff here.
104 00:12:39.580 ⇒ 00:12:40.380 Luke Scorziell: Oh, cool.
105 00:12:41.810 ⇒ 00:12:43.070 Uttam Kumaran: This is on Wednesday?
106 00:12:43.900 ⇒ 00:12:44.640 Pranav Narahari: This is tomorrow.
107 00:12:45.260 ⇒ 00:12:45.900 Uttam Kumaran: Okay.
108 00:12:46.370 ⇒ 00:12:48.569 Uttam Kumaran: Yeah, it’s gonna be tough to, like.
109 00:12:49.460 ⇒ 00:12:55.460 Uttam Kumaran: I mean, I don’t know, I think, Luke, this is where you just have to be a little bit more clear with, like, what demo you want to put together, because…
110 00:12:55.790 ⇒ 00:12:59.990 Uttam Kumaran: As of now, like, this is too… like, I wouldn’t put this in front of them.
111 00:13:00.580 ⇒ 00:13:06.209 Uttam Kumaran: So… I think you just have to give… print off some, like, clear requirements of, like, what you want.
112 00:13:06.940 ⇒ 00:13:08.639 Uttam Kumaran: To show visually.
113 00:13:08.760 ⇒ 00:13:10.959 Uttam Kumaran: And, like, what you don’t want to show, cause…
114 00:13:11.370 ⇒ 00:13:18.689 Uttam Kumaran: Yeah, I don’t know, that’s my take. The WorkOS stuff is gonna take a little bit longer to, like, get all figured out.
115 00:13:19.860 ⇒ 00:13:24.209 Luke Scorziell: What are you… what are your… I mean, what are you seeing? So this is my first time giving demo, I guess, direction.
116 00:13:24.210 ⇒ 00:13:26.460 Uttam Kumaran: Well, yeah, I mean, I just think, like, this, like…
117 00:13:26.910 ⇒ 00:13:37.180 Uttam Kumaran: I… like, right now, I think Pranav basically was like, cool, I’m gonna… I think we probably just took the transcript and was like, make something that looks like it. Like, right now, this is not… I wouldn’t put this in front of the client.
118 00:13:37.470 ⇒ 00:13:38.380 Pranav Narahari: Yay.
119 00:13:38.380 ⇒ 00:13:40.149 Uttam Kumaran: Right? In terms of a demo.
120 00:13:40.580 ⇒ 00:13:46.870 Uttam Kumaran: Because what are you gonna do? Like, if you… if we walk through the demo, like, let’s say today was tomorrow, and we were just, like, walking through, like.
121 00:13:47.200 ⇒ 00:14:04.129 Uttam Kumaran: I would say this isn’t polished enough. Like, you kind of need to have… a couple things would be, one is, like, think through what you want to demo. It’s easy for us to make dummy data and stuff like that, but what you don’t want to do is to walk in that call and be like, here’s, like, something we mocked up, and then, like, have no path.
122 00:14:04.240 ⇒ 00:14:09.789 Uttam Kumaran: Like, you wanna… what you wanna do is script, like, what the demo is you wanna give them, like.
123 00:14:09.960 ⇒ 00:14:15.989 Uttam Kumaran: Do you want to show, like, they may have labeled a couple of use cases, do you want to show examples of, like, how that can be done?
124 00:14:16.120 ⇒ 00:14:20.570 Uttam Kumaran: The other piece of, like, if you don’t feel like doing a demo at all, just put us all into a DAC.
125 00:14:20.940 ⇒ 00:14:24.989 Uttam Kumaran: And, like, just put screenshots of, like, what things could look like.
126 00:14:25.250 ⇒ 00:14:31.140 Uttam Kumaran: I don’t know, this is my take, based on, like, you know, just what I’m hearing.
127 00:14:32.420 ⇒ 00:14:40.980 Pranav Narahari: Yeah, I was kind of feeling the same thing as I was building this. I was thinking, like, I don’t know if this is something we want to put in front of the client, and…
128 00:14:41.320 ⇒ 00:14:46.670 Pranav Narahari: like, a demo sounds obviously great, and maybe I was even just, like, a more…
129 00:14:46.860 ⇒ 00:14:57.780 Pranav Narahari: optimistic about getting, like, a super polished demo in front of them, like, something that you can click around with live. But maybe, yeah, it is, like, a deck with, like, screenshots that we could
130 00:14:57.880 ⇒ 00:15:04.649 Pranav Narahari: you know, easily create, of, like, different views within a Teams app. And honestly, that might…
131 00:15:04.930 ⇒ 00:15:09.010 Pranav Narahari: Send a message even better to… to the client.
132 00:15:10.590 ⇒ 00:15:16.089 Luke Scorziell: Yeah, well, I mean, so is there a question of whether we can build this or not, or do you feel pretty confident that we can build it?
133 00:15:16.540 ⇒ 00:15:17.179 Uttam Kumaran: Build water.
134 00:15:17.180 ⇒ 00:15:17.540 Pranav Narahari: Yeah, though.
135 00:15:17.540 ⇒ 00:15:20.030 Uttam Kumaran: like, you’re… I don’t even know what we’re trying… like…
136 00:15:20.920 ⇒ 00:15:30.939 Uttam Kumaran: One, there’s no deal on the table, so, like, I’m not willing to say, like, we’re able to build something or not, right? Like, what’s the point of even getting to that? There’s no deal right now.
137 00:15:31.510 ⇒ 00:15:38.380 Luke Scorziell: Well, I think what we talked about last week was just building a… a… app within,
138 00:15:39.260 ⇒ 00:15:43.089 Luke Scorziell: Teams that they can ask back and forth,
139 00:15:43.090 ⇒ 00:15:50.750 Uttam Kumaran: But, like, there’s… there’s no deal, so why would… what are we gonna build right now? Like, we could do… I think long story short is yes, but, like.
140 00:15:50.930 ⇒ 00:15:57.369 Uttam Kumaran: Building it for the demo, It’s much different than building it for, like, once they sign a deal.
141 00:16:00.720 ⇒ 00:16:05.220 Uttam Kumaran: Like, we’re not… like, I am not gonna sign up for us to build shit for…
142 00:16:05.590 ⇒ 00:16:11.649 Uttam Kumaran: the demo, like, so that’s what I think you need to think about, like, what is convincing for the next sales call?
143 00:16:12.700 ⇒ 00:16:18.560 Uttam Kumaran: like, spinning it up in Teams and stuff is gonna take time. Like, we don’t have a ready-to-go demo for that right now.
144 00:16:19.430 ⇒ 00:16:20.080 Luke Scorziell: Yeah.
145 00:16:21.170 ⇒ 00:16:26.630 Luke Scorziell: Yeah, I mean, I think in the case… like, he literally, I think, just wanted to see on the call
146 00:16:27.400 ⇒ 00:16:34.049 Luke Scorziell: what it even looked like to query, I think, an MCP server. So I don’t even think that he had very high, like.
147 00:16:34.240 ⇒ 00:16:40.100 Luke Scorziell: expectations for… so I, like, we were literally on the call, and Pranav, I think, was…
148 00:16:40.360 ⇒ 00:16:53.030 Luke Scorziell: would have been able to pull up, like, a general MCP server, and show. So I don’t think he has huge expectations, so if it’s literally just took… we can walk something up with dummy data and do that, just so he can see how it works.
149 00:16:53.480 ⇒ 00:17:02.140 Uttam Kumaran: But I guess, like, what is… what is it? Like, I’m even trying to get to, like, what are we trying to show? Like, is this a web page with, like, a chat? Is it, like…
150 00:17:02.680 ⇒ 00:17:05.630 Uttam Kumaran: Yeah, like, what is… what are we trying to demo?
151 00:17:06.069 ⇒ 00:17:13.659 Uttam Kumaran: like, I hear you on the functionality, like, we want a demo talking to an MCP server, but is that on a website? Can that just be a slide?
152 00:17:15.160 ⇒ 00:17:18.110 Uttam Kumaran: Like, what is a… what would be effective?
153 00:17:20.069 ⇒ 00:17:20.959 Pranav Narahari: One thing I’m thinking of…
154 00:17:20.960 ⇒ 00:17:27.520 Uttam Kumaran: You can also just pull up, like, your cursor and show, like, how chatting with an MCP works. So, like, I’m trying to get to the sense of, like.
155 00:17:27.520 ⇒ 00:17:31.380 Luke Scorziell: What is it… what is it… what is it you even need for a demo? Like…
156 00:17:31.380 ⇒ 00:17:38.920 Uttam Kumaran: what is unique about this client that we have to… like, what… what is it… what would be the convincing demo? Again, like, your options are, like.
157 00:17:39.040 ⇒ 00:17:45.079 Uttam Kumaran: For some clients, we have literally put together, like, a fake app, where everything’s kind of, like.
158 00:17:45.860 ⇒ 00:17:50.459 Uttam Kumaran: hard-coded, and it’s, like, dummy data, but it looks clickable, right? That’s, like, the max.
159 00:17:50.950 ⇒ 00:17:51.430 Pranav Narahari: Yeah.
160 00:17:51.430 ⇒ 00:18:05.639 Uttam Kumaran: in 24 hours, I don’t think we’re gonna get there right now, because the requirements are still super loose. So your options are like, okay, well, can we just do, like, some simple… some better, like, UI mocks, and, like, put that into a slide, right?
161 00:18:06.010 ⇒ 00:18:08.899 Uttam Kumaran: And then even simpler, you could, like, hey, you know.
162 00:18:09.260 ⇒ 00:18:14.599 Uttam Kumaran: we can… our team is totally capable of developing MCPs that you need.
163 00:18:14.960 ⇒ 00:18:24.730 Uttam Kumaran: like, of course, we’re not… we can’t, like, do that just for this demo, those things take time, so here is an example of, like, how you could use an MCP server in cursor.
164 00:18:24.860 ⇒ 00:18:34.480 Uttam Kumaran: And, like, we’re actively using this, so, like, I would think about it in that… in that sense. Like, later on, it will get easier and easier to do these… to develop these demos, but, like.
165 00:18:34.820 ⇒ 00:18:42.510 Uttam Kumaran: we need… yeah, I don’t think it’s… I think just showing, like, a broad platform that we vibeco the whole thing is gonna fall flat.
166 00:18:43.010 ⇒ 00:18:43.740 Pranav Narahari: Yeah.
167 00:18:44.070 ⇒ 00:18:54.630 Pranav Narahari: So, my understanding, too, like, from the call we had, like, on Tuesday, was he didn’t even really know what an MCP server was, he didn’t know what, like, a knowledge base was, and so…
168 00:18:54.790 ⇒ 00:19:05.540 Pranav Narahari: I think, yeah, what Tom was saying, just, like, if we just pull up Cursor, if we just create, like, a custom GPT with just, like, throwing in documents in there and showing that, like, yeah, you can query these things, like, even if you have, like.
169 00:19:06.110 ⇒ 00:19:12.539 Pranav Narahari: thousands of files, like, you can still query them, with natural language. I think that would be…
170 00:19:12.820 ⇒ 00:19:15.110 Pranav Narahari: a good enough demo.
171 00:19:15.110 ⇒ 00:19:25.550 Luke Scorziell: Yeah, I think that’d be great. I mean, because the two things that you told him on the call was, yeah, that we could show them the MCP server and then the RAG, like, how those two things work, so I think…
172 00:19:25.930 ⇒ 00:19:27.890 Luke Scorziell: I guess, yeah, maybe it’s…
173 00:19:28.120 ⇒ 00:19:34.059 Luke Scorziell: me, from the conceptual, like, place that I’m at, to… you are wanting… you need more of, like, a…
174 00:19:34.380 ⇒ 00:19:37.300 Luke Scorziell: Like, line for line, this is what we…
175 00:19:37.590 ⇒ 00:19:47.560 Luke Scorziell: what we need, so… so I don’t know, maybe… maybe we can help translate that between the two, because I think if it’s just pulling up cursor, and I think a cool name for this would be, like, Slingshot.
176 00:19:47.880 ⇒ 00:19:52.569 Luke Scorziell: Because it’s, like, David and Goliath, and David kills Goliath, the slingshot.
177 00:19:53.850 ⇒ 00:20:00.540 Luke Scorziell: And so, if it’s, like, something like that, and then we just, yeah, put together, like, a mock-up
178 00:20:00.970 ⇒ 00:20:04.099 Luke Scorziell: Of, like, you know, here’s us showing this to you, and, like.
179 00:20:04.570 ⇒ 00:20:12.320 Luke Scorziell: Cursor, which wouldn’t really be what you guys would use. And then we just show, like, here’s what it could look like, and…
180 00:20:16.120 ⇒ 00:20:22.139 Luke Scorziell: Teams, or we could do, like, a screen recording, even, of, like, us querying, like, the Brainforge bot, I guess, and…
181 00:20:22.300 ⇒ 00:20:27.179 Luke Scorziell: Slack, I don’t know if that’s the same… same relying technology.
182 00:20:28.460 ⇒ 00:20:35.080 Pranav Narahari: Yeah, I think it would show the… the product in a very similar way, right? It’s just, like, Teams versus Slack.
183 00:20:36.580 ⇒ 00:20:39.880 Uttam Kumaran: Yeah, I mean, but I mean, like, again, like, I still think a lot of… I still…
184 00:20:40.490 ⇒ 00:20:44.460 Uttam Kumaran: I agree with some of that, but, like, it’s still not clear, like, what the demo is, like…
185 00:20:44.770 ⇒ 00:20:50.419 Uttam Kumaran: what are you planning on showing? Like, for example, you could just pull up ChatGPT and, like.
186 00:20:51.140 ⇒ 00:21:02.150 Uttam Kumaran: this is, like, probably a good enough demo for you to be like, we could develop you something like this that has all your stuff in it, right? Like, so what’s… so my challenge, if I would say it a different way, is, like, why isn’t this… this just a demo?
187 00:21:04.160 ⇒ 00:21:05.889 Pranav Narahari: Yeah, I think…
188 00:21:06.000 ⇒ 00:21:10.979 Pranav Narahari: this could just be the demo. Besides, it just doesn’t have their branding, so it just doesn’t…
189 00:21:11.320 ⇒ 00:21:16.889 Uttam Kumaran: No, no, no, but, like, that’s what I’m saying, is, like, what are you trying to get across them? Are you trying to get across to them that, like.
190 00:21:16.960 ⇒ 00:21:31.650 Uttam Kumaran: this is how it would work. Are you trying to get across to them that, like, this is a… this is… this would be the platform we would build? So, like, what I’m not hearing is, like, what we need to show for them. So, are your requirements, we need to put together a branded
191 00:21:32.150 ⇒ 00:21:33.270 Uttam Kumaran: demo.
192 00:21:33.510 ⇒ 00:21:41.810 Uttam Kumaran: where you’re showing… toggling MCPs and asking a question over this. If that’s it, then I don’t think you’re gonna be able to do this by tomorrow.
193 00:21:42.470 ⇒ 00:21:46.000 Uttam Kumaran: I don’t think we can… I mean, like, yeah, I guess, like, I…
194 00:21:46.360 ⇒ 00:21:48.920 Uttam Kumaran: Yeah, I think it’s gonna be hard for tomorrow.
195 00:21:52.020 ⇒ 00:21:52.380 Pranav Narahari: I’m trying.
196 00:21:52.380 ⇒ 00:21:53.769 Uttam Kumaran: Because we just… yeah.
197 00:21:54.010 ⇒ 00:22:00.110 Pranav Narahari: I felt like the… the client really just didn’t know that this was possible, that you can, like…
198 00:22:00.360 ⇒ 00:22:03.459 Pranav Narahari: integrate… like, he didn’t know what an MCP was.
199 00:22:03.600 ⇒ 00:22:08.919 Pranav Narahari: So it’s just like… showing this, in and of itself, I think just is like…
200 00:22:09.910 ⇒ 00:22:14.900 Pranav Narahari: this shows that, okay, this is actually, like, something that is possible with AI, and so…
201 00:22:15.520 ⇒ 00:22:29.709 Pranav Narahari: if we were to just show ChatGPT like this, just, like, yeah, having, like, a company knowledge base, and, like, having an MCP server enabled, and just, like, being able to ask it questions, and it’s… he’s able to see that it’s… it’s working, based on, like, how he described it, like, last week.
202 00:22:30.210 ⇒ 00:22:33.880 Pranav Narahari: I think that’s a pretty good demo.
203 00:22:36.440 ⇒ 00:22:41.549 Pranav Narahari: Yeah, Luke, do you disagree with that? Do you think, like, he would still be like, no, we’re looking for something else?
204 00:22:42.690 ⇒ 00:22:44.140 Luke Scorziell: No, I think he would be…
205 00:22:45.020 ⇒ 00:22:49.049 Luke Scorziell: Yeah, I think he would be pretty impressed, honestly, by that.
206 00:22:49.960 ⇒ 00:22:53.220 Luke Scorziell: So, I mean, my… the… yeah, the main concern that he had was
207 00:22:53.370 ⇒ 00:22:59.610 Luke Scorziell: adoption, and so that’s just why I was thinking if we could build something within teams, but if it’s…
208 00:22:59.720 ⇒ 00:23:00.639 Luke Scorziell: Like, going kind of the.
209 00:23:00.640 ⇒ 00:23:02.529 Uttam Kumaran: Yeah, it’s gonna take way too long, dude.
210 00:23:02.780 ⇒ 00:23:04.259 Uttam Kumaran: That’s, like, actual work.
211 00:23:06.380 ⇒ 00:23:10.259 Uttam Kumaran: Like, we don’t have, like, ready-to-go Teams demos, basically.
212 00:23:11.260 ⇒ 00:23:21.770 Luke Scorziell: Yeah, which, I mean, I think it was just a miscommunication, like, I didn’t… me not the exact scope of, like… like a… if it’s, like, I don’t want you to have spent hours or something on this, it’s like…
213 00:23:22.450 ⇒ 00:23:27.480 Luke Scorziell: It seemed like… like, I think just from the conversations that Pradav and I had, it seemed like it was…
214 00:23:27.900 ⇒ 00:23:33.119 Luke Scorziell: not gonna be, like, that big of a lift, so I guess knowing that it is a bigger lift.
215 00:23:34.760 ⇒ 00:23:47.249 Pranav Narahari: Yeah, I think I was… it’s on me, too, like, I was thinking from the perspective of, like, what we already had with Lilo, like, okay, just, like, changing, basically, the theme on that, just having,
216 00:23:47.620 ⇒ 00:23:56.400 Pranav Narahari: you know, no, like, client information on there, and then just putting in, like, a DNG, like, logo or whatever, we could…
217 00:23:56.780 ⇒ 00:24:01.849 Pranav Narahari: that would be one thing, and we already had, like, built an entire app on top of that.
218 00:24:01.980 ⇒ 00:24:03.819 Pranav Narahari: But yeah, if we want to build it, integrate.
219 00:24:03.820 ⇒ 00:24:06.340 Uttam Kumaran: Yeah, but dude, even that, I think, is gonna take, like, time.
220 00:24:07.800 ⇒ 00:24:08.210 Pranav Narahari: Yeah.
221 00:24:08.210 ⇒ 00:24:23.730 Uttam Kumaran: I think one… I think one thing, yeah, I think you just overshot how long this is gonna take. That’s fine. I think second thing is Clarence. I wouldn’t, and even if Clarence is here, I would say this. Clarence is… thinks this takes… is gonna be way faster. The difference is you guys aren’t capable
222 00:24:24.150 ⇒ 00:24:39.660 Uttam Kumaran: of, like, and neither am I right now, of using his work platform to do any of this shit that he said. So, I would not… I would not, like, rely on that. I would not spend any of your time, like, in that platform doing anything until, like, I… I’m, like, figuring out
223 00:24:39.680 ⇒ 00:24:49.290 Uttam Kumaran: how to do all the things that he’s mentioning, but he’s gonna be aggressive with, like, oh, go ahead and use that to do stuff. He’s not trained, you guys, and you guys haven’t done it, so don’t bank on that.
224 00:24:49.420 ⇒ 00:24:50.300 Uttam Kumaran: Like…
225 00:24:50.490 ⇒ 00:24:58.820 Uttam Kumaran: just hear what he has to say about that, but, like, you guys are actively selling shit today that needs to work today, that’s not like a… it can’t be, like, fake.
226 00:24:58.910 ⇒ 00:25:14.430 Uttam Kumaran: Or it can’t be, like, oh yeah, you should just go ahead and use it, and, like, you haven’t touched it in the demos tomorrow. So, like, do something that’s gonna work today. You can always use slides. So again, like, worst case, Luke, just, like, use Gemini and just make the fucking slides the UI that you want to use.
227 00:25:14.730 ⇒ 00:25:20.850 Uttam Kumaran: And then… then you can take that and be like, Pranav, can you get me anywhere clickable? Can you… can you turn this clickable?
228 00:25:21.660 ⇒ 00:25:29.170 Uttam Kumaran: Okay. So, like, I would say this is where, like, in working, like, and this is where, like, again, I think, think about separation responsibilities.
229 00:25:29.360 ⇒ 00:25:37.369 Uttam Kumaran: all three of us should not be thinking of what do we demo to the client. Like, Luke, you’re the salesperson, you think about what that demo is.
230 00:25:37.450 ⇒ 00:25:54.079 Uttam Kumaran: try your best to come up with the lowest fidelity, like, most static thing first, so that no matter what, even if none of our shit that we’re gonna promise to you gets out the door, you control your own destiny. And that’s not… that’s not a ding on me, or Pranav or our team, but, like.
231 00:25:54.370 ⇒ 00:26:01.759 Uttam Kumaran: you are the salesperson, if you rely on us, and then that shit’s not ready, then you’re screwed. So, like, control your own destiny.
232 00:26:01.940 ⇒ 00:26:08.139 Uttam Kumaran: Try to just have slides with, like, the thing that you want to pitch, right? You already have Cursor, you kind of know how it works.
233 00:26:08.140 ⇒ 00:26:22.389 Uttam Kumaran: I think it’ll be pretty easy for you to either do two things. One, you can use Gemini or whatever to do slides. You could also just have Cursor, or lovable, or whatever, just, like, produce something quickly. So do that. Do what’s in your wheelhouse, and then be like, hey guys, I worked on this.
234 00:26:22.500 ⇒ 00:26:30.689 Uttam Kumaran: it’s like, I wish it could be, like, a little bit more real in this way, can you help me? And then that’ll be easy to pass to Pranav, right? Versus from our side.
235 00:26:30.970 ⇒ 00:26:36.509 Uttam Kumaran: We’re thinking about, like, if we have to build something real, there’s all these pieces of it that, like.
236 00:26:36.860 ⇒ 00:26:44.320 Uttam Kumaran: It’s gonna get overwhelming, so if you give us that, like, first pass, and then you’re like, make this alive, okay, like, we can go from that.
237 00:26:45.160 ⇒ 00:26:49.410 Luke Scorziell: Okay, yeah, I mean, I can… I can do that. I think I was just, like, Yeah, so…
238 00:26:49.410 ⇒ 00:26:53.990 Uttam Kumaran: No, no, I think this is this… this is what we were, like, I wanted to see, like, how this process is gonna work.
239 00:26:54.150 ⇒ 00:27:00.079 Uttam Kumaran: We’re not yet at the point where you can do… you can generate demos end-to-end by yourself.
240 00:27:00.300 ⇒ 00:27:06.420 Uttam Kumaran: So, but you can develop slides and images and sort of, like, levelables, you already have done that, so, like.
241 00:27:06.570 ⇒ 00:27:08.160 Uttam Kumaran: Bank on that first.
242 00:27:09.210 ⇒ 00:27:19.739 Uttam Kumaran: See how far that gets you, and then… and then be like, hey, this was okay, but, like, it’s not sharp enough, or it doesn’t have a branding, or, like, it needs fake data, because then we’ll enable that.
243 00:27:19.860 ⇒ 00:27:22.359 Uttam Kumaran: Well, we can do, we can take it from there, for sure.
244 00:27:22.770 ⇒ 00:27:26.489 Luke Scorziell: Okay. Yeah, so, I mean, if that’s the flow, then,
245 00:27:28.320 ⇒ 00:27:35.050 Luke Scorziell: Yeah, that helps me, because then I can… I mean, I could probably just put together, like, a lovable that looks cool and is branded.
246 00:27:35.740 ⇒ 00:27:38.000 Uttam Kumaran: Yeah, I would just… I would rip that first.
247 00:27:39.360 ⇒ 00:27:41.850 Uttam Kumaran: And then at least you’re like, okay, I have something.
248 00:27:42.600 ⇒ 00:27:45.769 Luke Scorziell: Yeah. And then we’ll take it, because always is, like, the engineering piece.
249 00:27:45.770 ⇒ 00:27:48.050 Uttam Kumaran: Anytime you ask somebody for an engineer, just…
250 00:27:48.310 ⇒ 00:27:50.700 Uttam Kumaran: expect that it’s gonna take 2X longer.
251 00:27:50.940 ⇒ 00:27:57.890 Uttam Kumaran: And so that way, for the sales team, you just have to give our team a little bit of a buffer to generate a demo.
252 00:27:59.090 ⇒ 00:28:07.409 Uttam Kumaran: But, like, if, again, if you come to the table with a lovable, like, that will speed up the process, or even, like, some type of mock.
253 00:28:07.720 ⇒ 00:28:14.079 Uttam Kumaran: Because then we don’t have, like, on our side, we then have to think about all the creative stuff like that, it’s just gonna become tough.
254 00:28:14.730 ⇒ 00:28:15.350 Luke Scorziell: Okay.
255 00:28:15.350 ⇒ 00:28:20.640 Uttam Kumaran: Just on, like, last resort, you could just display this, but, like, it seems like within… between now and then, you can get
256 00:28:21.090 ⇒ 00:28:25.030 Uttam Kumaran: A lovable thing, going, at least for… so that at least there’s something we have.
257 00:28:25.260 ⇒ 00:28:28.600 Luke Scorziell: Yeah, no, I could get a lovable, and I’m sure I could ask it to, like.
258 00:28:29.360 ⇒ 00:28:33.400 Luke Scorziell: Create fake data within it for, like.
259 00:28:33.510 ⇒ 00:28:40.939 Luke Scorziell: a knowledge base, custom… like, an agency. Like, I don’t… I… and that would be helpful, too, because it’s like, I don’t know exactly what they have sitting in.
260 00:28:41.260 ⇒ 00:28:44.070 Luke Scorziell: each of those SaaS things. So,
261 00:28:45.750 ⇒ 00:28:48.779 Luke Scorziell: Okay, and then, well, maybe just while we’re here then, like.
262 00:28:49.340 ⇒ 00:28:52.549 Luke Scorziell: What is an effective process, then, on some of these, like.
263 00:28:52.740 ⇒ 00:28:59.429 Luke Scorziell: sales calls, at least as you’ve experienced, because I found it helpful to have Pranav there, just to be able to talk about, like, the technical side of things that
264 00:28:59.910 ⇒ 00:29:03.870 Luke Scorziell: like, the… Engineering team can do.
265 00:29:03.970 ⇒ 00:29:12.070 Luke Scorziell: But then obviously, like, if I’m the one kind of driving the sale, so I don’t know, maybe if you have feedback for…
266 00:29:12.460 ⇒ 00:29:17.329 Luke Scorziell: Myself, Pranav, and I, just going forward on some of these calls, that would be helpful.
267 00:29:17.900 ⇒ 00:29:18.510 Uttam Kumaran: Cool.
268 00:29:19.100 ⇒ 00:29:20.420 Uttam Kumaran: Yeah, happy to.
269 00:29:24.850 ⇒ 00:29:27.810 Luke Scorziell: Sweetball.
270 00:29:27.810 ⇒ 00:29:33.020 Uttam Kumaran: Yeah, I’m gonna jump to a client call, but yeah, just let me know. I feel like… I feel like you guys have a good path forward.
271 00:29:35.650 ⇒ 00:29:38.710 Luke Scorziell: So yeah, okay, well, Pranava, I guess we can stay on.
272 00:29:39.620 ⇒ 00:29:40.930 Uttam Kumaran: Okay, I guess.
273 00:29:41.590 ⇒ 00:29:42.180 Pranav Narahari: Yep.
274 00:29:47.200 ⇒ 00:29:52.649 Pranav Narahari: Yeah, I think that’s a little bit on me, too, for sure, like…
275 00:29:54.620 ⇒ 00:29:57.659 Pranav Narahari: I think as I started, like, hopping into it.
276 00:29:58.200 ⇒ 00:30:04.140 Pranav Narahari: I realized, okay, like, with the Teams app integration,
277 00:30:04.210 ⇒ 00:30:23.669 Pranav Narahari: I just felt like I was doing too much, but then also, like, not hitting the bar exactly, like, what we wanted to show them. Like, we can’t show them something that’s not super polished, I guess, with, like, that thing that I showed. I mean, I wasn’t sure about, like, okay, is this okay to show? Is this not okay to show? But then Utam was like, yeah, this is not okay to show, so…
278 00:30:23.730 ⇒ 00:30:27.120 Luke Scorziell: Yeah, I mean, it’s our first time, and we’re kind of, like.
279 00:30:27.780 ⇒ 00:30:32.160 Luke Scorziell: So, I mean, I can just try to generate something unlovable right now.
280 00:30:32.160 ⇒ 00:30:32.970 Pranav Narahari: Okay.
281 00:30:32.970 ⇒ 00:30:35.899 Luke Scorziell: I guess, like, yeah, cause I… maybe…
282 00:30:36.830 ⇒ 00:30:43.350 Luke Scorziell: Maybe there’s, like, between us communicating, so it sounds like what you needed from me was more, like, specific…
283 00:30:43.630 ⇒ 00:30:52.750 Luke Scorziell: like, you kind of needed me to give you, like, the UI, I guess? Or, like, this is what it should look like, and what I kind of… the functionalities, like, the three functionalities I want, and then…
284 00:30:53.120 ⇒ 00:30:56.750 Luke Scorziell: And then you could kind of build a demo off of that, or is that…
285 00:30:57.480 ⇒ 00:31:01.070 Pranav Narahari: May, like, I mean… That wouldn’t…
286 00:31:01.070 ⇒ 00:31:02.929 Luke Scorziell: I thought we were kind of aligned
287 00:31:03.080 ⇒ 00:31:08.340 Luke Scorziell: Maybe, I don’t know if Utam maybe was… yeah, or just coming in without context, or… but I…
288 00:31:08.340 ⇒ 00:31:09.060 Pranav Narahari: Yeah.
289 00:31:09.060 ⇒ 00:31:11.860 Luke Scorziell: We were more aligned than maybe he was thinking.
290 00:31:11.860 ⇒ 00:31:18.780 Pranav Narahari: Yeah, I did too. I thought, like, this is more of, like, okay, we both think about, like, what that product looks like together. Yeah.
291 00:31:19.140 ⇒ 00:31:19.720 Luke Scorziell: What else?
292 00:31:19.720 ⇒ 00:31:22.739 Pranav Narahari: But… At the same time, too, like…
293 00:31:23.120 ⇒ 00:31:25.559 Pranav Narahari: If it’s more helpful for just, like.
294 00:31:25.860 ⇒ 00:31:32.269 Pranav Narahari: you to figure that out, and then maybe you ask me, like, questions that I can help with, then maybe that makes us…
295 00:31:32.730 ⇒ 00:31:34.950 Pranav Narahari: Because I think at this point, we just were…
296 00:31:35.170 ⇒ 00:31:40.029 Pranav Narahari: I don’t know if we were even confused, but it was just like… we didn’t…
297 00:31:41.000 ⇒ 00:31:47.010 Pranav Narahari: Again, we kind of did define, like, exactly what we wanted to show, too, but we didn’t talk about, like…
298 00:31:47.330 ⇒ 00:31:54.060 Pranav Narahari: maybe, like, the individual, like, screen states that, like, we want to display.
299 00:31:54.730 ⇒ 00:31:58.009 Luke Scorziell: Yeah, and also, like, what kind of data and stuff we could show, and that…
300 00:31:58.010 ⇒ 00:31:58.400 Pranav Narahari: Yeah.
301 00:31:58.400 ⇒ 00:32:00.860 Luke Scorziell: Specific problems that he has.
302 00:32:00.860 ⇒ 00:32:01.770 Pranav Narahari: Yeah.
303 00:32:02.190 ⇒ 00:32:12.489 Pranav Narahari: And so, yeah, I was also thinking, though, like, what Uten was just saying, like, does it actually…
304 00:32:14.660 ⇒ 00:32:18.679 Pranav Narahari: makes more sense to show a slide deck or a lovable?
305 00:32:19.170 ⇒ 00:32:26.969 Pranav Narahari: Or does it make more sense to just show, like, that chat GPT, like, adding a knowledge base, and then also, like, an MCP server?
306 00:32:27.390 ⇒ 00:32:33.119 Pranav Narahari: Maybe… I think probably both is what we want to do, but what are you thinking?
307 00:32:33.590 ⇒ 00:32:38.220 Luke Scorziell: I guess my thoughts with the ChatGBT thing is, like, do we show this to him, and then it’s just like a, like…
308 00:32:38.670 ⇒ 00:32:41.929 Luke Scorziell: He’s like, okay, I’ll just go figure out how to do this on my own.
309 00:32:42.160 ⇒ 00:32:46.009 Luke Scorziell: Like, how complicated was the ChatGPT setup that Tom showed?
310 00:32:47.010 ⇒ 00:32:50.360 Pranav Narahari: Oh, that’s, like, super simple, like, I could probably do it, like, in this call.
311 00:32:51.020 ⇒ 00:32:59.280 Pranav Narahari: But the thing is, is, like, for the MCP, we would likely have to create our own MCP, and that’s where, like, the work would…
312 00:32:59.880 ⇒ 00:33:01.119 Pranav Narahari: Would come about.
313 00:33:01.270 ⇒ 00:33:06.040 Pranav Narahari: And then also for the knowledge base, too,
314 00:33:06.470 ⇒ 00:33:09.079 Pranav Narahari: Depending on the type of data that they have.
315 00:33:09.690 ⇒ 00:33:29.050 Pranav Narahari: If it’s not just text documents, and there’s also certain limitations with ChatGPT. And then another thing with ChatGPT, too, is, like, that’s not integrated into their tools, right? Like, they’re using Copilot, they’re also using Teams. I’m thinking they want something integrated into, like, Microsoft Teams.
316 00:33:29.160 ⇒ 00:33:42.449 Pranav Narahari: ChatGPT… we can also talk about the limitations with what Utam just showed, like, the knowledge bases there have to be static documents. If you want to update the files, then you need to manually, like, pull in files into that custom GPT.
317 00:33:42.580 ⇒ 00:33:46.760 Pranav Narahari: So…
318 00:33:47.090 ⇒ 00:33:53.310 Pranav Narahari: what that shows is basically just like, oh yeah, this is what it will look like from a user’s perspective.
319 00:33:54.790 ⇒ 00:34:12.019 Pranav Narahari: However, we can talk about, like, it’s not as simple as just this. Like, it’s not like we’ve developed your app right here. We would still need to, set up all the backend logic to, like, dynamically pull data from all of your different sources into the knowledge base.
320 00:34:12.530 ⇒ 00:34:16.299 Pranav Narahari: And then also developing the MCP for each of the tools.
321 00:34:16.460 ⇒ 00:34:18.410 Pranav Narahari: Is something we would need to do as well.
322 00:34:19.340 ⇒ 00:34:20.010 Luke Scorziell: Okay.
323 00:34:25.639 ⇒ 00:34:32.510 Luke Scorziell: I mean, is it easy to show that same thing in Copilot that Utam just showed us, or is that…
324 00:34:32.969 ⇒ 00:34:33.869 Pranav Narahari: Let’s…
325 00:34:34.969 ⇒ 00:34:37.499 Luke Scorziell: Maybe we can just spend, like…
326 00:34:37.780 ⇒ 00:34:44.499 Luke Scorziell: I don’t know, 10, 15 minutes, like, you can kind of chop some stuff up, I can kind of work on a lovable thing, and then we can compare.
327 00:34:45.090 ⇒ 00:34:46.810 Pranav Narahari: Yeah, like right now, you’re thinking?
328 00:34:46.810 ⇒ 00:34:50.190 Luke Scorziell: Yeah, I don’t know, just to get somewhere instead of.
329 00:34:50.560 ⇒ 00:34:57.639 Pranav Narahari: Yeah, I agree. I think… Let’s do that. Let’s see…
330 00:35:02.670 ⇒ 00:35:12.489 Pranav Narahari: Yeah, Copilot looks very similar, so… Can probably… add empty somewhere.
331 00:35:41.090 ⇒ 00:35:42.479 Pranav Narahari: Share my screen.
332 00:35:45.330 ⇒ 00:35:49.060 Pranav Narahari: Yeah, so Copilot here… let’s see…
333 00:36:01.350 ⇒ 00:36:05.370 Pranav Narahari: Hi, I’m Hugada.
334 00:36:15.590 ⇒ 00:36:28.269 Pranav Narahari: It calculates the runtime for me.
335 00:36:28.860 ⇒ 00:36:35.960 Pranav Narahari: change from us, later.
336 00:36:51.160 ⇒ 00:36:55.160 Pranav Narahari: Yeah, that’s dope, that’s dope, but I think December 2 is, like, poor.
337 00:36:58.400 ⇒ 00:37:00.670 Pranav Narahari: Yeah, that’s what I…
338 00:37:18.480 ⇒ 00:37:34.569 Pranav Narahari: So it’s hard to think about that as well.
339 00:37:39.120 ⇒ 00:37:39.910 Pranav Narahari: Weird.
340 00:38:15.250 ⇒ 00:38:16.040 Luke Scorziell: Hmm.
341 00:38:36.200 ⇒ 00:38:37.189 Pranav Narahari: Keep walking.
342 00:39:39.060 ⇒ 00:39:54.060 Pranav Narahari: And maybe you could just referring right now.
343 00:40:39.260 ⇒ 00:40:40.450 Pranav Narahari: Nope, okay.
344 00:41:23.510 ⇒ 00:41:24.490 Pranav Narahari: Okay.
345 00:41:50.150 ⇒ 00:41:50.990 Pranav Narahari: I agree.
346 00:41:51.110 ⇒ 00:41:57.540 Pranav Narahari: Once you go into payout, you should have one of…
347 00:42:42.880 ⇒ 00:42:51.510 Pranav Narahari: Looks like there is a way to do this in Copilot. It’s, like, using, like, the Azure AI search plugin.
348 00:42:54.930 ⇒ 00:43:02.729 Pranav Narahari: Do you think I should go about doing it that way, or chat… like, is there a big difference between doing it within ChatGPT or doing it in Copilot?
349 00:43:04.750 ⇒ 00:43:13.500 Luke Scorziell: I mean, do you think… do you feel pretty confident that we can do the same thing in Copilot? How long do you… is that, like, a… how long would that… is that, like, a long process to do? What…
350 00:43:14.350 ⇒ 00:43:20.049 Pranav Narahari: Mmm… Connect Azure AI Search.
351 00:43:21.370 ⇒ 00:43:27.660 Pranav Narahari: Yeah, I’m just like… learning this for the first time, so… I don’t think it should be too long.
352 00:43:28.120 ⇒ 00:43:28.490 Luke Scorziell: I mean…
353 00:43:28.490 ⇒ 00:43:32.170 Pranav Narahari: I feel like, at this point, I could have already had it done in, like, ChatGPT.
354 00:43:33.000 ⇒ 00:43:33.740 Luke Scorziell: Yeah.
355 00:43:34.390 ⇒ 00:43:38.769 Pranav Narahari: But I, yeah, I see what… like, they are built using Microsoft.
356 00:43:39.060 ⇒ 00:43:45.619 Pranav Narahari: Right, so… It’d probably be best. From Utam’s perspective, do you think it’s just fine to be, like.
357 00:43:45.850 ⇒ 00:43:49.330 Luke Scorziell: Copilot? Or, I mean, ChatGPT, like, it’s not, like, a huge…
358 00:43:51.370 ⇒ 00:44:03.839 Luke Scorziell: I guess in my mind, I’m just thinking… I mean, I can build a lovable app. I’m working on a prompt right now that just creates, like, an app that looks like Microsoft Teams, so if you just want to show them in ChatGPT and then say, we can do the same thing in Copilot, and then I can…
359 00:44:04.090 ⇒ 00:44:10.630 Luke Scorziell: Kind of run through the, like, this lovable app.
360 00:44:12.900 ⇒ 00:44:18.060 Pranav Narahari: Yeah, I think… I don’t think it really makes a huge difference.
361 00:44:18.510 ⇒ 00:44:20.690 Luke Scorziell: Okay. If we put it in ChatGPT.
362 00:44:27.700 ⇒ 00:44:28.660 Pranav Narahari: Yeah.
363 00:44:35.240 ⇒ 00:44:37.950 Pranav Narahari: Let me just do it in ChatGPT first, and then…
364 00:44:37.950 ⇒ 00:44:38.590 Luke Scorziell: Oh my god.
365 00:44:39.160 ⇒ 00:44:42.339 Pranav Narahari: I feel like I could literally do it in, like, 2 seconds.
366 00:46:40.050 ⇒ 00:46:44.260 Luke Scorziell: This cursor has decided to build the app for me instead of giving me a lovable prompt.
367 00:46:45.390 ⇒ 00:46:46.340 Pranav Narahari: Oh, really?
368 00:46:47.740 ⇒ 00:46:48.899 Luke Scorziell: Yeah, it’s kinda weird.
369 00:54:05.420 ⇒ 00:54:07.699 Luke Scorziell: Yeah, I mean, they built this thing for me.
370 00:54:10.330 ⇒ 00:54:11.480 Pranav Narahari: It did a good job.
371 00:54:12.960 ⇒ 00:54:13.929 Luke Scorziell: Seems decent.
372 00:54:15.750 ⇒ 00:54:17.040 Luke Scorziell: Alright, look…
373 00:54:17.040 ⇒ 00:54:17.950 Pranav Narahari: Damn. Okay.
374 00:54:21.880 ⇒ 00:54:22.689 Luke Scorziell: I like.
375 00:54:35.580 ⇒ 00:54:37.170 Luke Scorziell: I don’t know, what do you think?
376 00:54:37.800 ⇒ 00:54:39.269 Pranav Narahari: I mean, it looks great.
377 00:55:07.910 ⇒ 00:55:14.219 Pranav Narahari: Yeah, and the idea here could be, like, we show him this, This kind of shows…
378 00:55:15.560 ⇒ 00:55:21.089 Pranav Narahari: most of it. If he’s like, okay, I understand all this, I just want to see the…
379 00:55:21.290 ⇒ 00:55:25.109 Pranav Narahari: the MCPs and the knowledge bases in action.
380 00:55:25.390 ⇒ 00:55:28.910 Pranav Narahari: that’s where we can probably show, like, ChatGPT.
381 00:55:30.570 ⇒ 00:55:33.300 Luke Scorziell: Okay, yeah, I like this to kind of show, like.
382 00:55:33.760 ⇒ 00:55:39.060 Luke Scorziell: It’d be nice if this could kind of be, like, at least a slide deck, so that I could, like… or not a slides, but…
383 00:55:41.230 ⇒ 00:55:44.290 Luke Scorziell: Can I ask things here? Tell me about…
384 00:56:14.110 ⇒ 00:56:15.170 Luke Scorziell: So, I mean
385 00:56:29.280 ⇒ 00:56:30.450 Luke Scorziell: I don’t know.
386 00:56:30.810 ⇒ 00:56:34.440 Luke Scorziell: Utah. I guess this looks good, pretty good.
387 00:56:37.830 ⇒ 00:56:38.740 Pranav Narahari: Yeah.
388 00:56:39.210 ⇒ 00:56:44.390 Luke Scorziell: Is this, like, realistic, I guess?
389 00:56:47.000 ⇒ 00:56:54.030 Pranav Narahari: Looks very realistic, like, yeah, this looks similar to, like, something we built in… Net for Lilo, so…
390 00:57:07.390 ⇒ 00:57:16.910 Pranav Narahari: And so… they’re… so, like, the Adobe CC, Frame.io, like, these are essentially doing tool calls.
391 00:57:17.450 ⇒ 00:57:22.289 Pranav Narahari: That’s the way we can describe it. Like, doing two calls to the MCP.
392 00:57:23.880 ⇒ 00:57:33.289 Pranav Narahari: And, you know, we’ll just reiterate, like, yeah, this is just for demo purposes, but this is, like, what you can expect as a user to, like, see.
393 00:57:33.910 ⇒ 00:57:42.730 Pranav Narahari: And then… Yeah, so…
394 00:57:43.000 ⇒ 00:57:44.200 Luke Scorziell: Search our archives.
395 00:57:44.200 ⇒ 00:57:44.690 Pranav Narahari: Figure.
396 00:57:46.340 ⇒ 00:57:48.220 Luke Scorziell: So this is the rag thing, I think.
397 00:57:50.060 ⇒ 00:57:50.960 Pranav Narahari: Gotcha.
398 00:57:55.150 ⇒ 00:58:00.309 Pranav Narahari: Okay, archive search is, like… Checking the knowledge base.
399 00:58:00.740 ⇒ 00:58:01.380 Luke Scorziell: Yeah.
400 00:58:03.690 ⇒ 00:58:04.420 Pranav Narahari: Okay.
401 00:58:08.350 ⇒ 00:58:09.929 Pranav Narahari: What do you think, like…
402 00:58:11.190 ⇒ 00:58:16.969 Pranav Narahari: is gonna be the agenda for tomorrow’s… tomorrow’s call, like, we’ll just basically jump right into this, and then…
403 00:58:17.190 ⇒ 00:58:18.570 Pranav Narahari: What’s after that?
404 00:58:21.130 ⇒ 00:58:25.200 Luke Scorziell: Yeah, great question. That’s kind of what I was thinking about as well.
405 00:58:25.420 ⇒ 00:58:26.590 Luke Scorziell: So…
406 00:58:33.970 ⇒ 00:58:34.990 Luke Scorziell: That’s cute.
407 00:58:49.320 ⇒ 00:58:55.030 Luke Scorziell: Yeah, I think, I mean, okay, so we’ll show him. I think we specifically talked…
408 00:58:56.710 ⇒ 00:59:01.370 Luke Scorziell: Okay, this is our first time, so it’s okay that we don’t get it perfect. But…
409 00:59:05.900 ⇒ 00:59:10.380 Luke Scorziell: I think the outcome that we want to aim towards tomorrow is…
410 00:59:12.620 ⇒ 00:59:17.140 Luke Scorziell: Like, probably getting on another call to show this to other internal stakeholders.
411 00:59:17.600 ⇒ 00:59:19.939 Pranav Narahari: Right? And asking if he’d be interested.
412 00:59:20.150 ⇒ 00:59:30.250 Luke Scorziell: So… I don’t know, I might have to spend a little more time, kind of.
413 00:59:30.480 ⇒ 00:59:37.200 Luke Scorziell: working through an agenda. I guess, are there any things that you would… I think there’s probably, like.
414 00:59:37.550 ⇒ 00:59:41.359 Luke Scorziell: I don’t know if this is too early still to have a pricing conversation,
415 00:59:43.190 ⇒ 00:59:48.330 Luke Scorziell: Or if it’s just, like, passing it along and kind of asking, like, okay, what’s the next thing that we can…
416 00:59:53.790 ⇒ 00:59:54.960 Luke Scorziell: Yeah, okay.
417 00:59:55.740 ⇒ 00:59:57.300 Pranav Narahari: Yeah, probably…
418 01:00:00.630 ⇒ 01:00:08.869 Pranav Narahari: maybe getting, also, a better understanding of, like, what they would want built. Like, we’re kind of just demoing, like, what is possible here.
419 01:00:09.230 ⇒ 01:00:16.559 Pranav Narahari: We haven’t really, like, They are using Copilot, are they looking for something to replace Copilot?
420 01:00:17.390 ⇒ 01:00:22.710 Pranav Narahari: Or just, like, something to fill in the gaps that Copilot isn’t hitting.
421 01:00:23.090 ⇒ 01:00:26.879 Pranav Narahari: We’ll also probably talk after this
422 01:00:27.270 ⇒ 01:00:31.339 Pranav Narahari: And so this is basically supposed to simulate, like, Microsoft Teams, right?
423 01:00:32.540 ⇒ 01:00:33.260 Luke Scorziell: Yeah.
424 01:00:34.130 ⇒ 01:00:37.840 Pranav Narahari: So we’ll talk about, like, how does that solution sound for y’all, like…
425 01:00:38.160 ⇒ 01:00:48.770 Pranav Narahari: Maybe we’ll get feedback there that is like, oh, we’d rather have, like, a standalone app versus Teams, or maybe it’s like, yeah, we would like to have Teams
426 01:00:54.260 ⇒ 01:01:00.190 Pranav Narahari: Yeah, I wonder… Not exactly sure, to be honest, like…
427 01:01:00.380 ⇒ 01:01:02.849 Pranav Narahari: how to structure tomorrow’s call, because, like.
428 01:01:03.190 ⇒ 01:01:07.359 Pranav Narahari: Yeah, we could probably maybe talk about timelines as well.
429 01:01:07.680 ⇒ 01:01:08.929 Pranav Narahari: Yeah. Something built.
430 01:01:10.850 ⇒ 01:01:27.000 Luke Scorziell: So… I’m lying away… I mean, and then, as we’re, like, budgeting and scoping for pricing, Like…
431 01:01:28.630 ⇒ 01:01:32.610 Luke Scorziell: I guess we need to ask, like, what is out of the range?
432 01:01:33.080 ⇒ 01:01:35.479 Luke Scorziell: What’s in the range?
433 01:01:36.400 ⇒ 01:01:39.720 Luke Scorziell: So, like, what would we do for 50K?
434 01:01:41.090 ⇒ 01:01:46.220 Luke Scorziell: Well, I mean, we don’t… I think if we proposed 50K, it’s gonna seem like he just gave us a number, and then we were like.
435 01:01:46.620 ⇒ 01:01:50.229 Luke Scorziell: I mean, maybe it’s fine to just run with that, but
436 01:01:50.590 ⇒ 01:01:54.430 Luke Scorziell: I’d say, what’s like a… what’s like a… Easy.
437 01:01:54.930 ⇒ 01:01:55.740 Luke Scorziell: like…
438 01:01:56.420 ⇒ 01:02:05.029 Luke Scorziell: An easy, like, low-budget range, an in-the-range, and then maybe a stretch of, like, what are the features that we would include, not include.
439 01:02:06.600 ⇒ 01:02:11.939 Luke Scorziell: in each of those would kind of be what I would think. So, but I don’t know, maybe this might be too early to…
440 01:02:14.300 ⇒ 01:02:19.760 Luke Scorziell: To start having that conversation, because it sounds like you’re still saying we need to figure out what they even want built.
441 01:02:20.330 ⇒ 01:02:23.039 Luke Scorziell: Before we come up with a proposal.
442 01:02:24.220 ⇒ 01:02:31.410 Pranav Narahari: Yeah, we could probably, like… High level, talk about… Roadmap?
443 01:02:32.370 ⇒ 01:02:37.859 Pranav Narahari: Roadmap… well, maybe roadmap’s the wrong way to think about it. Maybe just, just timeline.
444 01:02:40.600 ⇒ 01:02:44.060 Pranav Narahari: Because how it… and it would also work, like, on a…
445 01:02:46.370 ⇒ 01:02:49.370 Pranav Narahari: I guess I don’t even know if that’s even a relevant question.
446 01:02:49.960 ⇒ 01:02:51.049 Pranav Narahari: Trying to think…
447 01:02:52.570 ⇒ 01:02:55.520 Luke Scorziell: Yeah. You have… how does the rest of your day look?
448 01:02:56.040 ⇒ 01:03:01.439 Pranav Narahari: I actually do kind of have a little bit of a busy day, because I’m, like, onboarding onto ABC.
449 01:03:01.670 ⇒ 01:03:02.700 Luke Scorziell: Okay.
450 01:03:03.280 ⇒ 01:03:06.630 Pranav Narahari: So… But…
451 01:03:07.560 ⇒ 01:03:17.629 Pranav Narahari: I have a 4PM that I could move, and then I just have my CSO call, and then I’m chilling outside of that, so we could…
452 01:03:18.080 ⇒ 01:03:20.769 Pranav Narahari: Yeah, do you wanna, like, touch base a little bit later today again?
453 01:03:21.080 ⇒ 01:03:25.340 Luke Scorziell: Yeah, maybe if there’s, like, an end of day…
454 01:03:25.460 ⇒ 01:03:31.080 Luke Scorziell: I just have a meeting that I need to prep for, it’s probably gonna take me about an hour to prep.
455 01:03:32.040 ⇒ 01:03:32.640 Pranav Narahari: Okay.
456 01:03:34.250 ⇒ 01:03:44.319 Luke Scorziell: And then… I can work on a script and an agenda, and then maybe I can just Slack it to you, and then if you want to give me, kind of, we can go back and forth on Slack, maybe?
457 01:03:44.590 ⇒ 01:03:45.269 Pranav Narahari: Yeah, that’s perfect.
458 01:03:45.270 ⇒ 01:03:48.230 Luke Scorziell: If it makes sense to meet later today, we can huddle or something.
459 01:03:48.720 ⇒ 01:03:49.900 Pranav Narahari: Perfect, yeah.
460 01:03:50.230 ⇒ 01:03:50.830 Luke Scorziell: Okay.
461 01:03:50.960 ⇒ 01:03:56.950 Luke Scorziell: Cool. Alright, well, we’ll iron this out, and then, yeah, getting this one down will make other ones easier, too, so…
462 01:03:57.160 ⇒ 01:03:58.360 Pranav Narahari: Yeah, totally.
463 01:03:58.600 ⇒ 01:03:59.689 Pranav Narahari: Sounds good, Luke.
464 01:03:59.870 ⇒ 01:04:01.470 Luke Scorziell: Cool, alright, thanks, Pranav.
465 01:04:01.470 ⇒ 01:04:02.839 Pranav Narahari: Yeah, thank you. See ya.
466 01:04:03.240 ⇒ 01:04:03.710 Luke Scorziell: But…