Meeting Title: ABC Central Docs Improvement Planning Date: 2026-02-04 Meeting participants: Mustafa Raja, Casie Aviles, Pranav Narahari, Samuel Roberts, Amber Lin
WEBVTT
1 00:00:00.000 ⇒ 00:00:01.490 Samuel Roberts: I don’t know what this means.
2 00:00:02.130 ⇒ 00:00:03.969 Samuel Roberts: Q&A Standardizer?
3 00:00:04.590 ⇒ 00:00:06.320 Amber Lin: They have a map… Question and answers?
4 00:00:06.320 ⇒ 00:00:06.960 Samuel Roberts: or…
5 00:00:06.960 ⇒ 00:00:08.510 Amber Lin: Yeah, question and answer.
6 00:00:08.930 ⇒ 00:00:10.790 Samuel Roberts: I know it was quality assurance, I was like.
7 00:00:11.690 ⇒ 00:00:14.230 Samuel Roberts: They have a lot of… What?
8 00:00:19.200 ⇒ 00:00:21.539 Samuel Roberts: You said they had a lot of something? I didn’t catch the end of that.
9 00:00:22.460 ⇒ 00:00:25.279 Amber Lin: Oh, sorry, they had a lot of,
10 00:00:25.430 ⇒ 00:00:32.460 Amber Lin: As the Central Doc got updated, people asked questions, and the trainers just dumped it as a question and answer format.
11 00:00:32.640 ⇒ 00:00:33.409 Amber Lin: Like, I don’t know…
12 00:00:33.410 ⇒ 00:00:34.430 Samuel Roberts: Oh!
13 00:00:34.430 ⇒ 00:00:36.499 Amber Lin: having that, or they have other.
14 00:00:36.500 ⇒ 00:00:36.950 Samuel Roberts: commercials.
15 00:00:36.950 ⇒ 00:00:51.739 Amber Lin: It’s like, oh, do we do microwave in this? And they’ll have an answer, like, no, go ask this. So I guess Andy could benefit from, like, these explicit question and answer formats, so I don’t know how we’ll deal with them just yet.
16 00:00:52.140 ⇒ 00:00:57.370 Samuel Roberts: Yeah, I think that’s a… that’s a bigger, like… question about… you know.
17 00:00:57.780 ⇒ 00:01:07.959 Samuel Roberts: what’s the best format for some of this information? Is it just to say, like, heading microwave, you know, we don’t do microwaves, talk to this person, or is it, like, here are some possible questions, like…
18 00:01:08.110 ⇒ 00:01:11.840 Samuel Roberts: I think the way we’ve set this up at RAG is probably the first one, but…
19 00:01:14.440 ⇒ 00:01:16.379 Samuel Roberts: We could… we could definitely have something that’s standardized with that.
20 00:01:16.380 ⇒ 00:01:17.500 Mustafa Raja: Yeah, yeah. Okay.
21 00:01:17.500 ⇒ 00:01:19.509 Amber Lin: I’ll put it there. Yeah, go ahead.
22 00:01:19.510 ⇒ 00:01:28.629 Mustafa Raja: Yeah, I’m also wondering if, if we would want to do a weekly analysis of what were some of the most asked questions.
23 00:01:30.140 ⇒ 00:01:31.170 Mustafa Raja: make this…
24 00:01:31.170 ⇒ 00:01:31.510 Samuel Roberts: I think.
25 00:01:31.510 ⇒ 00:01:32.680 Mustafa Raja: do a better.
26 00:01:33.230 ⇒ 00:01:33.680 Amber Lin: Yeah.
27 00:01:33.680 ⇒ 00:01:36.880 Samuel Roberts: at some of the Google stuff, I think?
28 00:01:36.880 ⇒ 00:01:41.909 Amber Lin: here. Should we do, like, query categorization and report?
29 00:01:42.880 ⇒ 00:01:43.780 Mustafa Raja: Oh, yeah.
30 00:01:43.780 ⇒ 00:01:45.240 Samuel Roberts: I thought that was kind of part of that.
31 00:01:45.710 ⇒ 00:01:48.430 Samuel Roberts: Yeah, let’s do weekly report, let’s say weekly report.
32 00:01:49.100 ⇒ 00:01:50.280 Mustafa Raja: So we know…
33 00:01:50.280 ⇒ 00:01:52.259 Amber Lin: Yeah, the client would really like that.
34 00:01:53.850 ⇒ 00:01:55.360 Samuel Roberts: Yeah, that’s good. Yeah, beautiful.
35 00:01:55.360 ⇒ 00:02:07.960 Amber Lin: Awesome. And then this, like, tag generator, do you think it’s important? I guess, like, it’s just an add-on of our current, Supabase, like, rag stuff.
36 00:02:08.810 ⇒ 00:02:12.310 Amber Lin: So… I would put it as low, but open.
37 00:02:12.310 ⇒ 00:02:15.640 Mustafa Raja: I want to understand the humor first, maybe a few discussions.
38 00:02:15.640 ⇒ 00:02:16.090 Samuel Roberts: I think…
39 00:02:16.090 ⇒ 00:02:16.720 Mustafa Raja: loose.
40 00:02:17.410 ⇒ 00:02:27.200 Samuel Roberts: Yeah, I think we can maybe come up with some more ideas for it, but I think also with the, like, query categorization, we might realize what tags would be useful, so I think that’s a general thing.
41 00:02:27.200 ⇒ 00:02:27.600 Amber Lin: Okay.
42 00:02:28.300 ⇒ 00:02:36.849 Amber Lin: Awesome. Now we have, I think these two sections, because that’s a lot later. So just these two sections,
43 00:02:38.120 ⇒ 00:02:41.049 Amber Lin: I think this is when people add new content.
44 00:02:41.560 ⇒ 00:02:44.860 Amber Lin: I don’t… like, this is… Hmm.
45 00:02:46.040 ⇒ 00:02:50.780 Amber Lin: Okay. I feel like we should name these better, but… Like, this is…
46 00:02:50.780 ⇒ 00:02:51.350 Samuel Roberts: content.
47 00:02:51.350 ⇒ 00:02:56.020 Amber Lin: analysis… Go ahead.
48 00:02:56.900 ⇒ 00:03:01.859 Samuel Roberts: I’m just trying to think, like, the new content validator, I don’t know how we know when it’s new content, necessarily.
49 00:03:02.210 ⇒ 00:03:11.050 Samuel Roberts: But it definitely could be… when something gets added, it might trick… or, trip the contradictory duplicate content detector.
50 00:03:12.310 ⇒ 00:03:21.780 Amber Lin: Yeah. Like, right now, we have the… we get a Slack report when the rag is updated, so meaning, like, new content or edits were made, like…
51 00:03:21.780 ⇒ 00:03:22.290 Samuel Roberts: Exactly.
52 00:03:22.290 ⇒ 00:03:22.860 Amber Lin: quote?
53 00:03:23.980 ⇒ 00:03:29.279 Samuel Roberts: So yeah, we can pull… we could actually, yeah, I’m sure… because right now we just… we… do we re-embed everything?
54 00:03:29.420 ⇒ 00:03:33.579 Mustafa Raja: Yeah, no, not everything. So it’s a difficult calculator.
55 00:03:33.800 ⇒ 00:03:34.450 Mustafa Raja: So it kind of.
56 00:03:34.450 ⇒ 00:03:35.409 Samuel Roberts: Okay, okay, good.
57 00:03:35.410 ⇒ 00:03:39.040 Mustafa Raja: So, you know, updated deletes those sections only.
58 00:03:39.780 ⇒ 00:03:41.829 Mustafa Raja: And re-embed those only.
59 00:03:42.620 ⇒ 00:03:47.460 Samuel Roberts: Good. Yeah, so then we can definitely do something where, like, we could take a look at that diff.
60 00:03:47.990 ⇒ 00:03:51.019 Samuel Roberts: See if it is already contained somewhere.
61 00:03:51.600 ⇒ 00:03:52.250 Amber Lin: Yeah, yeah.
62 00:03:52.250 ⇒ 00:03:56.139 Samuel Roberts: You know, that, that makes sure it’s not doing something that’s already, or, you know.
63 00:03:56.760 ⇒ 00:04:00.519 Samuel Roberts: validate whatever other language we want to add to it. Yeah, I like that.
64 00:04:00.700 ⇒ 00:04:07.310 Amber Lin: Yeah. So should that be the same as this policy adherence checker? So that will also be, like, hey.
65 00:04:07.440 ⇒ 00:04:08.010 Samuel Roberts: is referred.
66 00:04:08.010 ⇒ 00:04:13.200 Amber Lin: formatting making sense? Is that making sense? Like, should this be combined?
67 00:04:13.970 ⇒ 00:04:16.220 Samuel Roberts: Yeah, if that’s what they mean by policy, yeah.
68 00:04:16.860 ⇒ 00:04:17.560 Amber Lin: Okay.
69 00:04:17.560 ⇒ 00:04:18.240 Samuel Roberts: What was I must stop?
70 00:04:19.149 ⇒ 00:04:24.079 Mustafa Raja: Yeah, I was saying this new content validator would be checking against some policy, right?
71 00:04:24.930 ⇒ 00:04:25.510 Samuel Roberts: Yeah.
72 00:04:26.810 ⇒ 00:04:32.080 Amber Lin: Cool, so let me… Delete this… Great.
73 00:04:32.220 ⇒ 00:04:37.040 Amber Lin: And then… Do we?
74 00:04:40.210 ⇒ 00:04:41.440 Samuel Roberts: Yeah, I think that’s…
75 00:04:42.740 ⇒ 00:04:47.139 Samuel Roberts: Like, an outcome, potentially, of the detector anyway, so, like, we’ll see if we get to that, if it…
76 00:04:47.750 ⇒ 00:04:48.550 Amber Lin: Okay.
77 00:04:48.850 ⇒ 00:04:49.890 Amber Lin: So this…
78 00:04:49.890 ⇒ 00:04:52.209 Samuel Roberts: It’s like, hey, these things look like they’re duplicates, can, you know…
79 00:04:52.830 ⇒ 00:05:05.060 Amber Lin: Let me see. The new content validator, I think, is… it’s important, right? Because that just reduces the chaos that increases. I feel like we… we would be…
80 00:05:05.290 ⇒ 00:05:07.950 Amber Lin: Like, we first need a policy.
81 00:05:08.050 ⇒ 00:05:15.669 Amber Lin: And then we need a structure. Like, let me add a… Policy for medicine.
82 00:05:16.390 ⇒ 00:05:17.370 Samuel Roberts: How,
83 00:05:22.120 ⇒ 00:05:23.240 Samuel Roberts: Oh, yeah.
84 00:05:24.030 ⇒ 00:05:25.430 Samuel Roberts: I think,
85 00:05:27.000 ⇒ 00:05:32.780 Samuel Roberts: Do you guys have a sense of how many, like, how big these documents are in terms of, like, a context window, if we were to…
86 00:05:33.280 ⇒ 00:05:37.260 Amber Lin: They’re, like, 100 pages?
87 00:05:37.360 ⇒ 00:05:38.420 Amber Lin: 70 to 100.
88 00:05:38.420 ⇒ 00:05:38.880 Samuel Roberts: ages.
89 00:05:39.960 ⇒ 00:05:47.580 Amber Lin: Like, the shorter… the shortest ones are, like, 50, 60, I think. PES are about 70, mechanicals 100-something.
90 00:05:47.670 ⇒ 00:05:48.480 Mustafa Raja: Yo.
91 00:05:48.480 ⇒ 00:05:49.030 Samuel Roberts: Okay.
92 00:05:49.210 ⇒ 00:05:56.459 Mustafa Raja: That’s one thing, and then for embeddings, I think we have a character limit of 3000 or something.
93 00:05:57.160 ⇒ 00:06:04.189 Mustafa Raja: I feel it might be… 3,000, seems like, on the shorter side. We might also want to increase that.
94 00:06:06.020 ⇒ 00:06:11.780 Samuel Roberts: Okay. I’m wondering… I’m just trying to think through, like, how… if we like the structure of pests and want to…
95 00:06:12.180 ⇒ 00:06:19.019 Samuel Roberts: pull out what makes that good? Like, how much of that can we dump into a context window of, like, Opus or something?
96 00:06:20.340 ⇒ 00:06:27.189 Pranav Narahari: I just looked it up, a 100-page tech stock is roughly 60K tokens, which we could definitely…
97 00:06:27.480 ⇒ 00:06:28.210 Pranav Narahari: Toss it.
98 00:06:28.210 ⇒ 00:06:28.960 Samuel Roberts: Okay.
99 00:06:29.210 ⇒ 00:06:37.039 Samuel Roberts: Awesome. So yeah, my thought is, like, at a very first level, just, like, what kind of stuff can we learn about this stock versus the other ones?
100 00:06:37.260 ⇒ 00:06:46.339 Amber Lin: Yeah. Oh, I also ran it… when I was generating this notion, I also ran it through, like, I ran both the pest and mechanical through, like.
101 00:06:46.340 ⇒ 00:06:47.339 Samuel Roberts: Oh, nice, okay.
102 00:06:47.340 ⇒ 00:06:48.670 Amber Lin: So, it should be fine.
103 00:06:49.000 ⇒ 00:06:52.160 Samuel Roberts: Alright, cool, yeah, I think that’ll be… then we can get some good structured recommends on them.
104 00:06:52.160 ⇒ 00:06:58.880 Amber Lin: Yeah, okay. For the structure, do you think it’s the same thing as a central dog policy?
105 00:06:59.360 ⇒ 00:07:09.049 Amber Lin: Like, does it… is it just the same thing, or should we do this recommendation much earlier? So, like, after we do the analysis, and then we’ll do this recommendation?
106 00:07:10.590 ⇒ 00:07:12.499 Amber Lin: Maybe I’ll go move it here.
107 00:07:16.890 ⇒ 00:07:20.450 Samuel Roberts: Yeah, I guess I don’t necessarily know the difference now that you… Stay there.
108 00:07:20.450 ⇒ 00:07:33.170 Amber Lin: Yeah, this is more so, like, early on, just tell the… like, I guess it’s, like, a deliverable for the client. I can make this a milestone of, like, hey, analysis results and recommendations.
109 00:07:33.720 ⇒ 00:07:38.489 Samuel Roberts: Yeah, but yeah, I mean, I think that’s… yeah, that’s fine. I feel like it’s very similar to, like.
110 00:07:39.880 ⇒ 00:07:46.020 Samuel Roberts: the central dock policy will have… like, this is gonna have to be, like, a living central dock policy anyway, as we find things that…
111 00:07:46.180 ⇒ 00:07:49.169 Samuel Roberts: You know, but I think that, yeah, changing this is better, that’s good. Okay.
112 00:07:49.170 ⇒ 00:07:52.780 Amber Lin: Cool, okay. How do I make this a milestone, by the way?
113 00:07:52.780 ⇒ 00:07:54.289 Samuel Roberts: I think you gotta go into it.
114 00:07:54.780 ⇒ 00:07:55.480 Amber Lin: Okay.
115 00:07:56.330 ⇒ 00:07:58.589 Samuel Roberts: And then… it might be up here?
116 00:07:58.980 ⇒ 00:08:00.639 Samuel Roberts: Yeah, I forgot what it looked like.
117 00:08:01.260 ⇒ 00:08:02.080 Amber Lin: Okay.
118 00:08:04.620 ⇒ 00:08:10.760 Amber Lin: Alright. Okay. And then, our last part, which I think…
119 00:08:11.230 ⇒ 00:08:13.780 Amber Lin: this, what I wanted to do.
120 00:08:14.560 ⇒ 00:08:15.539 Samuel Roberts: Yeah, I mean…
121 00:08:15.540 ⇒ 00:08:16.080 Amber Lin: Hmm.
122 00:08:17.750 ⇒ 00:08:20.550 Samuel Roberts: All of those are pretty… useful.
123 00:08:20.550 ⇒ 00:08:21.730 Amber Lin: Did I just put…
124 00:08:21.970 ⇒ 00:08:27.520 Amber Lin: important for all of these. I don’t really know what the regression test suite, I think, is more.
125 00:08:27.520 ⇒ 00:08:32.070 Samuel Roberts: What is eval data set that runs weekly? Like, certain questions we want to check?
126 00:08:32.539 ⇒ 00:08:38.219 Amber Lin: I believe so, yes. Utam said… The docs. Okay. Yes.
127 00:08:39.230 ⇒ 00:08:39.780 Samuel Roberts: Okay.
128 00:08:40.150 ⇒ 00:08:45.180 Samuel Roberts: Yeah, I think… We’ve sort of started that, like.
129 00:08:45.310 ⇒ 00:08:50.540 Samuel Roberts: Where we have a little bit of a golden data set that’s not really too golden yet, but… .
130 00:08:50.540 ⇒ 00:08:54.300 Amber Lin: I guess once we do the, like, this, the…
131 00:08:54.300 ⇒ 00:08:54.620 Samuel Roberts: Yeah.
132 00:08:54.620 ⇒ 00:08:55.440 Amber Lin: Sweet.
133 00:08:56.780 ⇒ 00:09:02.599 Amber Lin: I’m gonna say, like, question categorization. The query trips me up.
134 00:09:03.480 ⇒ 00:09:03.909 Samuel Roberts: That’s fine.
135 00:09:03.910 ⇒ 00:09:11.590 Amber Lin: Okay, so… We’ll see this is 1, 2, 5… Live.
136 00:09:12.320 ⇒ 00:09:13.380 Amber Lin: Ugh.
137 00:09:15.400 ⇒ 00:09:16.440 Amber Lin: Okay.
138 00:09:20.080 ⇒ 00:09:22.300 Samuel Roberts: Cross-department consistency.
139 00:09:22.500 ⇒ 00:09:32.380 Amber Lin: They have some similar stuff, but maybe it doesn’t belong to this section. Like, different departments will have the same, like, hey, this is how you do a,
140 00:09:32.380 ⇒ 00:09:34.789 Samuel Roberts: Oh, I see what you’re saying, yes, okay.
141 00:09:34.790 ⇒ 00:09:40.610 Amber Lin: And then they have the same building stuff, the same cancellation stuff, so, like, they want to.
142 00:09:40.610 ⇒ 00:09:41.020 Samuel Roberts: combined.
143 00:09:41.020 ⇒ 00:09:47.170 Amber Lin: the departments eventually, and then some departments have different stuff, so, like, I don’t know if it’s…
144 00:09:47.170 ⇒ 00:09:48.910 Samuel Roberts: I would probably put that up to here, then.
145 00:09:50.580 ⇒ 00:09:51.260 Amber Lin: Oh, okay.
146 00:09:51.260 ⇒ 00:09:53.440 Samuel Roberts: It feels like more formatting to me than testing.
147 00:09:53.440 ⇒ 00:09:54.060 Mustafa Raja: Nope.
148 00:09:54.060 ⇒ 00:09:54.780 Amber Lin: Okay.
149 00:09:55.080 ⇒ 00:09:55.960 Amber Lin: Great.
150 00:09:56.650 ⇒ 00:09:57.620 Mustafa Raja: Okay.
151 00:09:57.620 ⇒ 00:10:00.600 Amber Lin: And then these… should I just put a blanket?
152 00:10:02.110 ⇒ 00:10:04.090 Amber Lin: What would you assign?
153 00:10:04.090 ⇒ 00:10:08.869 Samuel Roberts: I mean, I think… I think the idea here is good, in terms of it’ll,
154 00:10:09.040 ⇒ 00:10:13.179 Samuel Roberts: improve Andy, but I don’t think it’s gonna fix the Central Doc stuff.
155 00:10:13.620 ⇒ 00:10:16.340 Amber Lin: Yeah, okay, so I’ll leave it there.
156 00:10:16.910 ⇒ 00:10:21.550 Samuel Roberts: Yeah, I think, like, you know, like, doing re-ranks and things will be very helpful
157 00:10:21.750 ⇒ 00:10:25.860 Samuel Roberts: once the data’s good and we know it’s being asked, but I don’t think that really…
158 00:10:26.130 ⇒ 00:10:29.050 Samuel Roberts: Is gonna move the needle on anything right now.
159 00:10:29.890 ⇒ 00:10:33.440 Amber Lin: Awesome. Okay. Let’s…
160 00:10:37.670 ⇒ 00:10:41.539 Amber Lin: Okay, these are ongoing.
161 00:10:42.330 ⇒ 00:10:49.240 Amber Lin: Alright, and then, adding all zip code sheets.
162 00:10:51.730 ⇒ 00:10:54.030 Amber Lin: Casey’s currently doing that.
163 00:11:00.190 ⇒ 00:11:07.770 Amber Lin: And then… Validating missing code assignment.
164 00:11:12.270 ⇒ 00:11:15.019 Amber Lin: Well, that’s also very important.
165 00:11:17.450 ⇒ 00:11:20.350 Amber Lin: Alright, so…
166 00:11:21.120 ⇒ 00:11:31.219 Amber Lin: We have 20 minutes left. I want to use, like, 15 minutes to update this… this workflow of the scan. I think this is… this is something that’s ongoing.
167 00:11:31.580 ⇒ 00:11:35.550 Amber Lin: And then, I feel like the lawn stuff…
168 00:11:36.940 ⇒ 00:11:43.549 Amber Lin: Casey, do you think you’ll be able to start the lawn, like, validation this week, or should I start it next week?
169 00:11:44.850 ⇒ 00:11:46.430 Casie Aviles: Law invalidation.
170 00:11:47.180 ⇒ 00:11:49.060 Amber Lin: Yeah, the lawn technicians.
171 00:11:49.380 ⇒ 00:11:52.239 Amber Lin: That sheet. So this is, like, this includes…
172 00:11:52.730 ⇒ 00:11:56.049 Amber Lin: The lawn stuff, I can grab the ticket.
173 00:12:00.370 ⇒ 00:12:04.839 Casie Aviles: I think I can start adding it today. I can start adding it today.
174 00:12:05.970 ⇒ 00:12:09.640 Casie Aviles: And then I might finish… Tomorrow.
175 00:12:10.130 ⇒ 00:12:10.830 Amber Lin: Okay.
176 00:12:11.080 ⇒ 00:12:12.470 Casie Aviles: Sometime tomorrow.
177 00:12:12.990 ⇒ 00:12:13.550 Amber Lin: Cool.
178 00:12:14.810 ⇒ 00:12:17.159 Amber Lin: Copy this here…
179 00:12:22.320 ⇒ 00:12:23.930 Amber Lin: And then…
180 00:12:35.290 ⇒ 00:12:37.110 Amber Lin: Cool, so I’ll start this.
181 00:12:38.400 ⇒ 00:12:42.000 Amber Lin: Here, we’re maybe more expanded if needed.
182 00:12:42.260 ⇒ 00:12:50.130 Amber Lin: Alright, and then, on the transcript side, I know…
183 00:12:51.190 ⇒ 00:12:58.199 Amber Lin: Sam, you were… are you working on a… is there another task I should add here? I know you’re already working on this.
184 00:12:59.070 ⇒ 00:13:01.000 Samuel Roberts: Yeah, I mean, that first one is,
185 00:13:01.680 ⇒ 00:13:04.799 Samuel Roberts: what I’m working on. I had a plan, it put some code together, it’s not…
186 00:13:05.270 ⇒ 00:13:06.650 Amber Lin: quite there yet.
187 00:13:06.950 ⇒ 00:13:07.850 Samuel Roberts: Okay.
188 00:13:07.850 ⇒ 00:13:12.730 Amber Lin: Do you think it will take until end of this week, midweek next week, and then next week?
189 00:13:13.970 ⇒ 00:13:18.829 Samuel Roberts: Let’s say mid-next week, as a conservative estimate.
190 00:13:18.940 ⇒ 00:13:27.039 Samuel Roberts: I think I’ve got transcripts coming, I just want to make sure that, like, we’re getting all the data with it, and then I need to run it and everything too, and that could take a while, so…
191 00:13:27.220 ⇒ 00:13:33.749 Amber Lin: Awesome, okay. And then after that is… are these accurate tasks for you?
192 00:13:36.060 ⇒ 00:13:38.139 Samuel Roberts: To be honest, I have no idea.
193 00:13:38.660 ⇒ 00:13:40.040 Amber Lin: I think, like.
194 00:13:40.150 ⇒ 00:13:45.280 Samuel Roberts: we were working on this stuff, and I was throwing some things together. I don’t really know, like, what the full…
195 00:13:47.360 ⇒ 00:13:49.940 Samuel Roberts: Scope of this is gonna look like at this point?
196 00:13:49.940 ⇒ 00:13:51.299 Amber Lin: I see. Okay.
197 00:13:51.300 ⇒ 00:13:55.270 Samuel Roberts: Like, Utam literally just said on Slack to, like, you know, start pulling those again.
198 00:13:55.270 ⇒ 00:13:56.710 Amber Lin: Yeah, makes sense.
199 00:13:56.710 ⇒ 00:13:57.409 Samuel Roberts: on hold.
200 00:13:57.410 ⇒ 00:14:00.970 Amber Lin: want to tell them that, hey, we have the 8x8 stuff.
201 00:14:00.970 ⇒ 00:14:01.669 Samuel Roberts: Exactly, yeah.
202 00:14:01.690 ⇒ 00:14:02.599 Amber Lin: That’s all we want.
203 00:14:02.600 ⇒ 00:14:04.600 Samuel Roberts: That’s something I wouldn’t… I’m not even sure I would…
204 00:14:04.760 ⇒ 00:14:07.820 Samuel Roberts: schedule all that yet until I finish this.
205 00:14:07.820 ⇒ 00:14:08.150 Amber Lin: Okay.
206 00:14:08.150 ⇒ 00:14:08.470 Samuel Roberts: Like, this…
207 00:14:08.470 ⇒ 00:14:09.140 Amber Lin: Oh, my friend, there’s no…
208 00:14:09.140 ⇒ 00:14:14.600 Samuel Roberts: priority, but yeah, that’s fine. Yeah, the highest one is getting it into the database, certainly, and getting the sync and everything.
209 00:14:14.820 ⇒ 00:14:15.590 Samuel Roberts: And we can…
210 00:14:15.590 ⇒ 00:14:16.709 Amber Lin: We can figure out…
211 00:14:16.710 ⇒ 00:14:22.129 Samuel Roberts: full scope after that, yeah. But I think this is not… Yeah, I think it’s good.
212 00:14:23.240 ⇒ 00:14:24.120 Amber Lin: Okay.
213 00:14:24.440 ⇒ 00:14:34.289 Amber Lin: And then over here. So, like, when should we start? I think we have this one’s super important, that one’s super important.
214 00:14:34.720 ⇒ 00:14:35.540 Amber Lin: I guess.
215 00:14:35.540 ⇒ 00:14:36.429 Samuel Roberts: Yeah, let’s…
216 00:14:36.430 ⇒ 00:14:39.970 Amber Lin: These two sections, we’re gonna have to do, like, this.
217 00:14:40.070 ⇒ 00:14:41.480 Amber Lin: these few weeks.
218 00:14:42.610 ⇒ 00:14:48.480 Samuel Roberts: Yeah, I would say… The contradiction Duplicate Content Detector to get started on that.
219 00:14:48.880 ⇒ 00:14:52.110 Samuel Roberts: Get started on… categorization…
220 00:14:57.000 ⇒ 00:14:58.169 Samuel Roberts: What else is hot up here?
221 00:15:00.560 ⇒ 00:15:08.740 Samuel Roberts: Maybe the… formatter… As well, because I think that’ll help us with a lot.
222 00:15:09.590 ⇒ 00:15:11.080 Samuel Roberts: Like, it’s highly rich, so…
223 00:15:11.740 ⇒ 00:15:22.080 Samuel Roberts: Because I think if we can, like, the process that’s gonna go into this, in terms of, like, figuring out what makes it good and being able to pass that to the AI to help us, I think that will be a very high leverage thing.
224 00:15:22.420 ⇒ 00:15:24.920 Amber Lin: Okay. Like, the content is one thing, but the format…
225 00:15:24.920 ⇒ 00:15:28.430 Samuel Roberts: Like, even if we don’t get to the full… like…
226 00:15:28.790 ⇒ 00:15:32.620 Samuel Roberts: it structures it for us. Just the first part of learning
227 00:15:33.450 ⇒ 00:15:38.040 Samuel Roberts: what’s gonna be good, and, like, having the AI know that is gonna be a huge…
228 00:15:38.760 ⇒ 00:15:39.460 Amber Lin: Okay.
229 00:15:39.910 ⇒ 00:15:42.040 Amber Lin: So…
230 00:15:45.700 ⇒ 00:15:48.640 Amber Lin: I guess… where’s the…
231 00:15:49.920 ⇒ 00:16:01.010 Amber Lin: That’s the milestone there. When do we want to achieve that milestone of telling them, like, hey, this is… this is our initial finding, like, this is what we found.
232 00:16:01.010 ⇒ 00:16:10.429 Amber Lin: here’s what we need from you, here’s what we’re gonna do, like, when are we gonna be able to do that? You think 2 weeks is enough, or do we need more time?
233 00:16:20.330 ⇒ 00:16:26.050 Samuel Roberts: I mean, I think two weeks could be fine, I don’t know, Casey Mustafa, what are your thoughts? You guys know this stuff better than I do.
234 00:16:26.580 ⇒ 00:16:27.540 Samuel Roberts: In detail.
235 00:16:29.380 ⇒ 00:16:32.120 Samuel Roberts: I’m like, I have ideas about how to go about this stuff, I just… yeah.
236 00:16:32.690 ⇒ 00:16:35.119 Samuel Roberts: Exactly what the implementation’s gonna be.
237 00:16:36.460 ⇒ 00:16:39.659 Casie Aviles: So for… that’s for the query analysis and feedback, right?
238 00:16:41.310 ⇒ 00:16:44.510 Casie Aviles: Or which one are we estimating for? Sorry, I was.
239 00:16:44.510 ⇒ 00:16:45.199 Amber Lin: I was trying to figure out the…
240 00:16:45.200 ⇒ 00:16:47.739 Samuel Roberts: Like, the mile… like, when…
241 00:16:47.900 ⇒ 00:16:52.500 Samuel Roberts: I guess the real question is, like, how much are we trying to squeeze in, how quickly, right?
242 00:16:52.950 ⇒ 00:16:53.630 Amber Lin: Yeah.
243 00:16:53.830 ⇒ 00:16:54.750 Amber Lin: You’re right.
244 00:16:54.750 ⇒ 00:17:02.400 Samuel Roberts: So, like, because once we do categorization, we can report the unanswered questions, we can do the thumbs-down analysis better.
245 00:17:03.330 ⇒ 00:17:03.850 Amber Lin: Mmm.
246 00:17:03.850 ⇒ 00:17:05.169 Samuel Roberts: Like, we can get that. So, like, we can kind of.
247 00:17:05.170 ⇒ 00:17:05.750 Casie Aviles: Like…
248 00:17:07.130 ⇒ 00:17:09.670 Samuel Roberts: fit some of those in. So I would say maybe, like.
249 00:17:10.220 ⇒ 00:17:12.440 Samuel Roberts: Two weeks for a preliminary one, maybe?
250 00:17:12.609 ⇒ 00:17:16.400 Samuel Roberts: Like, obviously this will be more ongoing, too, because we’ll have the analysis happening, but…
251 00:17:18.500 ⇒ 00:17:28.409 Amber Lin: So I’ll say, like, this… goes here… This also goes here.
252 00:17:30.640 ⇒ 00:17:33.100 Amber Lin: This also goes here.
253 00:17:40.530 ⇒ 00:17:45.860 Samuel Roberts: Bit of a wrench, migration stuff in terms of where the data’s gonna be when we’re doing this, but…
254 00:17:46.100 ⇒ 00:17:47.260 Amber Lin: Yeah…
255 00:17:47.690 ⇒ 00:17:48.610 Samuel Roberts: We’ll have to.
256 00:17:49.580 ⇒ 00:17:51.679 Samuel Roberts: sort that out anyway, so… okay.
257 00:17:51.680 ⇒ 00:17:58.430 Amber Lin: Okay, cool. So I’ll put the milestone there. Like, I think we’ll report to the clients once we do…
258 00:17:58.630 ⇒ 00:18:01.949 Amber Lin: This, and then…
259 00:18:02.290 ⇒ 00:18:02.980 Samuel Roberts: Yeah.
260 00:18:03.030 ⇒ 00:18:06.239 Amber Lin: Tell them what’s up, and then do these.
261 00:18:07.350 ⇒ 00:18:08.230 Amber Lin: Okay.
262 00:18:09.500 ⇒ 00:18:16.460 Amber Lin: And then, like, I think after we do these, the audit, like, the bi-weekly audit would go on.
263 00:18:17.190 ⇒ 00:18:18.850 Amber Lin: Make your fancy.
264 00:18:19.080 ⇒ 00:18:20.290 Amber Lin: Ongoing stuff.
265 00:18:20.290 ⇒ 00:18:23.240 Samuel Roberts: Yeah, I wouldn’t put it like that, though, because, like, we gotta…
266 00:18:23.390 ⇒ 00:18:26.780 Samuel Roberts: build the thing that’s gonna do the auditing, right? That’s really what this is.
267 00:18:26.780 ⇒ 00:18:27.960 Amber Lin: That’s true.
268 00:18:29.360 ⇒ 00:18:34.270 Samuel Roberts: Don’t delete it, don’t delete it, no, no. Just, like, so yeah, I would say just, like, shrink that to, like, maybe…
269 00:18:35.000 ⇒ 00:18:35.790 Amber Lin: A week?
270 00:18:36.590 ⇒ 00:18:37.710 Samuel Roberts: Yeah.
271 00:18:38.860 ⇒ 00:18:39.445 Amber Lin: 2…
272 00:18:42.760 ⇒ 00:18:49.309 Amber Lin: And then… I don’t know when we’re doing this with the ambiguity score calculator.
273 00:18:58.350 ⇒ 00:18:59.210 Amber Lin: Okay.
274 00:18:59.500 ⇒ 00:19:01.779 Amber Lin: What about these?
275 00:19:05.690 ⇒ 00:19:08.339 Amber Lin: And when are we doing the testing stuff?
276 00:19:08.820 ⇒ 00:19:14.439 Amber Lin: Like, how early should we start doing the testing stuff? I assume before we do the changes?
277 00:19:15.050 ⇒ 00:19:15.970 Amber Lin: whereas we’re.
278 00:19:15.970 ⇒ 00:19:16.460 Samuel Roberts: Yeah.
279 00:19:16.460 ⇒ 00:19:17.569 Amber Lin: I mean.
280 00:19:17.570 ⇒ 00:19:20.420 Samuel Roberts: As, I think that should be happening so that we have the data, you know?
281 00:19:20.680 ⇒ 00:19:22.859 Amber Lin: Okay, so when will we start that?
282 00:19:23.860 ⇒ 00:19:27.210 Samuel Roberts: What do we, scroll up a little bit, too, see the rest of the stuff.
283 00:19:27.510 ⇒ 00:19:30.280 Samuel Roberts: I’m just trying to think, like, what’s happening next week, overall.
284 00:19:30.780 ⇒ 00:19:31.440 Amber Lin: Hmm.
285 00:19:37.110 ⇒ 00:19:39.790 Samuel Roberts: Yeah, okay, I think we can maybe do…
286 00:19:40.290 ⇒ 00:19:57.059 Amber Lin: Yeah, I wanna… can we just consider for the central doc stuff, and then we can look at the contradictions, and then shift things back. I mostly just want to know, like, if there’s dependencies, if this should happen in parallel with these tasks, or with these tasks, or just dependent and have.
287 00:19:57.060 ⇒ 00:20:01.639 Samuel Roberts: Yeah, I don’t think there’s any dependencies for this. This stuff is just gonna influence, like, what we know about it.
288 00:20:01.720 ⇒ 00:20:06.080 Amber Lin: And so my thought was just, like, I want to get that done sooner rather than later, I just wasn’t sure.
289 00:20:06.080 ⇒ 00:20:09.979 Samuel Roberts: like, if we had other things we needed to bump, but I think next week, we could get started on…
290 00:20:12.920 ⇒ 00:20:13.920 Samuel Roberts: Yeah…
291 00:20:14.490 ⇒ 00:20:15.370 Amber Lin: Okay.
292 00:20:15.370 ⇒ 00:20:20.049 Samuel Roberts: Dataset and regression, and then A-B test doc changes is a much bigger…
293 00:20:20.710 ⇒ 00:20:21.470 Amber Lin: Yeah.
294 00:20:21.470 ⇒ 00:20:27.100 Samuel Roberts: potential thing, but it’s gonna depend on these two anyway to do that, so… This one is definitely dependent on…
295 00:20:27.250 ⇒ 00:20:28.489 Samuel Roberts: regression test.
296 00:20:28.650 ⇒ 00:20:33.150 Samuel Roberts: And our graphing test is dependent on createEval set, so I guess that makes sense, yeah. Okay. Those are pretty, pretty…
297 00:20:33.150 ⇒ 00:20:40.649 Amber Lin: Just to… Let’s do this… Alright, that’s good.
298 00:20:40.780 ⇒ 00:20:44.160 Amber Lin: Workflow and maintenance.
299 00:20:47.410 ⇒ 00:20:49.820 Amber Lin: I guess this is much later.
300 00:20:50.360 ⇒ 00:20:52.719 Amber Lin: Auto deduplication workflow.
301 00:20:53.370 ⇒ 00:20:56.100 Samuel Roberts: Yeah, new content validator, policy nutrients checker…
302 00:20:56.440 ⇒ 00:20:59.630 Samuel Roberts: The central dock policy, I think…
303 00:21:00.010 ⇒ 00:21:02.409 Amber Lin: We can do, like, right here, after this.
304 00:21:02.410 ⇒ 00:21:05.970 Samuel Roberts: Yeah, I think so. Like, we’ll probably have something…
305 00:21:06.260 ⇒ 00:21:09.060 Samuel Roberts: Like, work in progress up with that, if that makes sense?
306 00:21:09.330 ⇒ 00:21:14.059 Samuel Roberts: Because we’ll need something for other things, but it won’t be, like, a finalized, kind of…
307 00:21:14.390 ⇒ 00:21:17.909 Amber Lin: Cool, alright. And then, like, we can have…
308 00:21:19.060 ⇒ 00:21:26.239 Amber Lin: This… these stuff, I’ll just put it later. We can probably… Locate it again later.
309 00:21:27.350 ⇒ 00:21:33.960 Amber Lin: Alright, cross-department dependency can go much, much later.
310 00:21:34.430 ⇒ 00:21:38.330 Amber Lin: can probably also go later.
311 00:21:38.990 ⇒ 00:21:41.340 Amber Lin: Wording optimizer…
312 00:21:45.490 ⇒ 00:21:48.110 Amber Lin: I don’t know, when should we have these two?
313 00:21:48.110 ⇒ 00:21:51.679 Samuel Roberts: What is… Rewarding optimizer questionnaire.
314 00:21:53.040 ⇒ 00:21:56.929 Casie Aviles: Probably after we do the policies.
315 00:21:57.860 ⇒ 00:21:59.409 Samuel Roberts: Yeah, I think that’s a good point.
316 00:21:59.560 ⇒ 00:22:00.800 Amber Lin: Right, makes sense.
317 00:22:15.060 ⇒ 00:22:22.990 Amber Lin: Okay. Cool. Cool, that’s everything, so, A little bit of… P.
318 00:22:23.500 ⇒ 00:22:29.220 Amber Lin: 3 more minutes, so we get, like, a 5-minute break in between. So this week.
319 00:22:29.490 ⇒ 00:22:33.839 Amber Lin: Are these done? Like, are these stuff done? I know they were started.
320 00:22:36.480 ⇒ 00:22:38.940 Casie Aviles: Mmm… the edge case one is…
321 00:22:39.790 ⇒ 00:22:42.399 Casie Aviles: Yeah, that’s going to be ongoing.
322 00:22:42.600 ⇒ 00:22:43.400 Samuel Roberts: Yeah.
323 00:22:44.390 ⇒ 00:22:45.810 Samuel Roberts: The clouds…
324 00:22:46.680 ⇒ 00:22:55.970 Mustafa Raja: I think there’s this… Yeah, so set up Cloud SQL, we are going to send an email to Tim.
325 00:22:55.970 ⇒ 00:22:56.400 Samuel Roberts: Yes.
326 00:22:56.400 ⇒ 00:23:01.400 Mustafa Raja: Just wanted to, you know, verify, the prices with them.
327 00:23:02.360 ⇒ 00:23:05.240 Mustafa Raja: I think if you… Is Tim the one to ask? I wasn’t sure.
328 00:23:05.480 ⇒ 00:23:09.949 Mustafa Raja: Sorry? Oh, yeah, Utem said that we should pinged him about that.
329 00:23:10.640 ⇒ 00:23:13.270 Samuel Roberts: Okay, okay, cool. I just wasn’t sure if he was the one to, like.
330 00:23:13.270 ⇒ 00:23:19.130 Mustafa Raja: Oh, yeah, okay. Yeah, yeah. Okay, cool. Just confirm that. Okay, so if we scroll that,
331 00:23:19.390 ⇒ 00:23:26.160 Mustafa Raja: This should be, yeah, in the latency, 31. This should be good. We, yesterday I set up the hook.
332 00:23:27.740 ⇒ 00:23:30.240 Mustafa Raja: So, hooky setup, this should be done.
333 00:23:30.240 ⇒ 00:23:31.240 Amber Lin: Awesome.
334 00:23:32.210 ⇒ 00:23:34.420 Samuel Roberts: Was Black alerting done, too? Was that all part of that?
335 00:23:34.630 ⇒ 00:23:36.990 Mustafa Raja: Oh yeah, Slack. Slack also implemented.
336 00:23:37.700 ⇒ 00:23:41.390 Amber Lin: Yeah. Awesome, so I’m gonna scoot that back here.
337 00:23:41.390 ⇒ 00:23:41.950 Samuel Roberts: Exactly right.
338 00:23:41.950 ⇒ 00:23:43.749 Amber Lin: I’m gonna move…
339 00:23:43.750 ⇒ 00:23:48.189 Mustafa Raja: This evil logging in Marsha, it’s red. Can you scroll a little left?
340 00:23:48.890 ⇒ 00:23:50.349 Samuel Roberts: I think that was the…
341 00:23:51.000 ⇒ 00:23:51.949 Mustafa Raja: Whoa, hmm…
342 00:23:51.950 ⇒ 00:23:52.849 Samuel Roberts: Yeah, that kind of…
343 00:23:52.850 ⇒ 00:23:53.550 Amber Lin: Done.
344 00:23:54.090 ⇒ 00:23:56.240 Amber Lin: Or is it just so ongoing?
345 00:23:56.600 ⇒ 00:23:57.089 Samuel Roberts: That is…
346 00:23:57.090 ⇒ 00:23:58.240 Mustafa Raja: What’s a deliverable for this?
347 00:23:59.990 ⇒ 00:24:03.249 Samuel Roberts: I think that’s just the… the Cloud SQL stuff.
348 00:24:03.250 ⇒ 00:24:04.079 Mustafa Raja: Oh, yeah, yeah, yeah, yeah.
349 00:24:04.080 ⇒ 00:24:04.540 Samuel Roberts: King.
350 00:24:04.540 ⇒ 00:24:04.880 Mustafa Raja: Yeah.
351 00:24:04.880 ⇒ 00:24:05.279 Samuel Roberts: Yeah, yeah.
352 00:24:05.280 ⇒ 00:24:11.639 Mustafa Raja: I think, I think we have all this set up, that’s just in, what’s it called? Sqlite, right?
353 00:24:11.950 ⇒ 00:24:12.480 Mustafa Raja: So…
354 00:24:12.480 ⇒ 00:24:13.449 Samuel Roberts: Yeah, I think you’re good on that.
355 00:24:13.450 ⇒ 00:24:18.899 Mustafa Raja: we’re just going to move it over to Postgres. So, market is done, you know.
356 00:24:19.900 ⇒ 00:24:20.690 Amber Lin: Okay.
357 00:24:21.610 ⇒ 00:24:23.650 Mustafa Raja: It’s in the top, yeah, mark as complete, yeah.
358 00:24:24.290 ⇒ 00:24:25.280 Amber Lin: Cool.
359 00:24:25.280 ⇒ 00:24:27.840 Samuel Roberts: Oh, there’s finished and completed? What?
360 00:24:27.840 ⇒ 00:24:30.689 Amber Lin: Yeah, that’s, that’s weird, yeah.
361 00:24:30.930 ⇒ 00:24:32.300 Amber Lin: I see, okay.
362 00:24:32.300 ⇒ 00:24:34.439 Samuel Roberts: This app is something else, I’ll tell you.
363 00:24:35.310 ⇒ 00:24:37.220 Amber Lin: It’s so, so much stuff.
364 00:24:39.170 ⇒ 00:24:40.680 Samuel Roberts: Kick off on Codex.
365 00:24:41.510 ⇒ 00:24:41.950 Mustafa Raja: Oh!
366 00:24:41.950 ⇒ 00:24:43.629 Samuel Roberts: build a better Instagant for us.
367 00:24:43.900 ⇒ 00:24:53.070 Amber Lin: What about this? Like, text to SQL accuracy? Is it started, or is it… like, is it started? Should I move it back? Like, how is this?
368 00:24:53.070 ⇒ 00:24:54.130 Samuel Roberts: I think moving…
369 00:24:55.230 ⇒ 00:24:55.720 Casie Aviles: Yeah.
370 00:24:57.490 ⇒ 00:25:00.180 Samuel Roberts: as critical as some of the other stuff, like…
371 00:25:01.070 ⇒ 00:25:04.850 Amber Lin: the data was not in the database, so you know what I mean? Like, I feel like once we…
372 00:25:05.330 ⇒ 00:25:06.270 Amber Lin: Yeah, okay.
373 00:25:06.270 ⇒ 00:25:09.440 Samuel Roberts: Make sure the zips are right, we’ll make sure that it’s accurate, but it’s hard to do that without…
374 00:25:09.440 ⇒ 00:25:13.629 Amber Lin: So, I will also… I mean, it’s just getting…
375 00:25:13.950 ⇒ 00:25:18.350 Amber Lin: just getting so messy, but I’ll say, like, once we…
376 00:25:20.120 ⇒ 00:25:21.510 Samuel Roberts: Yeah, that’s perfect.
377 00:25:21.510 ⇒ 00:25:22.819 Amber Lin: I’m gonna put it there.
378 00:25:23.660 ⇒ 00:25:26.650 Amber Lin: I… okay.
379 00:25:28.630 ⇒ 00:25:31.679 Amber Lin: There’s so many errors. Errors.
380 00:25:32.290 ⇒ 00:25:36.250 Samuel Roberts: It’s fine. It’s fine. The process, you know? We love it. Gotta love it.
381 00:25:36.890 ⇒ 00:25:43.609 Amber Lin: And so, moving this later, so we don’t have to worry about it this week. This current week, we still have…
382 00:25:44.170 ⇒ 00:25:51.699 Amber Lin: Edge cases set up… Initialized dbt, I’m gonna move that as well.
383 00:25:53.290 ⇒ 00:25:54.130 Mustafa Raja: Boom.
384 00:25:55.060 ⇒ 00:25:57.600 Amber Lin: Are we initializing dbt this week?
385 00:25:57.820 ⇒ 00:25:58.850 Amber Lin: No, right?
386 00:25:58.850 ⇒ 00:25:59.470 Samuel Roberts: What?
387 00:26:00.500 ⇒ 00:26:03.870 Samuel Roberts: What is… the dbt-4 here.
388 00:26:04.150 ⇒ 00:26:05.560 Amber Lin: I don’t know.
389 00:26:07.130 ⇒ 00:26:11.100 Mustafa Raja: Yeah, the, yeah, I don’t see the purpose of that, to be honest, so…
390 00:26:11.470 ⇒ 00:26:12.110 Samuel Roberts: Okay.
391 00:26:12.490 ⇒ 00:26:13.320 Mustafa Raja: It’s there.
392 00:26:13.320 ⇒ 00:26:19.650 Samuel Roberts: Yeah, I wasn’t sure if that was related to the other, like, stuff, but I think… because we needed BQ access for.
393 00:26:19.650 ⇒ 00:26:22.240 Amber Lin: us and for the discovery kind of stuff, right?
394 00:26:22.240 ⇒ 00:26:22.960 Mustafa Raja: Yeah.
395 00:26:23.410 ⇒ 00:26:23.780 Mustafa Raja: Yeah.
396 00:26:23.780 ⇒ 00:26:25.370 Samuel Roberts: I think dbt might be…
397 00:26:25.370 ⇒ 00:26:28.710 Mustafa Raja: Yeah, I just don’t know if we want to model the data, you know?
398 00:26:29.550 ⇒ 00:26:32.300 Samuel Roberts: For our, like, the logs, I don’t know if we knew that yet.
399 00:26:32.300 ⇒ 00:26:32.740 Mustafa Raja: to look down.
400 00:26:32.810 ⇒ 00:26:34.110 Samuel Roberts: So I would say, yeah.
401 00:26:34.110 ⇒ 00:26:39.660 Amber Lin: I’m going to… Delete the dates…
402 00:26:40.430 ⇒ 00:26:42.930 Amber Lin: I just don’t want all the timeline anymore.
403 00:26:43.300 ⇒ 00:26:44.290 Samuel Roberts: Yeah, yeah.
404 00:26:45.380 ⇒ 00:26:46.559 Samuel Roberts: Just get rid of it.
405 00:26:46.560 ⇒ 00:26:48.949 Amber Lin: If we need dbt, we’ll talk about it.
406 00:26:48.950 ⇒ 00:26:51.909 Samuel Roberts: Yeah, I think, I think that was just because, like, yeah, I think that’s.
407 00:26:53.280 ⇒ 00:26:56.559 Samuel Roberts: Master logs to BQ is…
408 00:26:58.310 ⇒ 00:27:01.570 Mustafa Raja: Yeah, that depends on what’s it called, Cloud SQL.
409 00:27:01.570 ⇒ 00:27:02.160 Samuel Roberts: Yeah.
410 00:27:02.550 ⇒ 00:27:03.690 Samuel Roberts: Where is that?
411 00:27:03.880 ⇒ 00:27:12.480 Amber Lin: Yeah, I’m going to move that down here, just to make it more clear. What about rail? Like, should I move it to…
412 00:27:12.480 ⇒ 00:27:15.440 Samuel Roberts: Also further down, yeah, just put that, put that down.
413 00:27:15.690 ⇒ 00:27:20.550 Casie Aviles: That’s gonna depend on… getting mashro logs to BQ.
414 00:27:22.050 ⇒ 00:27:25.940 Samuel Roberts: So, like… 42 goes to…
415 00:27:25.940 ⇒ 00:27:26.690 Amber Lin: this.
416 00:27:28.750 ⇒ 00:27:29.290 Samuel Roberts: You can move it?
417 00:27:29.290 ⇒ 00:27:30.310 Amber Lin: up here.
418 00:27:30.800 ⇒ 00:27:32.290 Amber Lin: Oh my.
419 00:27:32.340 ⇒ 00:27:33.940 Samuel Roberts: Man, I’m telling you, I’m telling you.
420 00:27:34.790 ⇒ 00:27:39.510 Amber Lin: So… What? Okay.
421 00:27:39.510 ⇒ 00:27:40.330 Samuel Roberts: Probably, probably didn’t.
422 00:27:40.330 ⇒ 00:27:40.790 Mustafa Raja: I’m married.
423 00:27:40.790 ⇒ 00:27:41.120 Samuel Roberts: drag.
424 00:27:44.450 ⇒ 00:27:45.750 Samuel Roberts: Yeah, cool, yes.
425 00:27:45.750 ⇒ 00:27:49.359 Amber Lin: Now, hold on, go back down to a thing? Oh, yeah, you’re good, you’re good.
426 00:27:49.970 ⇒ 00:27:51.370 Amber Lin: That’s also done.
427 00:27:51.740 ⇒ 00:27:54.880 Samuel Roberts: And then you can draw… Here to here.
428 00:27:58.220 ⇒ 00:28:00.769 Amber Lin: What are… these are all, like…
429 00:28:02.250 ⇒ 00:28:03.519 Samuel Roberts: Well, this one is now…
430 00:28:03.860 ⇒ 00:28:08.140 Mustafa Raja: I think, 18… 18 we might just want to get rid of.
431 00:28:08.790 ⇒ 00:28:10.960 Samuel Roberts: Yeah, because now we’ve made a whole new project out of that.
432 00:28:10.960 ⇒ 00:28:12.580 Casie Aviles: Oh, yeah. Okay.
433 00:28:13.590 ⇒ 00:28:14.360 Amber Lin: Cool.
434 00:28:16.070 ⇒ 00:28:18.690 Amber Lin: These… Blah blah blah blah. Okay.
435 00:28:18.690 ⇒ 00:28:19.420 Samuel Roberts: Pretty good.
436 00:28:20.260 ⇒ 00:28:22.140 Amber Lin: Alright.
437 00:28:26.420 ⇒ 00:28:27.000 Samuel Roberts: Cool.
438 00:28:28.750 ⇒ 00:28:34.420 Amber Lin: And then, going on, zip code, these are very important, that’s just ongoing.
439 00:28:34.630 ⇒ 00:28:40.649 Amber Lin: That you are doing… And so, like…
440 00:28:41.760 ⇒ 00:28:44.969 Amber Lin: Who has capacity to do these two things?
441 00:28:45.720 ⇒ 00:28:46.769 Samuel Roberts: Which two, sorry?
442 00:28:46.960 ⇒ 00:29:00.489 Amber Lin: like, the contradictory Doofly content creator, and the question categorization weekly report. These two things are… Immediate and urgent.
443 00:29:01.410 ⇒ 00:29:04.269 Mustafa Raja: I think I might have, capacity.
444 00:29:04.740 ⇒ 00:29:08.410 Amber Lin: Okay, and I know Pradnav has a little bit of capacity.
445 00:29:08.410 ⇒ 00:29:10.600 Samuel Roberts: I was gonna ask, I was gonna say, yeah, Pranath, how are you?
446 00:29:12.250 ⇒ 00:29:15.800 Mustafa Raja: Yeah, I think maybe one can… one I can take, and one Punav can take.
447 00:29:16.280 ⇒ 00:29:16.900 Amber Lin: Okay.
448 00:29:17.250 ⇒ 00:29:18.310 Pranav Narahari: Yeah.
449 00:29:18.310 ⇒ 00:29:18.820 Samuel Roberts: maybe…
450 00:29:19.700 ⇒ 00:29:22.139 Mustafa Raja: Let me know what you’re comfortable with.
451 00:29:22.140 ⇒ 00:29:25.899 Pranav Narahari: with Lilo, but, how, like, what do we think…
452 00:29:26.290 ⇒ 00:29:29.520 Pranav Narahari: These are both, like, the low-effort ones, so it shouldn’t take too much time.
453 00:29:29.730 ⇒ 00:29:33.139 Amber Lin: Yeah, I think this one is… yeah, go ahead, sorry.
454 00:29:33.620 ⇒ 00:29:35.289 Samuel Roberts: Oh, no, no, sorry, I forgot what you said.
455 00:29:36.410 ⇒ 00:29:44.879 Amber Lin: Oh, this we just need to download or link it to all the question inputs, and then, like, run the AI query, like this one.
456 00:29:44.880 ⇒ 00:29:46.750 Pranav Narahari: Yeah, formatting is going on.
457 00:29:46.750 ⇒ 00:29:47.410 Amber Lin: Next.
458 00:29:47.410 ⇒ 00:29:50.309 Pranav Narahari: weekly report. Yeah, that makes sense. I can… I can do this one.
459 00:29:50.310 ⇒ 00:29:52.110 Amber Lin: Okay Okay.
460 00:29:53.280 ⇒ 00:29:59.439 Amber Lin: I’ll assign it to you. How many hours do you think that will take? Or Sam, how many hours do you think that will take?
461 00:30:02.750 ⇒ 00:30:09.800 Samuel Roberts: So… what exactly we’re expecting the output? Like, are we building a recurring thing, or just we’re doing an.
462 00:30:09.800 ⇒ 00:30:14.129 Amber Lin: I guess the weekly report can come later, I guess the initial part just.
463 00:30:14.130 ⇒ 00:30:18.299 Samuel Roberts: That’s what I was… yeah, I was gonna say, if this is kind of being split in two, the first one, I would say, is probably…
464 00:30:20.360 ⇒ 00:30:21.160 Amber Lin: 4 hours?
465 00:30:21.160 ⇒ 00:30:24.500 Samuel Roberts: you’re… Yeah, I would say that’s probably reasonable, 4 or 5.
466 00:30:25.780 ⇒ 00:30:27.810 Samuel Roberts: It might be a little… you know.
467 00:30:28.380 ⇒ 00:30:29.270 Amber Lin: Okay.
468 00:30:29.870 ⇒ 00:30:34.159 Amber Lin: And that sounds like it’ll take a little bit… I don’t know how long that would take.
469 00:30:34.280 ⇒ 00:30:39.260 Amber Lin: Is it just an AI query, or is this, like, a big, big task?
470 00:30:39.260 ⇒ 00:30:40.669 Samuel Roberts: Yeah, I feel like it…
471 00:30:41.530 ⇒ 00:30:46.360 Samuel Roberts: it might be a big thing we might want to break down into a couple things, so I would call it maybe 8.
472 00:30:46.360 ⇒ 00:30:47.070 Amber Lin: Okay.
473 00:30:47.220 ⇒ 00:30:48.900 Samuel Roberts: And we can break it down in linear.
474 00:30:49.440 ⇒ 00:30:53.000 Amber Lin: Cool. Alright, that sounds good.
475 00:30:53.000 ⇒ 00:30:55.919 Samuel Roberts: Opus nails it, and we don’t have to worry, but…
476 00:30:55.920 ⇒ 00:30:56.470 Amber Lin: Yeah.
477 00:30:56.470 ⇒ 00:30:58.050 Samuel Roberts: I don’t want to count on that, so… Alright.
478 00:30:58.600 ⇒ 00:31:01.089 Amber Lin: Okay, so next week we have…
479 00:31:01.260 ⇒ 00:31:04.219 Mustafa Raja: Do we have an Opus API key?
480 00:31:08.320 ⇒ 00:31:08.930 Mustafa Raja: API.
481 00:31:08.930 ⇒ 00:31:19.289 Samuel Roberts: Oh, we… might… I mean, we could have it in the Claude account somewhere? I’ll… I’ll look.
482 00:31:19.590 ⇒ 00:31:20.070 Mustafa Raja: Okay, okay.
483 00:31:20.070 ⇒ 00:31:25.639 Samuel Roberts: When you… when you… did you just use Claude online for this, when you put that Notion doc together, Amber?
484 00:31:26.470 ⇒ 00:31:32.070 Mustafa Raja: I ran it in cursor, so I… I downloaded… oh, it’s not here.
485 00:31:32.070 ⇒ 00:31:32.630 Samuel Roberts: Even better.
486 00:31:32.630 ⇒ 00:31:40.800 Amber Lin: I downloaded the central docs, I gave it the transcripts, and then I just asked Kirscher in the query.
487 00:31:41.050 ⇒ 00:31:42.690 Amber Lin: So, it’s…
488 00:31:42.690 ⇒ 00:31:43.889 Samuel Roberts: Yeah, I would say…
489 00:31:45.090 ⇒ 00:31:49.190 Samuel Roberts: Mustafa, you and I can talk a little bit more about, like, how to do this, but I think
490 00:31:49.750 ⇒ 00:31:55.800 Samuel Roberts: I think first pass is literally just chatting with it, with the doc exactly the way Ember did.
491 00:31:55.800 ⇒ 00:31:56.900 Amber Lin: So you can get the…
492 00:31:56.900 ⇒ 00:31:57.290 Samuel Roberts: a trend.
493 00:31:57.290 ⇒ 00:32:05.499 Amber Lin: script, where we talked about it last time, like, this is, this was the plan, the Notion doc.
494 00:32:05.730 ⇒ 00:32:11.030 Amber Lin: And then they gave me this, like, the table that I gave you guys, so these are.
495 00:32:11.030 ⇒ 00:32:14.400 Samuel Roberts: Can you go back to the Markdown file? What extension are you using there for…
496 00:32:14.400 ⇒ 00:32:21.460 Amber Lin: Some… this I won’t recommend, like, half of the time it breaks, and I have to change the extension.
497 00:32:21.460 ⇒ 00:32:23.149 Samuel Roberts: Got it, okay.
498 00:32:23.150 ⇒ 00:32:26.559 Amber Lin: do it, but it is very pretty. I’m using this.
499 00:32:26.560 ⇒ 00:32:31.289 Samuel Roberts: Review and… oh, okay, I ac- you know what, I actually had that and got rid of it, so I didn’t… yes, okay.
500 00:32:31.290 ⇒ 00:32:39.370 Mustafa Raja: I use… I use one Markdown, previewer. That’s pretty good, but it… but we cannot edit Markdown within it.
501 00:32:40.280 ⇒ 00:32:42.700 Mustafa Raja: This one, I think we can… I did the markdown within…
502 00:32:42.700 ⇒ 00:32:50.120 Amber Lin: You can edit, but, like, the… sometimes the file fails, and I have to change it to TXT, and then change it back, so…
503 00:32:50.380 ⇒ 00:32:51.160 Mustafa Raja: happening.
504 00:32:51.160 ⇒ 00:32:55.709 Amber Lin: That’s the issue I have there. Anyways, I think this is good.
505 00:32:55.710 ⇒ 00:32:56.599 Samuel Roberts: Are you stunned, actually?
506 00:32:56.600 ⇒ 00:32:58.669 Amber Lin: like this, one of this.
507 00:32:58.670 ⇒ 00:32:59.489 Samuel Roberts: I saw that we…
508 00:32:59.810 ⇒ 00:33:00.590 Amber Lin: Yeah.
509 00:33:00.650 ⇒ 00:33:05.160 Mustafa Raja: Those are the things I do.
510 00:33:05.550 ⇒ 00:33:06.540 Mustafa Raja: We have…
511 00:33:07.410 ⇒ 00:33:11.379 Casie Aviles: Yeah, sorry, we just have, like, the development up.
512 00:33:11.380 ⇒ 00:33:13.080 Samuel Roberts: development version of…
513 00:33:13.540 ⇒ 00:33:15.810 Amber Lin: So, this is… this is done?
514 00:33:17.710 ⇒ 00:33:20.200 Casie Aviles: And the development and validate.
515 00:33:20.490 ⇒ 00:33:21.140 Samuel Roberts: Yeah.
516 00:33:21.560 ⇒ 00:33:24.949 Casie Aviles: Well, yeah, we do have the development version of Andy already.
517 00:33:25.800 ⇒ 00:33:26.500 Amber Lin: Cool.
518 00:33:26.900 ⇒ 00:33:35.669 Amber Lin: Alright, so I would say, like, staging. Next week, we have natural logs, staging, migrate air models.
519 00:33:36.920 ⇒ 00:33:39.530 Amber Lin: And then wrap up, like, these two.
520 00:33:39.800 ⇒ 00:33:42.740 Amber Lin: These two tasks. I think that’s, like, okay.
521 00:33:42.990 ⇒ 00:33:50.220 Mustafa Raja: I believe, do we, do we have, tasks for migrating the dbtool tables?
522 00:33:50.400 ⇒ 00:33:51.540 Mustafa Raja: From Superbase?
523 00:33:55.880 ⇒ 00:33:58.269 Casie Aviles: I don’t think we have, like, a explicit.
524 00:33:58.270 ⇒ 00:33:59.009 Samuel Roberts: Which one?
525 00:33:59.530 ⇒ 00:34:10.059 Mustafa Raja: Yeah, the DB tool, the tables for DBTool, no? Those are in… what’s it called? Superbase. We might want to transfer those to Cloud SQL when we have that.
526 00:34:13.489 ⇒ 00:34:15.659 Samuel Roberts: Yes, we’re gonna need to migrate those.
527 00:34:16.500 ⇒ 00:34:19.170 Amber Lin: When would that be?
528 00:34:19.170 ⇒ 00:34:22.210 Mustafa Raja: Cloud SQL, somewhere after Cloud SQL.
529 00:34:22.489 ⇒ 00:34:23.790 Samuel Roberts: Yeah, I…
530 00:34:23.790 ⇒ 00:34:26.940 Amber Lin: I’m gonna move it here.
531 00:34:29.239 ⇒ 00:34:39.299 Amber Lin: We’ll move… like, next week, we’ll check if, like, these are what we need to do. Okay, I will need to jump. I think they’re already having a meeting.
532 00:34:40.170 ⇒ 00:34:41.259 Samuel Roberts: Okay, yeah, go for it.
533 00:34:41.679 ⇒ 00:34:43.499 Pranav Narahari: Yeah, I think we’re good.
534 00:34:43.500 ⇒ 00:34:44.479 Amber Lin: Yeah, yeah.
535 00:34:44.480 ⇒ 00:34:51.820 Pranav Narahari: Just, you can probably do it at any point today, just, adding me to the central docs, I just had to request access as a viewer.
536 00:34:51.820 ⇒ 00:34:59.300 Amber Lin: Oh, could you actually try and log in of the Braveforged account? Like, I don’t think…
537 00:34:59.300 ⇒ 00:35:03.910 Samuel Roberts: Yeah, there’s issues with that, because I can’t get in, even, so… Oh.
538 00:35:04.330 ⇒ 00:35:06.079 Mustafa Raja: I don’t know, they’re… there’s…
539 00:35:06.290 ⇒ 00:35:12.489 Amber Lin: red, like, requirements are pretty strict, but I think I can share the past central doc with you, because I own…
540 00:35:12.490 ⇒ 00:35:13.110 Samuel Roberts: Can you…
541 00:35:13.110 ⇒ 00:35:14.080 Amber Lin: their voice.
542 00:35:14.260 ⇒ 00:35:14.890 Amber Lin: will be hard.
543 00:35:14.890 ⇒ 00:35:19.139 Samuel Roberts: Amber, can you export to, like, Markdown or something?
544 00:35:19.540 ⇒ 00:35:19.870 Amber Lin: Yeah.
545 00:35:19.870 ⇒ 00:35:20.380 Samuel Roberts: the central dog.
546 00:35:20.380 ⇒ 00:35:23.589 Amber Lin: Casey, can you help do that, actually? Yeah, yeah, yeah.
547 00:35:23.590 ⇒ 00:35:26.080 Samuel Roberts: Yeah, whoever has access is what you mean, yeah.
548 00:35:26.080 ⇒ 00:35:26.889 Casie Aviles: I can do that.
549 00:35:26.890 ⇒ 00:35:27.650 Samuel Roberts: Okay, yeah.
550 00:35:28.280 ⇒ 00:35:29.829 Samuel Roberts: Cool, that would be, that would be great.
551 00:35:30.170 ⇒ 00:35:34.439 Mustafa Raja: And Pranak, are you in, ABC Vault, in 1Pass?
552 00:35:35.400 ⇒ 00:35:37.890 Pranav Narahari: Let me… let me check.
553 00:35:38.280 ⇒ 00:35:47.769 Amber Lin: If not, ask Rico. I just shared the past one to you both. Okay. Actually, I’ll also, like, I’ll also try and share these ones with you guys.
554 00:35:48.040 ⇒ 00:35:49.289 Amber Lin: If I can.
555 00:35:50.560 ⇒ 00:35:51.010 Samuel Roberts: Sounds good.
556 00:35:51.010 ⇒ 00:35:52.250 Pranav Narahari: I’m not in ABC Vault.
557 00:35:52.250 ⇒ 00:35:55.610 Mustafa Raja: Yeah, so let’s just ask Rico if he can add in.
558 00:35:55.610 ⇒ 00:35:56.200 Pranav Narahari: I’ll do it.
559 00:35:56.510 ⇒ 00:35:57.070 Amber Lin: Okay.
560 00:35:57.070 ⇒ 00:35:57.800 Samuel Roberts: And then.
561 00:35:57.800 ⇒ 00:35:58.420 Amber Lin: Great.
562 00:35:58.700 ⇒ 00:36:03.619 Samuel Roberts: If there are new tickets that need to get added to the Gantt, Amber, can you just let me know when those are, and I’ll flesh them out a little bit?
563 00:36:04.300 ⇒ 00:36:06.230 Amber Lin: Okay, yeah, I’ll write that down.
564 00:36:06.900 ⇒ 00:36:07.400 Samuel Roberts: Thank you.
565 00:36:07.400 ⇒ 00:36:07.800 Amber Lin: Okay.
566 00:36:07.800 ⇒ 00:36:21.369 Mustafa Raja: One question, for Sam. Do you think Brainforge agent is the good one to report, the logs from Mastra in ABC Logs channel?
567 00:36:22.060 ⇒ 00:36:24.039 Samuel Roberts: Yeah, for now that’s fine.
568 00:36:24.040 ⇒ 00:36:25.960 Mustafa Raja: Okay, yeah, just wanted to confirm that.
569 00:36:25.960 ⇒ 00:36:28.850 Samuel Roberts: we will, but I didn’t… yeah, that’s perfect, actually, so, okay.
570 00:36:28.850 ⇒ 00:36:30.190 Mustafa Raja: Okay, thank you.
571 00:36:32.090 ⇒ 00:36:33.710 Amber Lin: Alright, bye everyone.
572 00:36:33.910 ⇒ 00:36:34.369 Pranav Narahari: See you guys.
573 00:36:34.370 ⇒ 00:36:34.970 Mustafa Raja: Thank you.