Meeting Title: Lilo Deliverables Estimates Sync Date: 2026-02-24 Meeting participants: Casie Aviles, Pranav Narahari, Samuel Roberts
WEBVTT
1 00:03:01.000 ⇒ 00:03:01.990 Pranav Narahari: Hey, Casey.
2 00:03:15.240 ⇒ 00:03:16.699 Casie Aviles: Waiting for some…
3 00:03:23.780 ⇒ 00:03:24.940 Pranav Narahari: Hey, can you hear me?
4 00:03:25.870 ⇒ 00:03:27.340 Casie Aviles: Hello? Hello?
5 00:03:27.340 ⇒ 00:03:29.469 Pranav Narahari: Yeah, I can hear you. No.
6 00:03:30.690 ⇒ 00:03:31.330 Casie Aviles: Okay.
7 00:03:34.130 ⇒ 00:03:37.130 Pranav Narahari: So I’m actually, like, on the road right now.
8 00:03:37.130 ⇒ 00:03:37.590 Casie Aviles: Whoa.
9 00:03:38.610 ⇒ 00:03:43.319 Pranav Narahari: Yeah, but, this still works, like, I knew I’d be on the road for this call.
10 00:03:44.040 ⇒ 00:03:58.220 Pranav Narahari: if you don’t mind driving this a little bit, well, I can, like… if you can just, like, share your screen with, like, that Excel, Google Sheet, that would be super helpful for me, just so, like, Sam has, like, additional context, and also you.
11 00:03:58.510 ⇒ 00:04:05.510 Pranav Narahari: It’s the one that I think I sent yesterday, and you were looking over the deliverables to make sure there was no, like, duplicate deliverables.
12 00:04:05.510 ⇒ 00:04:08.800 Casie Aviles: Oh, okay, okay. Yeah. Big time.
13 00:04:08.800 ⇒ 00:04:11.460 Pranav Narahari: And… Perfect, yeah.
14 00:04:13.660 ⇒ 00:04:16.540 Pranav Narahari: So… yeah, actually, I don’t know, is Sam in here?
15 00:04:16.610 ⇒ 00:04:17.320 Casie Aviles: Yeah, I just…
16 00:04:17.329 ⇒ 00:04:18.159 Samuel Roberts: I just joined in.
17 00:04:18.950 ⇒ 00:04:19.920 Pranav Narahari: Okay, perfect.
18 00:04:19.920 ⇒ 00:04:21.949 Samuel Roberts: I just didn’t want to interrupt while you were talking.
19 00:04:21.959 ⇒ 00:04:24.189 Pranav Narahari: Oh, yeah, no, no, no.
20 00:04:24.489 ⇒ 00:04:34.959 Pranav Narahari: Yeah, Casey’s just gonna share his screen, show the deliverables from that Google Sheet that I pasted yesterday, and I’ll talk a little bit about, like, what the priorities are,
21 00:04:35.189 ⇒ 00:04:42.379 Pranav Narahari: Basically, what the priorities are now, what we’re going to stick to, is just what was defined in the original SOW.
22 00:04:42.719 ⇒ 00:04:50.629 Pranav Narahari: And that’s gonna be the 14 deliverables, some of which we’ve already completed, but they’re at the top of this document.
23 00:04:50.800 ⇒ 00:04:51.210 Samuel Roberts: Cool.
24 00:04:51.210 ⇒ 00:04:53.400 Pranav Narahari: So… Yeah.
25 00:04:54.410 ⇒ 00:04:56.519 Casie Aviles: Okay, perfect. Can you guys see my screen now?
26 00:04:56.580 ⇒ 00:04:57.510 Samuel Roberts: Yeah, definitely.
27 00:04:57.510 ⇒ 00:04:58.180 Pranav Narahari: That’s it.
28 00:04:58.870 ⇒ 00:05:02.530 Pranav Narahari: Yeah, so…
29 00:05:03.380 ⇒ 00:05:14.220 Pranav Narahari: I’ll let y’all just look over these 14 real quick. If you… Casey, if you even want to add an additional filter to just, like, filter out the ones we’ve already completed, that works too, because we only really need to scope out
30 00:05:14.380 ⇒ 00:05:17.010 Pranav Narahari: And estimate the ones that we haven’t.
31 00:05:20.650 ⇒ 00:05:21.710 Casie Aviles: Okay.
32 00:05:21.710 ⇒ 00:05:25.500 Pranav Narahari: But, yeah, we can just scroll through all 14 of them, too, it’s not like… it’s not a big deal.
33 00:05:27.460 ⇒ 00:05:29.490 Pranav Narahari: Oh, I don’t think I can…
34 00:05:29.870 ⇒ 00:05:33.729 Pranav Narahari: Oh, do you not have edit access on it? Okay. Okay. Yeah.
35 00:05:33.930 ⇒ 00:05:37.879 Pranav Narahari: That’s totally fine. What we can do is, so… let’s just go line by line.
36 00:05:38.020 ⇒ 00:05:44.400 Pranav Narahari: figure out, like, what we have the status set at right now. I… some of this might not be accurate.
37 00:05:45.070 ⇒ 00:05:50.229 Samuel Roberts: I don’t even know if the client looked at this, to be honest, so… Okay.
38 00:05:50.230 ⇒ 00:05:54.690 Pranav Narahari: Another story, but yeah, basically,
39 00:05:56.720 ⇒ 00:06:02.540 Pranav Narahari: How many… so the data warehouse one, that’s the first one, that’s what I remember. Sorry, yeah, Sam, I’m also driving right now, so…
40 00:06:02.540 ⇒ 00:06:04.630 Samuel Roberts: Yeah, I was gonna say, where… where are you right now?
41 00:06:05.400 ⇒ 00:06:13.889 Pranav Narahari: Oh, I probably actually crossed by you recently. I was driving from Champaign, Illinois to Massachusetts.
42 00:06:14.450 ⇒ 00:06:15.810 Samuel Roberts: Oh, yeah, down 90?
43 00:06:16.550 ⇒ 00:06:19.349 Pranav Narahari: Yeah, I think I was taking 70 for most of it.
44 00:06:20.070 ⇒ 00:06:21.709 Samuel Roberts: Oh, 70, okay, okay.
45 00:06:21.870 ⇒ 00:06:23.829 Pranav Narahari: Yeah. Nice.
46 00:06:24.070 ⇒ 00:06:27.820 Pranav Narahari: But yeah, it’s just because of the storm that’s, like, in New England, like…
47 00:06:28.160 ⇒ 00:06:30.440 Pranav Narahari: You’re heading right towards it, yeah.
48 00:06:30.440 ⇒ 00:06:34.209 Samuel Roberts: Yeah, so right now, it’s calm, and so that’s why I’m just trying to get.
49 00:06:34.210 ⇒ 00:06:36.970 Pranav Narahari: Yeah. Before it gets, like, it picks up again.
50 00:06:36.970 ⇒ 00:06:39.490 Samuel Roberts: Crazy. Alright, well, be safe.
51 00:06:39.490 ⇒ 00:06:42.969 Pranav Narahari: I appreciate it, I appreciate it. Right now, it’s pretty clear skies, so, I’m doing okay.
52 00:06:42.970 ⇒ 00:06:43.590 Samuel Roberts: Good.
53 00:06:43.830 ⇒ 00:06:44.480 Samuel Roberts: Alrighty.
54 00:06:44.840 ⇒ 00:06:48.309 Pranav Narahari: But, yeah, there’s a few here that are already completed.
55 00:06:48.690 ⇒ 00:07:03.680 Pranav Narahari: The goal of this is just to make sure, like, the estimates that I already wrote here, like, look pretty accurate to y’all. You won’t see hours estimates in here, you’ll just see the total cost, but just take this number, like, the total cost, and divide it by…
56 00:07:04.950 ⇒ 00:07:15.040 Pranav Narahari: 200-ish, like, 225, I guess we can say. And then that’s, like, the hours that I said that it would take to accomplish this.
57 00:07:16.070 ⇒ 00:07:19.149 Pranav Narahari: And, yeah, we’re gonna start from there, so…
58 00:07:19.730 ⇒ 00:07:29.040 Pranav Narahari: Okay. Casey, if you want to, like, scroll over to the total cost estimate, I think it’s, like, InScope cost is the column name.
59 00:07:29.180 ⇒ 00:07:32.590 Pranav Narahari: You can just… let’s take a look at that. If you can…
60 00:07:33.370 ⇒ 00:07:42.930 Pranav Narahari: do the quick calculation. I can’t see it right now, if you can do the quick calculation of how many hours it could be, and just say it out loud, we can just figure out, like, if that sounds about right.
61 00:07:43.070 ⇒ 00:07:44.230 Pranav Narahari: And then go from there.
62 00:07:45.290 ⇒ 00:07:52.800 Samuel Roberts: Okay, yeah, so let’s see, the 10,000 is… 250? Okay.
63 00:07:53.210 ⇒ 00:07:56.840 Casie Aviles: We divide it by… 200, right?
64 00:07:57.850 ⇒ 00:08:06.100 Pranav Narahari: Yeah, you can divide by 200, yeah, so… that’s 50 hours. I think… That makes sense.
65 00:08:06.710 ⇒ 00:08:09.090 Pranav Narahari: There’s a few other tasks in here where we, like.
66 00:08:09.810 ⇒ 00:08:26.709 Pranav Narahari: where we’re talking about just, like, the integration of data within Polytomic or Airbyte, which is not included within the data warehouse portion of things, this data warehouse line item, how I think of it, is that… Sam, we kind of discussed this before, like, how do we want to structure the…
67 00:08:27.530 ⇒ 00:08:29.640 Pranav Narahari: All of these brands, and all of these…
68 00:08:29.640 ⇒ 00:08:30.260 Samuel Roberts: Yeah.
69 00:08:30.260 ⇒ 00:08:35.580 Pranav Narahari: for integrations within, Mother Duck? Like, what does that architecture look like?
70 00:08:35.809 ⇒ 00:08:41.240 Pranav Narahari: That’s… what I think is defined within this line item.
71 00:08:42.270 ⇒ 00:08:42.840 Samuel Roberts: Okay.
72 00:08:43.870 ⇒ 00:08:46.630 Pranav Narahari: Yeah, and does this sound, like, about right?
73 00:08:47.190 ⇒ 00:08:49.950 Samuel Roberts: 50-hour, connections, so we’re getting,
74 00:08:51.830 ⇒ 00:08:56.340 Samuel Roberts: AirByte Meta, Google, Automate Pipeline Per Brand… yeah…
75 00:08:57.010 ⇒ 00:08:58.669 Samuel Roberts: I don’t… yeah, I think so.
76 00:08:58.930 ⇒ 00:09:00.669 Samuel Roberts: Okay. I think you’re on track there.
77 00:09:00.970 ⇒ 00:09:06.960 Samuel Roberts: I think there’s a little bit of uncertainty there, but yeah, I don’t think that’s crazy. 50 hours… yeah, that’s probably pretty good.
78 00:09:07.570 ⇒ 00:09:13.770 Pranav Narahari: Okay, cool. Yeah, Casey, can you read out the next one that is incomplete?
79 00:09:14.380 ⇒ 00:09:16.739 Samuel Roberts: What is this? This is the MCP…
80 00:09:16.740 ⇒ 00:09:18.100 Casie Aviles: Progress…
81 00:09:18.100 ⇒ 00:09:19.020 Samuel Roberts: metadashboard data.
82 00:09:19.020 ⇒ 00:09:20.679 Casie Aviles: Meta Dashboard Data Revenue.
83 00:09:21.990 ⇒ 00:09:22.660 Pranav Narahari: Okay.
84 00:09:23.010 ⇒ 00:09:24.050 Casie Aviles: And the school…
85 00:09:24.050 ⇒ 00:09:24.770 Samuel Roberts: hours.
86 00:09:25.400 ⇒ 00:09:26.090 Casie Aviles: Yep.
87 00:09:26.520 ⇒ 00:09:27.520 Casie Aviles: 1200.
88 00:09:27.520 ⇒ 00:09:29.090 Samuel Roberts: I’ll let you talk, Casey. Go ahead.
89 00:09:29.090 ⇒ 00:09:32.449 Casie Aviles: Yeah, the scope cost here is 1200, so…
90 00:09:33.040 ⇒ 00:09:35.439 Casie Aviles: the RS estimate would be 6.
91 00:09:37.200 ⇒ 00:09:42.580 Samuel Roberts: This is, like… remaining, right? Not… Totally.
92 00:09:42.580 ⇒ 00:09:46.780 Pranav Narahari: Yeah, it’s what we have remaining. I think there’s already a lot of progress done here.
93 00:09:46.780 ⇒ 00:09:48.380 Samuel Roberts: But yeah, I think that’s probably fine. I mean…
94 00:09:48.650 ⇒ 00:09:51.939 Pranav Narahari: Yeah, the thing about this… this feature is…
95 00:09:52.260 ⇒ 00:10:01.009 Pranav Narahari: it’s just pulling the data from MCP, and what we would like is that it’s pulling it from the data warehouse, but even then, like, you know.
96 00:10:01.370 ⇒ 00:10:04.609 Pranav Narahari: We’re kind of in, like, a weird position right now, where we’re like…
97 00:10:05.220 ⇒ 00:10:18.330 Pranav Narahari: we just want to get things shipped, like, they’re actually happy with the data right now, so even if it is connected to the MCP and they’re happy with it, if it’s just, like, some UI tweaks, like, that’s fine, let’s just get that done within the 6 hours. If we have the capacity to then…
98 00:10:18.520 ⇒ 00:10:23.169 Pranav Narahari: instead of pulling from MCP, we can pull from the data warehouse, even better. Yeah.
99 00:10:23.170 ⇒ 00:10:23.780 Samuel Roberts: I think, yeah, I think we
100 00:10:24.110 ⇒ 00:10:25.610 Samuel Roberts: Eventually, but I think you’re right.
101 00:10:25.610 ⇒ 00:10:26.260 Pranav Narahari: Yeah.
102 00:10:26.890 ⇒ 00:10:27.680 Pranav Narahari: Cool.
103 00:10:28.150 ⇒ 00:10:28.680 Casie Aviles: Okay.
104 00:10:28.680 ⇒ 00:10:40.470 Pranav Narahari: And then, yeah, let me, like, let me just, like, mention, too, I think one thing I want to talk about here as well is just, like, the sequence of events. To be honest, like, we’ve already sent over this doc, so, like…
105 00:10:40.750 ⇒ 00:10:49.299 Pranav Narahari: some of the estimate stuff is, like, already, like, predefined. I feel like I did a pretty good job estimating, though, and, like, made it pretty conservative.
106 00:10:49.300 ⇒ 00:10:49.860 Samuel Roberts: Yeah.
107 00:10:49.860 ⇒ 00:10:52.190 Pranav Narahari: I think the sequencing thing is, like.
108 00:10:52.460 ⇒ 00:10:57.239 Pranav Narahari: the main thing that I want to, like, talk to you both about,
109 00:10:57.610 ⇒ 00:11:00.830 Pranav Narahari: Two things that we need to consider here is, like.
110 00:11:01.220 ⇒ 00:11:12.350 Pranav Narahari: The data warehouse is likely the first thing that we want to start with, just because that is what’s going to pull in the data for the forecasting dashboard, as well as the remaining Slack reports.
111 00:11:13.340 ⇒ 00:11:21.929 Pranav Narahari: The MCP stuff is kind of just, like, you know, completely separate, don’t really need to, like… doesn’t have really any dependency with any of the data warehouse or forecast.
112 00:11:21.930 ⇒ 00:11:22.370 Samuel Roberts: Yeah.
113 00:11:22.370 ⇒ 00:11:27.310 Pranav Narahari: Blackboard stuff. Yeah, so…
114 00:11:28.270 ⇒ 00:11:43.589 Pranav Narahari: And, yeah, the second thing is that, I don’t know if I just said this, my bad. We’re at… we should be, like, trying to be at 80 hours per month. Like, that’s what is, like, the healthy amount for this client, based on just, like, what our…
115 00:11:43.590 ⇒ 00:11:55.250 Pranav Narahari: deal is. We’re not at, like, we’re not on an hourly plan with them, we’re on a monthly, like, fixed cost with them, so we internally have, like, decided, like, 80 hours per week is where we want to be.
116 00:11:55.880 ⇒ 00:11:58.049 Pranav Narahari: Oh, sorry, not per week, per month.
117 00:11:58.620 ⇒ 00:12:00.410 Samuel Roberts: Yeah, yeah, right. Okay.
118 00:12:02.290 ⇒ 00:12:02.860 Pranav Narahari: Cool.
119 00:12:02.980 ⇒ 00:12:04.570 Pranav Narahari: Yeah, good to know.
120 00:12:05.200 ⇒ 00:12:08.639 Pranav Narahari: Yeah, yeah, I think it’s just good for everybody to, like, know that,
121 00:12:09.210 ⇒ 00:12:13.010 Pranav Narahari: I think what that’ll probably mean is, like.
122 00:12:13.190 ⇒ 00:12:22.250 Pranav Narahari: Yes, like, Casey, you and I will probably… and probably mostly me, will be, like, taking on the bulk of this stuff, and then…
123 00:12:22.610 ⇒ 00:12:37.319 Pranav Narahari: Sam, like, you’ll be there for, like, the… when I have, like, real, like, architectural decisions, like, maybe, like, cross… crossroads, that’s where I’ll probably be, like, Casey and I will be coming to you for, like, assistance.
124 00:12:38.190 ⇒ 00:12:39.070 Samuel Roberts: Sounds good.
125 00:12:39.070 ⇒ 00:12:40.240 Pranav Narahari: Yeah.
126 00:12:40.350 ⇒ 00:12:46.429 Pranav Narahari: Okay, perfect. So then, yeah, we could just, like, continue going down this list.
127 00:12:46.780 ⇒ 00:12:48.619 Pranav Narahari: Also, I want to know, like.
128 00:12:48.940 ⇒ 00:13:02.360 Pranav Narahari: I may have just missed, like, what the status on some of these things are, like, some that may say incomplete that might actually already be completed, so that would be good to know. Yeah, if you just want to, like, maybe peek at each line item, and then…
129 00:13:03.350 ⇒ 00:13:08.059 Pranav Narahari: I think, Casey, there’s a few that maybe I think… I think you would have the most,
130 00:13:08.210 ⇒ 00:13:09.670 Pranav Narahari: Up-to-date knowledge on.
131 00:13:12.050 ⇒ 00:13:16.399 Pranav Narahari: Yeah. Okay. Okay, so, yeah, the next item that we have is…
132 00:13:16.400 ⇒ 00:13:18.959 Casie Aviles: for automated Slack reporting.
133 00:13:19.850 ⇒ 00:13:20.840 Casie Aviles: Yeah.
134 00:13:21.290 ⇒ 00:13:22.489 Pranav Narahari: That one, bro.
135 00:13:22.490 ⇒ 00:13:23.180 Casie Aviles: Chris.
136 00:13:23.570 ⇒ 00:13:26.330 Pranav Narahari: Yeah, that’s in progress. There’s, like, the…
137 00:13:27.790 ⇒ 00:13:34.239 Pranav Narahari: there’s a few reports in there that I think are gonna be super easily generated based on
138 00:13:34.560 ⇒ 00:13:40.060 Pranav Narahari: just after we create the dashboard… oh, sorry, not the dashboard, the data warehouse. So… Yeah.
139 00:13:40.100 ⇒ 00:13:57.409 Pranav Narahari: it’s really just gonna be, like, a query to the data warehouse, and then that’s what’s going to populate the Slack report. We already have, like, the end-to-end process of creating one Slack report. Now we just need to be, instead of pulling the data from the MCPU, which I think is a little bit more difficult, just pulling it from the data warehouse.
140 00:14:00.990 ⇒ 00:14:03.379 Samuel Roberts: Yeah, okay, so 14 hours,
141 00:14:04.430 ⇒ 00:14:06.720 Samuel Roberts: Do we schedule a 5 minute, please?
142 00:14:08.230 ⇒ 00:14:13.450 Samuel Roberts: Do we have a list of, like, what those… ports are…
143 00:14:14.160 ⇒ 00:14:16.140 Pranav Narahari: Yeah, I have a…
144 00:14:16.310 ⇒ 00:14:25.669 Pranav Narahari: they’re not linked here, they’re probably somewhere in either linear… yeah, they should be in linear, probably in, like, the done column now.
145 00:14:26.100 ⇒ 00:14:34.970 Pranav Narahari: But, yeah, they’re pretty… pretty easy. Like, the main… the difficult one was Shopify,
146 00:14:35.420 ⇒ 00:14:36.080 Samuel Roberts: Okay.
147 00:14:36.300 ⇒ 00:14:39.300 Pranav Narahari: Meta will be pretty straightforward.
148 00:14:39.300 ⇒ 00:14:40.230 Samuel Roberts: Yeah.
149 00:14:40.230 ⇒ 00:14:41.440 Pranav Narahari: than so will Google.
150 00:14:42.530 ⇒ 00:14:44.640 Samuel Roberts: Okay, cool, then yeah, that seems reasonable.
151 00:14:46.910 ⇒ 00:14:47.870 Samuel Roberts: And…
152 00:14:47.870 ⇒ 00:14:48.350 Casie Aviles: Shit.
153 00:14:48.370 ⇒ 00:14:50.060 Samuel Roberts: Prototype…
154 00:14:50.310 ⇒ 00:14:54.599 Casie Aviles: Yeah, the… Yep. Where’s this?
155 00:14:56.330 ⇒ 00:15:03.520 Samuel Roberts: Oh, this is Nano Banana. Okay, yeah, so I saw the ad machine and forgot what that was. Yeah, I see Nano Banana over there now. Okay, yeah, is that… that says not started, but that should be…
156 00:15:03.680 ⇒ 00:15:07.169 Casie Aviles: Yeah, I think that should be completed, I think.
157 00:15:07.320 ⇒ 00:15:08.630 Pranav Narahari: Okay, perfect.
158 00:15:09.030 ⇒ 00:15:11.060 Samuel Roberts: Yeah, it’s completely Yeah.
159 00:15:11.060 ⇒ 00:15:12.899 Pranav Narahari: I don’t know, do you have a…
160 00:15:13.570 ⇒ 00:15:23.769 Pranav Narahari: If you can’t edit it directly, Casey, can you just, like, write in our, like, internal, like, channel, or you can even message me individually, and I can just, like, update that? Yeah.
161 00:15:23.770 ⇒ 00:15:26.579 Casie Aviles: I have a copy right now, I just made a copy.
162 00:15:26.990 ⇒ 00:15:27.359 Casie Aviles: Oh, bro.
163 00:15:27.360 ⇒ 00:15:30.160 Pranav Narahari: Perfect, yeah, you can send me the copy. Yeah.
164 00:15:31.050 ⇒ 00:15:32.309 Samuel Roberts: Cool. That’s perfect.
165 00:15:34.270 ⇒ 00:15:35.320 Samuel Roberts: Okay, awesome.
166 00:15:35.320 ⇒ 00:15:37.080 Casie Aviles: That seems good.
167 00:15:37.780 ⇒ 00:15:40.660 Casie Aviles: chat data… fix…
168 00:15:44.920 ⇒ 00:15:50.390 Samuel Roberts: This is… where I get a little nervous, because… To stop pure, like, just…
169 00:15:50.930 ⇒ 00:15:57.769 Samuel Roberts: Guaranteed no hallucinations is not something that we can do with AI, but I see, I know what they’re trying to get here. Okay.
170 00:15:58.710 ⇒ 00:16:01.009 Pranav Narahari: Oh, so, yeah, what does it say again? Can you read it?
171 00:16:01.010 ⇒ 00:16:07.010 Samuel Roberts: This is, address issue where AI chat returns fabricated slash hallucinated data instead of the actual data from connected sources.
172 00:16:08.450 ⇒ 00:16:14.749 Samuel Roberts: Gotcha. Chat responsibly, only verified data. Yeah, I’m trying to… I mean, this is, like, an industry problem, not, you know…
173 00:16:15.230 ⇒ 00:16:19.139 Pranav Narahari: Right, yeah, let’s not worry too much about this, like, how much…
174 00:16:19.140 ⇒ 00:16:23.810 Samuel Roberts: Yeah, I’m not… I think… the 16 hours… I think that’s…
175 00:16:23.810 ⇒ 00:16:24.130 Pranav Narahari: Okay.
176 00:16:24.350 ⇒ 00:16:29.169 Samuel Roberts: reasonable to get it better. I think part of this is also we should probably think about ways to, like.
177 00:16:29.350 ⇒ 00:16:32.950 Samuel Roberts: Show the data being processed, maybe?
178 00:16:35.310 ⇒ 00:16:39.699 Samuel Roberts: Or have a validation check or something? Yeah, I think that’s fine.
179 00:16:40.040 ⇒ 00:16:40.639 Samuel Roberts: Go ahead.
180 00:16:40.640 ⇒ 00:16:42.080 Pranav Narahari: Yeah, we can’t…
181 00:16:42.290 ⇒ 00:16:54.049 Pranav Narahari: make a non-deterministic thing, like, always right, right? And it’s not gonna be exactly what the client wants. So, like, the idea here is just, like, how do we make it a little bit better? It’s already…
182 00:16:54.050 ⇒ 00:16:57.400 Samuel Roberts: Fine, yeah, and that’s where I think 100%, yeah, you’re fine there.
183 00:16:58.060 ⇒ 00:16:58.830 Pranav Narahari: gold.
184 00:17:00.810 ⇒ 00:17:05.560 Samuel Roberts: Alright. Yeah, I think just some more… yeah. Alright, cool. UI enablement…
185 00:17:05.819 ⇒ 00:17:08.930 Casie Aviles: Google Ads… MCP, so…
186 00:17:08.930 ⇒ 00:17:09.890 Samuel Roberts: That shouldn’t be much.
187 00:17:09.890 ⇒ 00:17:10.280 Pranav Narahari: Yeah.
188 00:17:10.280 ⇒ 00:17:12.910 Samuel Roberts: I thought that we added a PR and dev for that, right?
189 00:17:12.910 ⇒ 00:17:18.189 Pranav Narahari: It’s already in dev, yeah, I think if we can just do some testing on that, if we haven’t already.
190 00:17:18.190 ⇒ 00:17:18.810 Samuel Roberts: Yeah.
191 00:17:18.819 ⇒ 00:17:23.769 Pranav Narahari: I mean, just pushing that to staging, production, like, if there’s any issues for that, yeah.
192 00:17:23.770 ⇒ 00:17:25.319 Samuel Roberts: Yeah, this should be fine, I think…
193 00:17:26.060 ⇒ 00:17:26.510 Pranav Narahari: Yeah.
194 00:17:26.510 ⇒ 00:17:27.140 Samuel Roberts: Go ahead.
195 00:17:27.560 ⇒ 00:17:30.929 Pranav Narahari: I think the estimate’s probably pretty low on that, like, we’re probably gonna…
196 00:17:30.930 ⇒ 00:17:43.099 Samuel Roberts: It’s just 8, it’s 1600, yeah, it’s 8. I think that’s fine. I think, really, the issue is that they have to connect Google so we can actually really test it, otherwise it’s hard to test without ad accounts, or we have to set up a bunch of fake ad accounts, but yeah, I think either way, that’s fine. Okay.
197 00:17:43.680 ⇒ 00:17:50.220 Samuel Roberts: The next one is, forecasting MVP…
198 00:17:50.490 ⇒ 00:17:57.330 Samuel Roberts: meta ads, data connection, connect Facebook Meta Ads data source to Polytomic using results… is this…
199 00:17:58.300 ⇒ 00:18:00.459 Samuel Roberts: Are we sticking with polytomic for that one? Is that…
200 00:18:01.000 ⇒ 00:18:01.630 Samuel Roberts: Or is that…
201 00:18:01.630 ⇒ 00:18:05.229 Pranav Narahari: Yeah, they were able to make the connection with Polyatomic,
202 00:18:05.790 ⇒ 00:18:11.450 Pranav Narahari: To be honest, at this point, it’s just like, wherever we can make the connection, that’s how I’m thinking about it.
203 00:18:11.450 ⇒ 00:18:12.240 Samuel Roberts: Okay. Yeah.
204 00:18:12.240 ⇒ 00:18:17.019 Pranav Narahari: comic, of course, is, like, where we are most confident, but also…
205 00:18:17.320 ⇒ 00:18:22.549 Pranav Narahari: airplane’s probably fine, too, I just won’t… wherever they can get it to work, like, this is kind of on them, like…
206 00:18:22.880 ⇒ 00:18:25.729 Pranav Narahari: Can they authenticate their Meta account?
207 00:18:25.870 ⇒ 00:18:30.999 Pranav Narahari: with either of them. They said they were able to do it for one brand on Polysomic.
208 00:18:31.200 ⇒ 00:18:31.760 Samuel Roberts: Oh, God.
209 00:18:31.760 ⇒ 00:18:32.680 Pranav Narahari: Oh.
210 00:18:32.680 ⇒ 00:18:34.480 Samuel Roberts: That’s… yeah.
211 00:18:34.480 ⇒ 00:18:35.150 Pranav Narahari: Okay.
212 00:18:35.960 ⇒ 00:18:41.549 Pranav Narahari: Yeah, so… Let’s assume… Okay.
213 00:18:42.110 ⇒ 00:18:47.990 Pranav Narahari: And if they’re the ones blocking this, then of course we’re not going to be held to, like, the…
214 00:18:48.220 ⇒ 00:18:51.610 Pranav Narahari: We’re not going to be held to, like, the estimate that’s provided here.
215 00:18:51.610 ⇒ 00:18:53.259 Samuel Roberts: Sure, sure, sure, yeah.
216 00:18:53.260 ⇒ 00:19:01.819 Pranav Narahari: So, like, I guess the assumption should be that, like, once they make all the connections, which is completely out of our hands, how then can we…
217 00:19:02.260 ⇒ 00:19:03.590 Pranav Narahari: proceed.
218 00:19:04.260 ⇒ 00:19:04.950 Samuel Roberts: Yeah.
219 00:19:04.950 ⇒ 00:19:07.799 Pranav Narahari: With, like, the actual, MVP.
220 00:19:07.970 ⇒ 00:19:13.199 Pranav Narahari: And I guess, to be fair, too, like, part of this is on us as well, is like.
221 00:19:13.470 ⇒ 00:19:19.520 Pranav Narahari: When they create a new brand, some of that authentication stuff and connection should be…
222 00:19:19.770 ⇒ 00:19:23.609 Pranav Narahari: We should see what we can do via the Stitch platform as well.
223 00:19:23.930 ⇒ 00:19:24.440 Pranav Narahari: Oh.
224 00:19:24.440 ⇒ 00:19:43.130 Samuel Roberts: That’s what I was thinking, yeah. That’s why I was a little nervous about the way Polytomic was doing it, but I figured there was a way to work around it once we had it set up, but… Like, once there was an account connected, I figured there’d be a way to see all the accounts on it programmatically, but I don’t know yet. But yeah, okay, that’s fine, I get that. Six hours on that might be a little…
225 00:19:43.420 ⇒ 00:19:45.450 Samuel Roberts: optimistic, but I think that’s fine.
226 00:19:46.680 ⇒ 00:19:48.129 Pranav Narahari: Oh, it’s only 6 hours.
227 00:19:48.590 ⇒ 00:19:49.910 Samuel Roberts: 1200, yeah.
228 00:19:50.450 ⇒ 00:19:52.500 Pranav Narahari: Oh, oh, I guess…
229 00:19:53.150 ⇒ 00:19:58.400 Pranav Narahari: Okay, yeah, if this is only the connection, yeah, 6 hours might just… That’s why I feel like it could be…
230 00:19:59.140 ⇒ 00:20:02.689 Samuel Roberts: It could be fine, yeah, it’s just… it’s just a mystery, you know, it’s a little…
231 00:20:02.870 ⇒ 00:20:03.799 Samuel Roberts: Okay. But I think it’s fine.
232 00:20:03.800 ⇒ 00:20:04.180 Pranav Narahari: I, I…
233 00:20:04.180 ⇒ 00:20:04.850 Samuel Roberts: We can work with that.
234 00:20:04.850 ⇒ 00:20:05.470 Pranav Narahari: Yeah, I agree.
235 00:20:05.580 ⇒ 00:20:17.619 Pranav Narahari: we’ll… we’ll work with it. If it goes… this is just an estimate, right? So, like, if it goes a little bit over, it’s okay, and, like, the whole idea with this, too, is, like, there’s some that I felt like were more conservative, some were less…
236 00:20:18.180 ⇒ 00:20:21.639 Pranav Narahari: I think we’ll be okay here a little bit.
237 00:20:22.020 ⇒ 00:20:22.630 Samuel Roberts: Right.
238 00:20:22.790 ⇒ 00:20:24.120 Pranav Narahari: Yeah, okay, cool.
239 00:20:26.810 ⇒ 00:20:27.520 Casie Aviles: And this is kind of.
240 00:20:27.520 ⇒ 00:20:30.280 Pranav Narahari: like, actually building the UI, dashboard.
241 00:20:30.280 ⇒ 00:20:40.160 Samuel Roberts: Yeah, 50 hours, that seems fine. Editable input, yeah, once the data’s all… yeah, I don’t think it’s a problem. Transition polyatomic mother up to paid accounts, revalidate everybody, yeah, okay, that all seems good.
242 00:20:40.830 ⇒ 00:20:49.980 Samuel Roberts: And then the last one is Future Enhancements, voice input, LangFuse Monster Readme… 25 hours…
243 00:20:51.650 ⇒ 00:20:52.210 Pranav Narahari: Yeah.
244 00:20:54.470 ⇒ 00:20:56.030 Samuel Roberts: Yeah…
245 00:20:57.460 ⇒ 00:20:57.920 Pranav Narahari: You know how, like.
246 00:20:57.920 ⇒ 00:20:58.440 Samuel Roberts: Okay.
247 00:20:58.440 ⇒ 00:21:02.079 Pranav Narahari: They may want to modify, like, some system prompt stuff, if they want to, like, also.
248 00:21:02.080 ⇒ 00:21:02.909 Samuel Roberts: Yeah, I still…
249 00:21:02.910 ⇒ 00:21:04.150 Pranav Narahari: Yes, yeah.
250 00:21:04.580 ⇒ 00:21:13.760 Samuel Roberts: I don’t know if length used… I mean, I still kind of have a feeling that if they’re gonna wanna do that, we might want to add a UI in Stitch for that, and store the prompts, but,
251 00:21:14.230 ⇒ 00:21:17.919 Samuel Roberts: I mean, lank views is fine, I just feel like giving them a separate UI is…
252 00:21:18.490 ⇒ 00:21:24.179 Samuel Roberts: not ideal, but if it works, it works, and they’re fine with it, you know, I’m not gonna argue that. I don’t know if there’s much,
253 00:21:24.640 ⇒ 00:21:34.569 Samuel Roberts: some monster… I don’t know… when we get into more agent workflows, yeah, monster makes sense, I just don’t know what those flows are gonna be yet, so I don’t even know if it’s worth, you know, until we get there. And then… Yeah.
254 00:21:34.570 ⇒ 00:21:35.149 Pranav Narahari: I think that might.
255 00:21:35.150 ⇒ 00:21:37.919 Samuel Roberts: Did you mean anything that it pulled from, like, the SOW. It was.
256 00:21:37.920 ⇒ 00:21:38.250 Pranav Narahari: So…
257 00:21:38.250 ⇒ 00:21:45.180 Samuel Roberts: It definitely was, yeah, and I was fine leaving it in there for that, I just… it was… we didn’t know exactly what we were getting into at that point. So I think that’s fine.
258 00:21:45.740 ⇒ 00:21:49.720 Samuel Roberts: You know, that’s probably an overestimate at this point, but it’s good for now.
259 00:21:49.720 ⇒ 00:21:51.530 Pranav Narahari: Perfect.
260 00:21:51.530 ⇒ 00:21:53.540 Samuel Roberts: little question. All right, cool.
261 00:21:53.840 ⇒ 00:21:59.770 Pranav Narahari: Cool. Can we do a quick, like, total for that? Yeah.
262 00:22:00.260 ⇒ 00:22:06.099 Pranav Narahari: for these 14 tasks, that are remaining, I guess. Like, so let’s filter out the ones that have been completed.
263 00:22:06.880 ⇒ 00:22:08.850 Samuel Roberts: Yeah, so just hours, horse-wise?
264 00:22:08.850 ⇒ 00:22:12.960 Pranav Narahari: Yeah, we just figure out the hours, and then… I just want to see how many weeks that would be.
265 00:22:13.770 ⇒ 00:22:14.370 Samuel Roberts: Okay.
266 00:22:16.730 ⇒ 00:22:20.869 Samuel Roberts: If you just highlight the columns with hours, it’ll show you in the corner, I think, what the sum is.
267 00:22:21.190 ⇒ 00:22:22.200 Casie Aviles: Oh, okay.
268 00:22:23.100 ⇒ 00:22:25.540 Samuel Roberts: And down here, 175 hours.
269 00:22:26.180 ⇒ 00:22:29.459 Pranav Narahari: But that’s probably including some of the ones that are completed, right?
270 00:22:29.780 ⇒ 00:22:34.289 Samuel Roberts: No, I think scroll… I think he filtered out those. Yeah, it’s just anything in progress.
271 00:22:34.820 ⇒ 00:22:36.190 Samuel Roberts: And… what’s this?
272 00:22:36.320 ⇒ 00:22:38.480 Samuel Roberts: I think Yeah, okay.
273 00:22:38.980 ⇒ 00:22:45.890 Pranav Narahari: Okay, 175 makes sense, 2 months to get everything ready. Yeah, okay.
274 00:22:46.290 ⇒ 00:22:47.710 Pranav Narahari: I’ll put that in again.
275 00:22:49.540 ⇒ 00:22:50.120 Samuel Roberts: Okay.
276 00:22:51.500 ⇒ 00:22:53.969 Samuel Roberts: Yeah, these seem like good estimates, I think.
277 00:22:54.760 ⇒ 00:23:00.640 Pranav Narahari: Perfect. Yeah, I think that overall makes sense, right? Like, all… Stuff getting done.
278 00:23:01.100 ⇒ 00:23:02.090 Pranav Narahari: once.
279 00:23:02.260 ⇒ 00:23:05.939 Pranav Narahari: We’ve been working at a way higher velocity for probably, like.
280 00:23:06.180 ⇒ 00:23:14.350 Pranav Narahari: 40-plus hours per week, now it’s gonna be 20. But I still think 2 months will… we’ll be able to get all this done, yeah.
281 00:23:15.520 ⇒ 00:23:18.539 Pranav Narahari: Any, any reservations there? Like, does that sound about right?
282 00:23:21.380 ⇒ 00:23:25.350 Samuel Roberts: Yeah… I think that’s fine,
283 00:23:26.800 ⇒ 00:23:28.589 Samuel Roberts: I’m just trying to think, like, is there a…
284 00:23:28.840 ⇒ 00:23:33.309 Samuel Roberts: I think we need to mitigate some of that, but I think, like, just planning ahead a little bit more is probably good.
285 00:23:33.500 ⇒ 00:23:34.950 Samuel Roberts: You know.
286 00:23:35.810 ⇒ 00:23:40.309 Samuel Roberts: Before we jump into certain features that might not be… you know, a lot of lives in progress, so I’m…
287 00:23:40.940 ⇒ 00:23:42.800 Pranav Narahari: It’ll be a… Yeah.
288 00:23:42.930 ⇒ 00:23:46.319 Pranav Narahari: In, like, our weekly calls, like, when we get, like.
289 00:23:46.450 ⇒ 00:23:49.129 Pranav Narahari: Hey, can we do this, can we do that? It’s gonna be just like…
290 00:23:49.520 ⇒ 00:23:59.719 Pranav Narahari: No, no, no, no, no. We’re gonna finish these things first. Yeah. So it’ll be a… it’ll be very different to how we were operating before.
291 00:23:59.720 ⇒ 00:24:00.330 Samuel Roberts: Yes.
292 00:24:00.330 ⇒ 00:24:05.580 Pranav Narahari: Like, the numbers were insane before, like, we were working, like, 60% of the tasks we were doing were…
293 00:24:05.700 ⇒ 00:24:09.409 Pranav Narahari: like, not as part of the SOW, so…
294 00:24:09.410 ⇒ 00:24:10.260 Samuel Roberts: Yeah, yeah.
295 00:24:10.260 ⇒ 00:24:17.679 Pranav Narahari: That’s gonna be completely different now. We’ll get a lot more structure, which I think will automatically just make us more efficient. We won’t be scrambling week to week.
296 00:24:17.860 ⇒ 00:24:23.120 Pranav Narahari: So… yeah. I think the next couple months should be much smoother.
297 00:24:23.930 ⇒ 00:24:24.830 Pranav Narahari: Hopefully.
298 00:24:25.470 ⇒ 00:24:26.539 Samuel Roberts: Cool, that sounds good.
299 00:24:26.540 ⇒ 00:24:27.180 Pranav Narahari: Yeah.
300 00:24:27.360 ⇒ 00:24:28.540 Pranav Narahari: Well, thanks, guys.
301 00:24:29.320 ⇒ 00:24:29.870 Samuel Roberts: Alright.
302 00:24:29.870 ⇒ 00:24:30.640 Pranav Narahari: Thanks.
303 00:24:31.310 ⇒ 00:24:31.850 Samuel Roberts: Have a good one.
304 00:24:31.850 ⇒ 00:24:32.350 Casie Aviles: Thank you.
305 00:24:33.110 ⇒ 00:24:33.979 Pranav Narahari: Thank you, thank you.
306 00:24:34.390 ⇒ 00:24:35.629 Pranav Narahari: Talk soon. See ya.
307 00:24:35.910 ⇒ 00:24:36.830 Samuel Roberts: Alright, bye.