Meeting Title: AI Team Standup Date: 2025-05-19 Meeting participants: Mustafa Raja, Uttam Kumaran, Miguel De Veyra, Casie Aviles, Awaish Kumar
WEBVTT
1 00:06:23.350 ⇒ 00:06:24.550 Miguel de Veyra: Morning, everyone.
2 00:08:13.520 ⇒ 00:08:15.020 Uttam Kumaran: Hey, guys, can you hear me?
3 00:08:17.070 ⇒ 00:08:17.740 Mustafa Raja: Yes.
4 00:08:18.600 ⇒ 00:08:19.120 Casie Aviles: Yeah.
5 00:08:19.940 ⇒ 00:08:21.019 Uttam Kumaran: Hey! How are you?
6 00:08:21.810 ⇒ 00:08:22.937 Mustafa Raja: Hi! How are you?
7 00:08:23.900 ⇒ 00:08:26.029 Uttam Kumaran: Good. How was the weekend.
8 00:08:26.470 ⇒ 00:08:27.609 Mustafa Raja: Yeah, it was good.
9 00:08:30.230 ⇒ 00:08:31.459 Uttam Kumaran: Do anything, fun.
10 00:08:33.011 ⇒ 00:08:35.119 Mustafa Raja: I I visited my family and all.
11 00:08:35.690 ⇒ 00:08:36.520 Mustafa Raja: Oh, my.
12 00:08:36.520 ⇒ 00:08:37.150 Uttam Kumaran: Oh, nice!
13 00:08:37.150 ⇒ 00:08:41.030 Mustafa Raja: My mom’s sister is visiting from England, so I spend time with them.
14 00:08:42.340 ⇒ 00:08:43.120 Uttam Kumaran: Oh, nice!
15 00:08:43.120 ⇒ 00:08:45.610 Miguel de Veyra: Oh, there you go! That’s why I can’t hear anyone.
16 00:08:46.810 ⇒ 00:08:47.590 Uttam Kumaran: Hey!
17 00:08:47.830 ⇒ 00:08:51.430 Miguel de Veyra: It was on different earphones. Hey, everyone! Morning.
18 00:08:51.870 ⇒ 00:08:52.410 Mustafa Raja: Hey!
19 00:08:52.410 ⇒ 00:08:53.530 Uttam Kumaran: Hey? Good morning.
20 00:08:57.420 ⇒ 00:09:03.959 Uttam Kumaran: Cool. Well, I sent a couple of notes in the chat yesterday, I guess. Would love to get everyone’s feedback.
21 00:09:04.399 ⇒ 00:09:15.560 Uttam Kumaran: Maybe I’ll start we can just dive right in. Maybe, Mustafa. I just wanna hear a little bit about if you were able to go any, go through anything last week, and maybe give me a sense of like
22 00:09:16.290 ⇒ 00:09:19.920 Uttam Kumaran: what you took a look at, and any initial feedback.
23 00:09:20.500 ⇒ 00:09:28.599 Mustafa Raja: Yeah, yeah, it I read the documents in the notion that you mentioned and also
24 00:09:29.235 ⇒ 00:09:33.780 Mustafa Raja: I went into linear and checked out the
25 00:09:34.100 ⇒ 00:09:49.710 Mustafa Raja: checked out the issue for centralization of prompts in Github, for I checked out the repo and I couldn’t find the workflow for it. So I guess we have to start that from the scratch.
26 00:09:52.430 ⇒ 00:09:53.810 Uttam Kumaran: For the prompts.
27 00:09:53.810 ⇒ 00:09:55.090 Mustafa Raja: Yeah, for the prompts.
28 00:09:55.900 ⇒ 00:10:01.100 Uttam Kumaran: So we actually do have one there. Maybe Casey can get you access or.
29 00:10:01.100 ⇒ 00:10:01.710 Mustafa Raja: Yeah.
30 00:10:03.740 ⇒ 00:10:09.670 Casie Aviles: Yeah, we do have a repository where we have, like stored our Bronx.
31 00:10:10.290 ⇒ 00:10:13.917 Mustafa Raja: Yeah, yeah, I saw that report for that. But
32 00:10:14.440 ⇒ 00:10:18.620 Mustafa Raja: the in the linear, there’s a workflow also mentioned there, right?
33 00:10:18.820 ⇒ 00:10:22.240 Mustafa Raja: So I couldn’t find the workflow in an.
34 00:10:22.240 ⇒ 00:10:23.350 Casie Aviles: You mean a workflow.
35 00:10:23.350 ⇒ 00:10:24.110 Mustafa Raja: Yeah, yeah.
36 00:10:24.920 ⇒ 00:10:29.730 Casie Aviles: Okay, yeah, we don’t. We don’t have like a workflow for Github yet.
37 00:10:30.080 ⇒ 00:10:33.809 Mustafa Raja: Okay, yeah, we can. We can get started from scratch.
38 00:10:34.600 ⇒ 00:10:38.250 Mustafa Raja: Okay, what do you think? You know, after taking a look at.
39 00:10:38.390 ⇒ 00:10:40.899 Uttam Kumaran: Some of the workflows like
40 00:10:41.256 ⇒ 00:10:46.229 Uttam Kumaran: do you kind of you kind of see the general theme of like some of the things that we’re trying to do.
41 00:10:48.250 ⇒ 00:10:54.139 Mustafa Raja: Yeah, yeah, I looked at some of them. And I I do. I
42 00:10:57.350 ⇒ 00:11:07.310 Mustafa Raja: I do. Looked at the task that you assigned me for centralization of the common things and for the Github one. So I took a look in that regard.
43 00:11:08.092 ⇒ 00:11:20.759 Mustafa Raja: I I need I need some. I need some assistance in, you know, understanding what’s going on and the current state at what we are, and then I can take off from that.
44 00:11:21.450 ⇒ 00:11:23.760 Uttam Kumaran: Okay, okay, cool. So yeah, I think,
45 00:11:24.650 ⇒ 00:11:35.340 Uttam Kumaran: I guess, for Miguel and Casey like, are there, is there low hanging? N, 8. N, stuff like organization centralization that we can hand to Masafa
46 00:11:36.970 ⇒ 00:11:38.690 Uttam Kumaran: and like, what would some of those be.
47 00:11:39.794 ⇒ 00:11:44.540 Miguel de Veyra: I guess a good sorry. Can everyone hear me?
48 00:11:48.150 ⇒ 00:11:52.889 Miguel de Veyra: wait! Let me think. It’s either the image thing or for slack
49 00:11:53.390 ⇒ 00:11:57.230 Miguel de Veyra: cause it’s just there, or we could try centralize something.
50 00:11:57.290 ⇒ 00:11:59.539 Casie Aviles: Switching them off the imaging works.
51 00:12:00.600 ⇒ 00:12:01.240 Miguel de Veyra: Yeah, yeah.
52 00:12:01.240 ⇒ 00:12:06.900 Uttam Kumaran: Oh, yeah, what is that which? Oh, okay, wait. Ex, maybe. Explain that one further.
53 00:12:07.810 ⇒ 00:12:08.830 Miguel de Veyra: Which one sorry.
54 00:12:09.442 ⇒ 00:12:10.329 Casie Aviles: Yeah, switching them on.
55 00:12:10.330 ⇒ 00:12:10.940 Uttam Kumaran: Models.
56 00:12:11.220 ⇒ 00:12:11.780 Miguel de Veyra: Yeah. Yeah.
57 00:12:11.780 ⇒ 00:12:19.999 Casie Aviles: I mean, yeah. So one of the things we want to do is so we have, like a general purpose, Gpt for all
58 00:12:20.270 ⇒ 00:12:22.619 Casie Aviles: agent on slack and
59 00:12:23.388 ⇒ 00:12:31.470 Casie Aviles: idea. The idea is we want to switch models sometimes. So maybe for all we don’t want to use for all. So maybe we want to use all 3.
60 00:12:31.720 ⇒ 00:12:34.490 Casie Aviles: So maybe there’s like a way we could.
61 00:12:35.332 ⇒ 00:12:39.060 Casie Aviles: You know, easily switch from between models.
62 00:12:39.340 ⇒ 00:12:40.330 Casie Aviles: Basically.
63 00:12:41.740 ⇒ 00:12:45.530 Uttam Kumaran: Hmm, and select which? Yeah.
64 00:12:47.535 ⇒ 00:12:52.570 Mustafa Raja: So what kind of in what kind of interface are we looking into for that?
65 00:12:53.159 ⇒ 00:12:56.830 Mustafa Raja: Are we? Are we? Do we want to do, do it from slack or something.
66 00:12:58.910 ⇒ 00:13:04.320 Casie Aviles: Yeah, that’s 1 option that we want to do, since it’s the bot itself is already on slack.
67 00:13:05.100 ⇒ 00:13:18.750 Mustafa Raja: Okay. So so what I understand is that we have a workflow in and we want the workflow has an AI agent and has a model connected to it, and we want to change model through slack right.
68 00:13:19.815 ⇒ 00:13:20.790 Casie Aviles: Yes, that’s it.
69 00:13:21.000 ⇒ 00:13:24.870 Mustafa Raja: Yeah, yeah, I can. Yeah, I can take over from this.
70 00:13:28.120 ⇒ 00:13:31.890 Uttam Kumaran: Okay, so let’s okay. So this is one.
71 00:13:32.020 ⇒ 00:13:37.849 Uttam Kumaran: Let’s also what about consolidating some of like the
72 00:13:38.130 ⇒ 00:13:41.499 Uttam Kumaran: the logic that’s being reused in in many places.
73 00:13:41.500 ⇒ 00:13:41.939 Miguel de Veyra: Yeah, I bet.
74 00:13:42.390 ⇒ 00:13:45.710 Miguel de Veyra: I was gonna suggest. I think that’s the better thing to work on.
75 00:13:47.970 ⇒ 00:13:48.669 Miguel de Veyra: Because like.
76 00:13:48.670 ⇒ 00:13:53.170 Uttam Kumaran: What are what are some of these? Reuse logic?
77 00:13:54.870 ⇒ 00:13:56.840 Miguel de Veyra: Across the client of agents.
78 00:13:59.550 ⇒ 00:14:00.260 Miguel de Veyra: Yeah.
79 00:14:03.500 ⇒ 00:14:07.790 Miguel de Veyra: the client of agents. We have to separate it. Right, Casey, because it’s triggered.
80 00:14:10.650 ⇒ 00:14:12.090 Miguel de Veyra: Yeah, I think we have this.
81 00:14:12.090 ⇒ 00:14:16.760 Uttam Kumaran: Well, the not. The. The only difference is like the knowledge, and the core prompts right?
82 00:14:16.900 ⇒ 00:14:19.599 Uttam Kumaran: You do have to. You do have to have some.
83 00:14:21.000 ⇒ 00:14:24.260 Uttam Kumaran: I mean, I guess you don’t necessarily need to.
84 00:14:24.420 ⇒ 00:14:27.400 Uttam Kumaran: You could have this whole thing into one workflow.
85 00:14:31.690 ⇒ 00:14:35.101 Uttam Kumaran: But like I mean, let’s take a look at
86 00:14:37.980 ⇒ 00:14:43.933 Uttam Kumaran: One of them, I guess one of my one of my
87 00:14:45.340 ⇒ 00:14:47.840 Uttam Kumaran: points is like, for example.
88 00:14:47.940 ⇒ 00:14:51.240 Uttam Kumaran: this is something that’s used in every single one.
89 00:14:52.540 ⇒ 00:14:59.490 Uttam Kumaran: Right? So how can we like? How can we consolidate this so it can be reused.
90 00:14:59.680 ⇒ 00:15:01.339 Uttam Kumaran: You see what I’m saying.
91 00:15:01.380 ⇒ 00:15:20.610 Mustafa Raja: Yeah, we can. We can transfer this into a new workflow. And here we can use a trigger. Another workflow trigger to trigger that workflow to use these notes and wherever these are being used we can use this logic.
92 00:15:22.370 ⇒ 00:15:27.860 Uttam Kumaran: So then let’s maybe let’s just write down some option, some some of the ones that we want to
93 00:15:29.480 ⇒ 00:15:36.589 Uttam Kumaran: consolidate. So for the. So we wanted to consolidate the slack reaction and loading.
94 00:15:41.660 ⇒ 00:15:45.529 Uttam Kumaran: I mean almost this entire thing. We can consolidate right.
95 00:15:46.960 ⇒ 00:15:47.640 Casie Aviles: Yeah.
96 00:15:54.758 ⇒ 00:15:59.639 Casie Aviles: This one uses like an old. The context approach. So it’s still different from
97 00:16:01.810 ⇒ 00:16:03.879 Casie Aviles: from like, for example, Javi.
98 00:16:07.795 ⇒ 00:16:08.920 Uttam Kumaran: I see. Okay.
99 00:16:31.520 ⇒ 00:16:37.689 Uttam Kumaran: yeah. So I would love to start to consolidate some of these, I think. Let’s take a look at a few others. So.
100 00:16:37.690 ⇒ 00:16:40.209 Miguel de Veyra: The formatter would be a good one to consolidate.
101 00:16:40.870 ⇒ 00:16:42.580 Uttam Kumaran: Yeah, so.
102 00:16:44.200 ⇒ 00:16:46.960 Casie Aviles: This one’s a huge workflow.
103 00:16:53.020 ⇒ 00:17:01.220 Uttam Kumaran: So there is this. Yeah, where there were is the slack formatter in here as well.
104 00:17:02.600 ⇒ 00:17:09.170 Casie Aviles: There’s a series of prompt Llm chain. Basic chain steps before that. That.
105 00:17:09.569 ⇒ 00:17:10.479 Miguel de Veyra: Formats.
106 00:17:11.299 ⇒ 00:17:12.419 Miguel de Veyra: You saw it! Tuta.
107 00:17:13.010 ⇒ 00:17:15.310 Casie Aviles: Right on the right side.
108 00:17:15.560 ⇒ 00:17:17.447 Miguel de Veyra: Right there you go!
109 00:17:18.319 ⇒ 00:17:19.489 Uttam Kumaran: Oh, okay. Okay.
110 00:17:19.490 ⇒ 00:17:20.109 Miguel de Veyra: Yeah.
111 00:17:21.069 ⇒ 00:17:25.370 Miguel de Veyra: And then I think even more, you can see the slack format, or there.
112 00:17:28.339 ⇒ 00:17:29.829 Uttam Kumaran: Where is the.
113 00:17:33.969 ⇒ 00:17:36.719 Uttam Kumaran: I guess? Is there? Is it just in this.
114 00:17:37.830 ⇒ 00:17:40.160 Miguel de Veyra: I think I saw it earlier. Can you zoom out a bit? Sorry.
115 00:17:41.170 ⇒ 00:17:43.719 Miguel de Veyra: I think it’s after this huge slack thing.
116 00:17:44.930 ⇒ 00:17:46.590 Miguel de Veyra: There you go! There, there, there!
117 00:17:46.730 ⇒ 00:17:51.020 Miguel de Veyra: After that slack, short summarizer.
118 00:17:52.100 ⇒ 00:17:53.630 Uttam Kumaran: I think it’s just this.
119 00:17:53.630 ⇒ 00:17:55.140 Miguel de Veyra: Yeah, I think it should be that one
120 00:17:58.545 ⇒ 00:18:06.529 Miguel de Veyra: but a good example. So we don’t have a hard time finding it. I think it’s 1. i think it’s in one of data without double check.
121 00:18:08.090 ⇒ 00:18:09.340 Casie Aviles: I think Javi also has.
122 00:18:09.340 ⇒ 00:18:10.800 Miguel de Veyra: Yeah, I think, yeah, we have some.
123 00:18:11.780 ⇒ 00:18:12.390 Uttam Kumaran: Okay.
124 00:18:12.920 ⇒ 00:18:14.300 Casie Aviles: Mine’s a little messy.
125 00:18:14.300 ⇒ 00:18:15.619 Miguel de Veyra: And the client Hub. Sorry.
126 00:18:16.120 ⇒ 00:18:16.690 Casie Aviles: Yeah.
127 00:18:24.990 ⇒ 00:18:27.330 Casie Aviles: yeah. There’s 1 node, I think, for formatting.
128 00:18:27.330 ⇒ 00:18:28.290 Miguel de Veyra: Yeah, there you go.
129 00:18:29.800 ⇒ 00:18:32.690 Miguel de Veyra: It’s it’s the note before sending it to slack.
130 00:18:33.170 ⇒ 00:18:34.909 Casie Aviles: Yes, mark down, clean.
131 00:18:34.910 ⇒ 00:18:35.259 Uttam Kumaran: Oh! This one!
132 00:18:35.260 ⇒ 00:18:37.070 Miguel de Veyra: Yeah. Markdown cleaner. Yep, that one.
133 00:18:41.960 ⇒ 00:18:42.610 Casie Aviles: Hello!
134 00:19:01.690 ⇒ 00:19:09.810 Uttam Kumaran: Okay, so I want to definitely see us like, Con, like, basically consolidate a lot of these pieces of logic and actually start to build
135 00:19:10.300 ⇒ 00:19:14.706 Uttam Kumaran: like systems right? Cause I don’t want each of us. I don’t want to reinvent
136 00:19:15.340 ⇒ 00:19:19.689 Uttam Kumaran: how we’re doing slack, formatting all power writing to different places. So
137 00:19:20.750 ⇒ 00:19:37.469 Uttam Kumaran: I think, Mustafa, it’d be very helpful. This is just like a short list, but I would love to consolidate as much as as we can. So do you. Do you mind like? I guess, Miguel and Casey, can we? Can you add other pieces to this that we want to consolidate? And then.
138 00:19:38.390 ⇒ 00:19:38.770 Miguel de Veyra: Sure.
139 00:19:38.770 ⇒ 00:19:45.949 Uttam Kumaran: Must, Mustafa as well as as you go through these? Can you? Can you nominate certain things that for consolidation.
140 00:19:46.370 ⇒ 00:19:47.340 Mustafa Raja: Yeah, sure.
141 00:19:48.470 ⇒ 00:19:59.919 Uttam Kumaran: Okay, exactly the way you described, though, like, we just call an external workflow. But some of these again, like we use so many places, and I don’t. It’s really painful to go edit this every time. So
142 00:20:02.510 ⇒ 00:20:05.140 Uttam Kumaran: cool anything else in.
143 00:20:06.580 ⇒ 00:20:12.230 Uttam Kumaran: And I then like sort of clean up, because I think some of these, like simple ones, will also help you.
144 00:20:12.750 ⇒ 00:20:15.680 Uttam Kumaran: you know. Learn about each of the core workflows.
145 00:20:18.140 ⇒ 00:20:18.750 Mustafa Raja: Yeah.
146 00:20:25.470 ⇒ 00:20:28.860 Uttam Kumaran: Anything else like Casey Miguel. We want to add to this list.
147 00:20:34.200 ⇒ 00:20:34.740 Casie Aviles: I don’t.
148 00:20:35.035 ⇒ 00:20:36.809 Casie Aviles: I don’t have any right now.
149 00:20:36.810 ⇒ 00:20:44.699 Miguel de Veyra: Yeah, right now. No cause I’m checking the github and slack and zoom, they’re pretty straightforward, so I don’t think it’s worth, you know.
150 00:20:44.700 ⇒ 00:20:49.059 Casie Aviles: I guess, for a cleanup we could do like folders for each.
151 00:20:51.090 ⇒ 00:20:54.579 Miguel de Veyra: Yeah, yeah. That one would be the good thing to do.
152 00:20:58.560 ⇒ 00:20:59.240 Uttam Kumaran: Okay?
153 00:21:04.870 ⇒ 00:21:12.160 Uttam Kumaran: And then the only other one I would ask is like, Create an O 3 slack agent.
154 00:21:15.790 ⇒ 00:21:16.660 Miguel de Veyra: Yeah.
155 00:21:17.270 ⇒ 00:21:22.810 Uttam Kumaran: Oh, well, well, this is actually gonna be what we mentioned, which is like being able to select models. So I’m sure.
156 00:21:22.810 ⇒ 00:21:23.180 Miguel de Veyra: Yeah.
157 00:21:26.410 ⇒ 00:21:29.310 Miguel de Veyra: Chat model familiar.
158 00:21:37.010 ⇒ 00:21:39.530 Miguel de Veyra: do we have access to? O 3 already?
159 00:21:41.384 ⇒ 00:21:43.980 Casie Aviles: We don’t. We haven’t deployed one on azure.
160 00:21:43.980 ⇒ 00:21:50.800 Miguel de Veyra: Yeah. And okay, okay, yeah, we don’t have it on openai. But yeah, maybe you’re sure.
161 00:22:18.710 ⇒ 00:22:19.500 Uttam Kumaran: Okay,
162 00:22:25.490 ⇒ 00:22:30.719 Uttam Kumaran: great. So then tell me where we are with some. Let’s go through some of these existing tickets and
163 00:22:31.280 ⇒ 00:22:33.910 Uttam Kumaran: tell me where we are with some of these.
164 00:22:39.710 ⇒ 00:22:42.449 Uttam Kumaran: So this rag pre retrieval, we’re happy with.
165 00:22:43.247 ⇒ 00:22:44.500 Miguel de Veyra: Yes, so far.
166 00:22:48.540 ⇒ 00:22:54.429 Uttam Kumaran: Okay, so let’s let’s talk about all of these. So which ones can I move to? Done like, where are where are these ones?
167 00:22:56.942 ⇒ 00:23:04.010 Miguel de Veyra: I think we start with client review the ones in there assure to S. 3.
168 00:23:06.970 ⇒ 00:23:08.839 Uttam Kumaran: Oh, wait! This is the usage reports.
169 00:23:14.340 ⇒ 00:23:14.900 Awaish Kumar: Yes.
170 00:23:15.860 ⇒ 00:23:17.250 Uttam Kumaran: Okay, you got it out.
171 00:23:18.113 ⇒ 00:23:25.659 Awaish Kumar: No, actually, I’ve been, I I work, I and Miguel actually try to find out. But the I think it says the
172 00:23:26.350 ⇒ 00:23:32.299 Awaish Kumar: like. It does not provide the granularity like the call level the cost.
173 00:23:32.640 ⇒ 00:23:34.490 Awaish Kumar: It’s because of subscriptions.
174 00:23:34.610 ⇒ 00:23:35.760 Awaish Kumar: I’m not sure.
175 00:23:38.140 ⇒ 00:23:39.649 Uttam Kumaran: Because we’re getting it for free.
176 00:23:39.870 ⇒ 00:23:41.360 Awaish Kumar: Yes, so it?
177 00:23:41.360 ⇒ 00:23:42.020 Awaish Kumar: Oh, yeah. Okay.
178 00:23:42.020 ⇒ 00:23:45.009 Awaish Kumar: So for each call, like, what is the
179 00:23:45.842 ⇒ 00:23:50.139 Awaish Kumar: how many calls are made? How much are successful? But it doesn’t show the cost because it
180 00:23:50.140 ⇒ 00:23:51.889 Awaish Kumar: do. We have the do. We have the tokens?
181 00:23:53.970 ⇒ 00:23:58.138 Awaish Kumar: Yeah, it’s it does show the tokens data.
182 00:23:58.660 ⇒ 00:24:03.580 Uttam Kumaran: Okay, that’s fine. Then as long as we have the tokens, then we can get the estimated cost. Basically.
183 00:24:03.990 ⇒ 00:24:04.900 Awaish Kumar: Okay.
184 00:24:06.670 ⇒ 00:24:11.330 Awaish Kumar: Okay. So then, is this already there in S. 3?
185 00:24:12.250 ⇒ 00:24:14.269 Awaish Kumar: Yeah, no, it’s it’s not yet.
186 00:24:14.950 ⇒ 00:24:15.580 Uttam Kumaran: Okay.
187 00:24:15.690 ⇒ 00:24:19.069 Awaish Kumar: I have been working on the this one. The notion, Doc.
188 00:24:19.630 ⇒ 00:24:27.360 Uttam Kumaran: Okay, so I’m gonna move this one back here. And then, yeah, how is this one.
189 00:24:27.660 ⇒ 00:24:32.830 Awaish Kumar: Yeah. So it’s kind of this. I have wrote the script, but I’ve been. I’ve been asking for Snowflake.
190 00:24:34.030 ⇒ 00:24:36.169 Uttam Kumaran: Oh, that’s right. That’s right. That’s right. Okay.
191 00:24:36.495 ⇒ 00:24:40.399 Awaish Kumar: So I can move sheets data there. So we we have to.
192 00:24:40.400 ⇒ 00:24:42.430 Uttam Kumaran: On zoom, yeah.
193 00:24:42.925 ⇒ 00:24:43.250 Awaish Kumar: Okay.
194 00:24:48.980 ⇒ 00:24:49.850 Uttam Kumaran: Yeah. Go ahead.
195 00:24:50.266 ⇒ 00:24:54.433 Awaish Kumar: Yeah, we have 2 things like mission docs and the
196 00:24:55.110 ⇒ 00:25:03.440 Awaish Kumar: the sheets data. So I think the notion dog goes to S 3 is that.
197 00:25:03.440 ⇒ 00:25:04.150 Uttam Kumaran: Yes.
198 00:25:04.150 ⇒ 00:25:10.330 Awaish Kumar: From it. But the sheets I don’t know like which which is better. Should I just like a story.
199 00:25:10.330 ⇒ 00:25:12.659 Uttam Kumaran: But what what sheets are we talking about?
200 00:25:12.850 ⇒ 00:25:17.350 Awaish Kumar: The-the-box sheet, like the documentation sheet.
201 00:25:17.350 ⇒ 00:25:18.250 Uttam Kumaran: Oh!
202 00:25:19.610 ⇒ 00:25:20.050 Awaish Kumar: Yeah.
203 00:25:21.600 ⇒ 00:25:23.729 Uttam Kumaran: Just store that as a Csv right?
204 00:25:24.550 ⇒ 00:25:31.669 Awaish Kumar: Yeah, yeah, we can store them as a Csv in S. 3. But that’s what I’m use. Use. You are saying that
205 00:25:32.530 ⇒ 00:25:39.489 Awaish Kumar: good one like when sometime we should spare store them in Snowflake. That’s why
206 00:25:39.600 ⇒ 00:25:44.959 Awaish Kumar: I was thinking, if like, if we want to store it snowflake like, please, why not now.
207 00:25:46.070 ⇒ 00:25:55.349 Uttam Kumaran: Oh, I guess my point is, make sure it all gets an s. 3 first, st and then we can bring it into Snowflake after that way. There’s it’s in a data lake. First.st
208 00:25:55.460 ⇒ 00:26:06.550 Awaish Kumar: Otherwise, like we’re gonna be bound by snowflake. So I would just make sure it’s landing somewhere, because then you could, you could use snowpipe, or like we could use dlt or something to move it for free.
209 00:26:07.570 ⇒ 00:26:12.599 Awaish Kumar: Okay, okay. Then, I just keep it in S. 3.
210 00:26:16.470 ⇒ 00:26:20.710 Uttam Kumaran: Okay, is that track? Is that tracked? Or should I create another ticket here.
211 00:26:23.379 ⇒ 00:26:24.639 Awaish Kumar: Sorry!
212 00:26:25.060 ⇒ 00:26:27.109 Uttam Kumaran: Should I just create another ticket for that.
213 00:26:28.750 ⇒ 00:26:30.449 Awaish Kumar: Okay, there’s no ticket. I just
214 00:26:30.560 ⇒ 00:26:34.254 Awaish Kumar: implicitly in the the same ticket. I I was doing that.
215 00:26:34.590 ⇒ 00:26:35.549 Uttam Kumaran: Okay? No. Problem.
216 00:26:42.740 ⇒ 00:26:43.415 Awaish Kumar: Okay.
217 00:26:44.546 ⇒ 00:26:47.283 Uttam Kumaran: Great. So then let’s talk about
218 00:26:48.710 ⇒ 00:26:49.729 Uttam Kumaran: This one Miguel.
219 00:26:52.821 ⇒ 00:26:55.069 Miguel de Veyra: Yeah, I think this one we’ve already.
220 00:26:55.624 ⇒ 00:26:57.360 Miguel de Veyra: Yeah, we always provided.
221 00:26:57.360 ⇒ 00:27:02.819 Uttam Kumaran: Connect with. Did you? Did you connect with Kyle about the tables that he had a bunch of questions about tables.
222 00:27:03.110 ⇒ 00:27:07.629 Miguel de Veyra: I believe we did. Yes, 2 weeks ago. Wait.
223 00:27:07.750 ⇒ 00:27:18.440 Casie Aviles: I replied to Kyle. He asked about tables, and I just told him that it should be doable as long as we can modify the Python script to ingest tables.
224 00:27:31.417 ⇒ 00:27:33.560 Uttam Kumaran: Can you make sure that gets into here.
225 00:27:34.692 ⇒ 00:27:35.480 Casie Aviles: Yes, yes, this.
226 00:27:35.480 ⇒ 00:27:39.596 Uttam Kumaran: Because right now it says you can’t do tape. It says, don’t do tables.
227 00:27:40.610 ⇒ 00:27:42.529 Casie Aviles: Yeah, this was before the tables.
228 00:27:42.780 ⇒ 00:27:44.630 Casie Aviles: I’m going to update that.
229 00:27:46.580 ⇒ 00:27:49.499 Miguel de Veyra: Because I think the script I provided only got the blocks.
230 00:27:50.750 ⇒ 00:27:51.410 Uttam Kumaran: Okay.
231 00:28:02.300 ⇒ 00:28:03.020 Uttam Kumaran: right?
232 00:28:10.020 ⇒ 00:28:15.590 Uttam Kumaran: Okay, let’s talk about this one. So linear tickets. S, 3. That’s good.
233 00:28:15.800 ⇒ 00:28:17.519 Awaish Kumar: That’s done, actually.
234 00:28:18.200 ⇒ 00:28:18.596 Uttam Kumaran: And
235 00:28:20.090 ⇒ 00:28:24.930 Uttam Kumaran: This is a come. This is every linear ticket across the entire app.
236 00:28:25.520 ⇒ 00:28:30.709 Awaish Kumar: So it’s basically just we are using this or atomic right?
237 00:28:31.030 ⇒ 00:28:36.899 Awaish Kumar: It just moves everything there. But now, after that, we might need some transformation
238 00:28:37.750 ⇒ 00:28:38.290 Uttam Kumaran: Okay.
239 00:28:38.290 ⇒ 00:28:41.150 Awaish Kumar: Based on how, how ait wants to use it.
240 00:28:41.680 ⇒ 00:28:42.700 Uttam Kumaran: Okay, okay?
241 00:28:43.140 ⇒ 00:28:46.804 Uttam Kumaran: So Miguel did, have you guys looked, have you guys looked at?
242 00:28:48.090 ⇒ 00:28:49.599 Uttam Kumaran: Have you guys looked at this.
243 00:28:51.648 ⇒ 00:28:54.440 Miguel de Veyra: No, no, we haven’t worked on any.
244 00:28:54.440 ⇒ 00:28:59.119 Uttam Kumaran: Well, Ca, can you just can you just go look and verify the format.
245 00:28:59.260 ⇒ 00:29:00.150 Miguel de Veyra: Okay. Sure.
246 00:29:01.510 ⇒ 00:29:05.300 Uttam Kumaran: I’m not gonna close this out until we do that, because then we’re gonna have to come back.
247 00:29:07.250 ⇒ 00:29:09.040 Uttam Kumaran: So, Miguel, do you want to take this.
248 00:29:09.660 ⇒ 00:29:10.729 Miguel de Veyra: Oh, yeah. Yeah. Sure.
249 00:29:22.710 ⇒ 00:29:24.499 Uttam Kumaran: Cool. How about this one.
250 00:29:27.460 ⇒ 00:29:28.220 Miguel de Veyra: Yes, I think.
251 00:29:28.220 ⇒ 00:29:31.119 Casie Aviles: This is just for initialization.
252 00:29:31.360 ⇒ 00:29:34.139 Miguel de Veyra: Yeah, yeah, this was, this should be done because it’s already there.
253 00:29:45.690 ⇒ 00:29:50.729 Uttam Kumaran: This one. The only thing is that the threading right is still a challenge.
254 00:29:52.683 ⇒ 00:30:02.500 Casie Aviles: Yeah, I mean, the we did have like a an implementation for it. But the thing is, the bot will reply, regardless right? I think you’ve tested.
255 00:30:04.040 ⇒ 00:30:04.360 Awaish Kumar: Okay.
256 00:30:04.360 ⇒ 00:30:05.750 Uttam Kumaran: So we’re blocked there.
257 00:30:07.820 ⇒ 00:30:08.470 Casie Aviles: Yeah.
258 00:30:09.590 ⇒ 00:30:10.320 Awaish Kumar: Okay.
259 00:30:10.490 ⇒ 00:30:17.472 Uttam Kumaran: Can we? Can you meet with Mustafa and and ask and and see if he has any ideas for this?
260 00:30:18.820 ⇒ 00:30:20.290 Uttam Kumaran: Because I I.
261 00:30:20.510 ⇒ 00:30:21.090 Awaish Kumar: I don’t.
262 00:30:21.090 ⇒ 00:30:22.019 Uttam Kumaran: I guess it’s
263 00:30:22.940 ⇒ 00:30:27.949 Uttam Kumaran: I guess it’s fine, for now that the bot responds, but it is like not the expected.
264 00:30:28.140 ⇒ 00:30:28.780 Awaish Kumar: Workflow.
265 00:30:28.780 ⇒ 00:30:36.810 Uttam Kumaran: So like I I want us to either meet. I want to ha hand this to Mustafa, or I want to talk to some preta and ask her.
266 00:30:41.740 ⇒ 00:30:44.460 Uttam Kumaran: can you guys grab time to meet today.
267 00:30:47.220 ⇒ 00:30:50.429 Uttam Kumaran: Okay. Thank you.
268 00:30:51.500 ⇒ 00:30:57.470 Miguel de Veyra: I have a question. Sorry is it linear tickets or linear? What’s the difference.
269 00:30:57.470 ⇒ 00:31:04.250 Awaish Kumar: Linear. Yeah, I was stored in linear tickets. I’m not sure how linear is. It was already there. But.
270 00:31:04.250 ⇒ 00:31:06.629 Miguel de Veyra: Okay? Okay, sure. Sure. Okay, thank you.
271 00:31:08.610 ⇒ 00:31:10.919 Miguel de Veyra: Linear ticket system.
272 00:31:13.330 ⇒ 00:31:14.590 Awaish Kumar: Oh, okay. Here.
273 00:31:14.590 ⇒ 00:31:15.210 Awaish Kumar: Yes.
274 00:31:19.660 ⇒ 00:31:26.709 Uttam Kumaran: Yeah, you can update the ticket with what you find. And then so rag, slack, free retrieval. So how do we feel about this one.
275 00:31:26.960 ⇒ 00:31:33.759 Uttam Kumaran: do we? Do we want Mustafa? Have you done any like rag like pre retrieval, pre filtering stuff like that?
276 00:31:35.140 ⇒ 00:31:37.590 Mustafa Raja: I haven’t, but I but I can look into this.
277 00:31:38.920 ⇒ 00:31:49.709 Uttam Kumaran: Okay, I kinda it would be great to get your review, Miguel, can you? Can you put in that? Can you send that article here and put that in here, and we can have Mustafa review this.
278 00:31:49.710 ⇒ 00:31:51.030 Miguel de Veyra: Okay. Yeah. Sure.
279 00:32:01.520 ⇒ 00:32:02.440 Uttam Kumaran: right?
280 00:32:06.080 ⇒ 00:32:09.559 Uttam Kumaran: Okay, how do we feel about this one adding, Github.
281 00:32:12.960 ⇒ 00:32:20.929 Miguel de Veyra: I think this should be. Wait cool part. Ciavi urban Eden. Yeah, I think we migrated this last week.
282 00:32:24.150 ⇒ 00:32:26.500 Miguel de Veyra: Wait, let me check pool parts. I’m not sure.
283 00:32:27.740 ⇒ 00:32:31.690 Miguel de Veyra: Wait, let me double check super.
284 00:32:32.250 ⇒ 00:32:40.130 Miguel de Veyra: because there, there was an extra step that we we needed to do here with them. Basically, we had to bring it into storage
285 00:32:41.900 ⇒ 00:32:43.490 Miguel de Veyra: and supervise.
286 00:32:50.150 ⇒ 00:32:57.190 Casie Aviles: Like initially, we were getting from Github directly, right? Or for these. But
287 00:32:57.360 ⇒ 00:33:03.860 Casie Aviles: there there was like the repository was not in the brain forge org anymore, right.
288 00:33:04.210 ⇒ 00:33:19.310 Miguel de Veyra: Yeah. So we had a ways get it from or for wherever the source are, and then S into S. 3, and then we moved it to S. 3 into super base. But yeah, the only thing we need there is pool parts. We still don’t have it in super base.
289 00:33:19.600 ⇒ 00:33:20.889 Miguel de Veyra: The rest are there.
290 00:33:28.220 ⇒ 00:33:38.929 Uttam Kumaran: So, okay, what? What’s what else is this that can that happen today?
291 00:33:39.651 ⇒ 00:33:44.620 Miguel de Veyra: Yeah, we just, I think, a wish. We just need to find where the pool parts code is right.
292 00:33:45.240 ⇒ 00:33:49.980 Awaish Kumar: Yeah, like we. I can just add it in config to bring in full part.
293 00:33:50.425 ⇒ 00:33:53.100 Miguel de Veyra: Okay, same storage and supervisor.
294 00:33:53.100 ⇒ 00:33:54.140 Awaish Kumar: Okay.
295 00:33:54.140 ⇒ 00:33:55.309 Miguel de Veyra: Okay. Thanks. Matt.
296 00:34:02.880 ⇒ 00:34:09.930 Uttam Kumaran: Okay, great. Let’s talk about these 3 are fine.
297 00:34:11.639 ⇒ 00:34:17.729 Uttam Kumaran: This one is gonna be a lower priority swim is higher priority.
298 00:34:18.952 ⇒ 00:34:24.180 Uttam Kumaran: Let’s talk about this zoom to s 3 backfill.
299 00:34:28.423 ⇒ 00:34:28.960 Casie Aviles: Kate.
300 00:34:28.969 ⇒ 00:34:33.519 Uttam Kumaran: Do, do you? Wanna yeah, tell me about this? Or if you want to hand this to Weish, yeah.
301 00:34:34.760 ⇒ 00:34:41.870 Casie Aviles: Yeah. So for this. I’ve only been able to backfill. For like the meeting topics and dates. And
302 00:34:42.150 ⇒ 00:34:50.990 Casie Aviles: yeah, the problem I am encountering is that we have, like the old videos on or like the old recording files on S. 3.
303 00:34:51.219 ⇒ 00:34:55.560 Casie Aviles: But we don’t have like the the summaries, for from those.
304 00:34:56.150 ⇒ 00:35:02.089 Casie Aviles: So that’s where I guess I would like oasis help in figuring that out.
305 00:35:02.580 ⇒ 00:35:08.459 Casie Aviles: Because, you know, for a certain point, we started getting the summaries and the
306 00:35:08.770 ⇒ 00:35:12.800 Casie Aviles: vector embeddings of the meetings. But before that we didn’t.
307 00:35:13.020 ⇒ 00:35:21.939 Casie Aviles: So basically, we have to go through the old videos and get their and get their summaries
308 00:35:22.100 ⇒ 00:35:25.050 Casie Aviles: from the transcripts, and then also metadata
309 00:35:25.560 ⇒ 00:35:27.960 Casie Aviles: and have it and then vectorize those.
310 00:35:31.390 ⇒ 00:35:34.459 Miguel de Veyra: This could be a good use case for the Pre.
311 00:35:35.000 ⇒ 00:35:36.600 Miguel de Veyra: The pre retrieval.
312 00:35:38.430 ⇒ 00:35:39.870 Uttam Kumaran: Oh, 100%.
313 00:35:40.790 ⇒ 00:35:47.859 Miguel de Veyra: Although that pre retrieval step is a bit tailored to slack, of course, but I think I can move that definitely here.
314 00:35:49.570 ⇒ 00:35:51.199 Uttam Kumaran: What do you have that article? By the way.
315 00:35:51.390 ⇒ 00:35:53.273 Miguel de Veyra: Wait! Let me find it.
316 00:35:53.830 ⇒ 00:35:55.370 Uttam Kumaran: Sorry. I keep forgetting.
317 00:35:55.830 ⇒ 00:35:56.320 Miguel de Veyra: Okay.
318 00:35:56.320 ⇒ 00:35:59.459 Miguel de Veyra: I think I saved it somewhere. Give me a moment.
319 00:35:59.460 ⇒ 00:36:00.810 Uttam Kumaran: Oh, it’s this one, right.
320 00:36:01.530 ⇒ 00:36:02.389 Miguel de Veyra: You found it?
321 00:36:03.885 ⇒ 00:36:04.410 Miguel de Veyra: No.
322 00:36:12.750 ⇒ 00:36:14.640 Casie Aviles: There another article, the second one.
323 00:36:16.360 ⇒ 00:36:16.839 Uttam Kumaran: This one right.
324 00:36:20.820 ⇒ 00:36:22.699 Miguel de Veyra: I think, yeah, yeah, yeah.
325 00:36:22.990 ⇒ 00:36:25.200 Miguel de Veyra: Wait. Oh, mine’s blocked. That’s why.
326 00:36:25.850 ⇒ 00:36:26.700 Miguel de Veyra: Okay.
327 00:36:32.080 ⇒ 00:36:35.580 Uttam Kumaran: Okay, it would. Honestly, it would be great for everybody here to read this
328 00:36:35.860 ⇒ 00:36:40.220 Uttam Kumaran: and understand like what these pieces are, especially Mustafa. I think
329 00:36:40.380 ⇒ 00:36:44.619 Uttam Kumaran: we’re sort of leveling up as a team to go do pretty serious rag work.
330 00:36:44.790 ⇒ 00:36:48.070 Uttam Kumaran: And I think this would be very helpful to look at.
331 00:36:48.990 ⇒ 00:36:50.420 Mustafa Raja: Yeah, I’ll look into this.
332 00:36:51.160 ⇒ 00:36:57.699 Uttam Kumaran: Okay, cool. So I wanna break this down as I want this to be
333 00:36:58.160 ⇒ 00:37:04.189 Uttam Kumaran: just a backfill, and then and then is the pre retrieval. Step after this right.
334 00:37:08.670 ⇒ 00:37:11.450 Miguel de Veyra: Yep, we have.
335 00:37:11.450 ⇒ 00:37:14.190 Miguel de Veyra: You have to. You kind of have to also clean the data.
336 00:37:16.130 ⇒ 00:37:25.969 Uttam Kumaran: Yeah, so this would be Zoom Meeting point and free retrieval.
337 00:37:38.920 ⇒ 00:37:45.050 Uttam Kumaran: Okay, I want. I want Casey to take this. But can you give him the the lowdown on how to do this, and then.
338 00:37:45.050 ⇒ 00:37:45.989 Miguel de Veyra: Oh, yeah, yeah, definitely.
339 00:37:46.520 ⇒ 00:37:48.270 Uttam Kumaran: Mustafa can take this.
340 00:37:48.270 ⇒ 00:37:49.230 Miguel de Veyra: Okay, sure.
341 00:37:50.690 ⇒ 00:37:51.680 Miguel de Veyra: And then.
342 00:37:51.680 ⇒ 00:37:52.500 Uttam Kumaran: Under this law.
343 00:37:52.500 ⇒ 00:38:09.859 Miguel de Veyra: Yeah, can we add some ticket? I mean later, about cause there’s that’s just I sent the correct link. By the way, it’s a 3 step process right? Pre retrieval is done. We have that down. Retrieval is one, and then post retrieval. I want to work on post retrieval ideally sometime this week.
344 00:38:10.420 ⇒ 00:38:11.509 Uttam Kumaran: This is the right one.
345 00:38:12.930 ⇒ 00:38:16.489 Miguel de Veyra: I don’t know the the right ticket in the AI team Channel.
346 00:38:18.240 ⇒ 00:38:19.040 Uttam Kumaran: Okay. Okay.
347 00:38:19.330 ⇒ 00:38:20.430 Miguel de Veyra: Yeah. It’s that one.
348 00:38:21.200 ⇒ 00:38:26.843 Uttam Kumaran: Cool, so I guess
349 00:38:28.450 ⇒ 00:38:33.890 Uttam Kumaran: So let’s talk about that. So you want to work on it. But for the for the slack, right.
350 00:38:35.064 ⇒ 00:38:36.759 Miguel de Veyra: No, I don’t.
351 00:38:36.760 ⇒ 00:38:37.210 Uttam Kumaran: Because I.
352 00:38:37.210 ⇒ 00:38:40.900 Miguel de Veyra: There is as much already. Pre retrieval is the most complicated one. I’d say.
353 00:38:41.430 ⇒ 00:38:41.960 Uttam Kumaran: I guess
354 00:38:41.960 ⇒ 00:38:46.930 Uttam Kumaran: what I want to. What I what I want to consider is that we just do it source by source.
355 00:38:47.310 ⇒ 00:38:48.170 Uttam Kumaran: right?
356 00:38:48.170 ⇒ 00:38:50.840 Miguel de Veyra: Yes, yes, tailored, basically.
357 00:38:58.110 ⇒ 00:38:59.859 Uttam Kumaran: So you’re gonna work on this.
358 00:39:00.230 ⇒ 00:39:04.710 Uttam Kumaran: You’re gonna are you working? You’re not working on retrieval. Right? You’re working on.
359 00:39:06.500 ⇒ 00:39:07.010 Miguel de Veyra: I did.
360 00:39:07.010 ⇒ 00:39:08.130 Uttam Kumaran: Post, retrieval.
361 00:39:09.130 ⇒ 00:39:15.220 Miguel de Veyra: Oh, no, no, I wanna do it from top down. So I was working on retrieval last week a bit. But then.
362 00:39:15.390 ⇒ 00:39:16.180 Miguel de Veyra: yeah.
363 00:39:16.410 ⇒ 00:39:17.000 Uttam Kumaran: Okay.
364 00:39:17.920 ⇒ 00:39:26.590 Uttam Kumaran: Well, let’s talk. Let’s add that. So I’ll do. I’ll do. Slack rag retrieval.
365 00:39:26.840 ⇒ 00:39:27.390 Miguel de Veyra: Yeah.
366 00:39:28.260 ⇒ 00:39:34.710 Uttam Kumaran: And then I’m gonna do slack rag post, retrieve.
367 00:39:37.920 ⇒ 00:39:46.769 Miguel de Veyra: Because I think I, from what I remember, I have my docs there, I’ll add it there. But from what I remember you ha! We kinda have to
368 00:39:47.040 ⇒ 00:39:48.480 Miguel de Veyra: code this a bit.
369 00:39:50.420 ⇒ 00:39:54.430 Miguel de Veyra: Actually, I would like, but yeah.
370 00:39:55.600 ⇒ 00:39:56.210 Uttam Kumaran: Okay.
371 00:40:01.930 ⇒ 00:40:05.960 Uttam Kumaran: Okay. Great. Linear. S, 3 to super base.
372 00:40:05.960 ⇒ 00:40:08.600 Miguel de Veyra: I don’t think we’re done on the in progress stuff.
373 00:40:09.650 ⇒ 00:40:12.990 Uttam Kumaran: So this one is blocked by Brit. Oh, but this one.
374 00:40:12.990 ⇒ 00:40:15.340 Awaish Kumar: There. So like, we have linear.
375 00:40:15.340 ⇒ 00:40:18.800 Uttam Kumaran: It’s already there. So yeah, yeah, this just needs to happen. You know.
376 00:40:18.800 ⇒ 00:40:19.400 Casie Aviles: Yes.
377 00:40:20.090 ⇒ 00:40:22.469 Uttam Kumaran: Like, okay. So give me, let me let me ask a question.
378 00:40:23.460 ⇒ 00:40:26.429 Uttam Kumaran: Do we want among this group?
379 00:40:26.770 ⇒ 00:40:31.430 Uttam Kumaran: You know, I sort of consider that Oasi is handling all data engineering.
380 00:40:31.540 ⇒ 00:40:37.710 Uttam Kumaran: Do we want to continue to have sort of like everybody takes everything or
381 00:40:37.870 ⇒ 00:40:43.709 Uttam Kumaran: like, how do we want to say cool like someone is the subject matter expert when it comes to
382 00:40:44.387 ⇒ 00:40:49.980 Uttam Kumaran: super base. Someone is a subject matter expert when it comes to like building agents like.
383 00:40:50.750 ⇒ 00:40:57.540 Uttam Kumaran: I guess that’s my. That’d be my question that way before we’ve sort of been assigning it, just like as capacity exists. Yeah.
384 00:40:57.540 ⇒ 00:40:58.060 Miguel de Veyra: But.
385 00:40:58.060 ⇒ 00:41:04.649 Uttam Kumaran: For example, if if this is all related to like pre filtering, then, Miguel, I’d rather you take on
386 00:41:06.620 ⇒ 00:41:08.459 Uttam Kumaran: moving this to super base.
387 00:41:09.690 ⇒ 00:41:16.390 Miguel de Veyra: Yeah, sure, I mean, I think that would also be like, basically expertise based matchmaking.
388 00:41:18.850 ⇒ 00:41:21.980 Uttam Kumaran: Yeah, I guess, is everybody okay with that? Or.
389 00:41:22.340 ⇒ 00:41:26.869 Miguel de Veyra: That’s I think that’s fine with me. Yes, cause yeah, I’m not sure with the rest.
390 00:41:29.590 ⇒ 00:41:30.340 Casie Aviles: Yeah, that’s crazy.
391 00:41:30.340 ⇒ 00:41:31.400 Uttam Kumaran: See. What do you think?
392 00:41:32.850 ⇒ 00:41:36.790 Uttam Kumaran: If you if you’re like, hey? I still want to learn that, then I’ll leave it for you.
393 00:41:38.596 ⇒ 00:41:42.839 Miguel de Veyra: Yeah, I think while there’s specialties, there should also always be an overlap.
394 00:41:44.760 ⇒ 00:41:46.810 Uttam Kumaran: That’s that’s what I’m saying. Yeah.
395 00:41:47.970 ⇒ 00:41:53.600 Casie Aviles: Yeah, I mean, I still, I mean, I don’t mind still figuring it out. I I also want to
396 00:41:54.340 ⇒ 00:41:56.519 Casie Aviles: learn how to do rug properly.
397 00:41:57.110 ⇒ 00:42:01.496 Uttam Kumaran: Okay. Okay? So then I’m gonna leave. I’m gonna leave this for you.
398 00:42:02.100 ⇒ 00:42:05.639 Casie Aviles: This I just made prior to the meeting.
399 00:42:06.190 ⇒ 00:42:06.550 Uttam Kumaran: Okay.
400 00:42:08.020 ⇒ 00:42:08.670 Casie Aviles: Just to check.
401 00:42:08.670 ⇒ 00:42:10.480 Uttam Kumaran: So this one, I’m yeah.
402 00:42:11.006 ⇒ 00:42:13.859 Uttam Kumaran: Okay. So this is like, once it’s available.
403 00:42:15.910 ⇒ 00:42:19.059 Uttam Kumaran: it’s making sure that the client hubs can pull from rag.
404 00:42:20.380 ⇒ 00:42:21.240 Casie Aviles: Yes.
405 00:42:22.770 ⇒ 00:42:23.590 Uttam Kumaran: Okay.
406 00:42:34.060 ⇒ 00:42:35.680 Uttam Kumaran: this one’s gonna be blocked.
407 00:42:38.850 ⇒ 00:42:43.640 Uttam Kumaran: Okay, so this one is really, really high priority.
408 00:42:44.645 ⇒ 00:42:49.100 Uttam Kumaran: I guess. Yeah, Miguel, let’s talk about so I review.
409 00:42:49.550 ⇒ 00:42:56.680 Uttam Kumaran: yeah. So I reviewed the work I reviewed the website. And then, Holly, I’ll I asked Haleem to start building. So
410 00:42:57.620 ⇒ 00:43:02.240 Uttam Kumaran: give me a sense of where we’re at with this or any other steps that I wanna that we want to add here.
411 00:43:02.982 ⇒ 00:43:09.830 Miguel de Veyra: Do we have 3 active, I mean active stuff in the website. Those I’ve I framed already.
412 00:43:10.490 ⇒ 00:43:25.499 Miguel de Veyra: But I think the next step for me to do is I mean, there’s 2 steps that I wanna ask for your opinion. Do we focus on the design like, match it? Whatever is there, or do I turn everything active first, st while they’re still working on the design.
413 00:43:25.970 ⇒ 00:43:30.370 Uttam Kumaran: Yeah, yeah, no, the design. They’ll they’ll finish it all. So don’t worry about that.
414 00:43:30.570 ⇒ 00:43:37.100 Miguel de Veyra: Oh, no, I mean cause. I mean, I still have to basically design the chat interface a bit. Right?
415 00:43:37.510 ⇒ 00:43:41.920 Miguel de Veyra: So it’s either that or do I turn everything on 1st Cause? There’s.
416 00:43:41.920 ⇒ 00:43:42.590 Uttam Kumaran: Like 7.
417 00:43:42.590 ⇒ 00:43:45.760 Miguel de Veyra: In there like 4 are like down.
418 00:43:45.760 ⇒ 00:43:48.270 Uttam Kumaran: I would turn it on, I would turn it on first.st
419 00:43:48.630 ⇒ 00:43:52.120 Miguel de Veyra: Okay, and then if we turn it on, sorry, I have a couple of questions.
420 00:43:52.120 ⇒ 00:43:53.229 Uttam Kumaran: No no problem.
421 00:43:53.946 ⇒ 00:43:59.499 Miguel de Veyra: The Csr assist co-pilot. The knowledge creation agent. They’re both trained on ABC. Documents.
422 00:43:59.850 ⇒ 00:44:00.859 Miguel de Veyra: Is that fine.
423 00:44:00.860 ⇒ 00:44:01.790 Uttam Kumaran: Yeah.
424 00:44:01.930 ⇒ 00:44:05.640 Uttam Kumaran: So I wanna only I wanna show this stuff where it’s not
425 00:44:07.891 ⇒ 00:44:10.720 Uttam Kumaran: it can’t be client stuff anymore.
426 00:44:11.320 ⇒ 00:44:17.299 Uttam Kumaran: So I guess my point to you would be either have AI go through that and and change it to a fictional.
427 00:44:17.950 ⇒ 00:44:22.519 Miguel de Veyra: Yeah, I mean, we did that already for that. But it’s still their data. It’s just never about them.
428 00:44:23.960 ⇒ 00:44:26.060 Uttam Kumaran: But it’s their data meaning.
429 00:44:26.570 ⇒ 00:44:30.180 Miguel de Veyra: Like the basically the older central Bible.
430 00:44:30.960 ⇒ 00:44:34.020 Uttam Kumaran: Well, then, I would run the Central Bible through.
431 00:44:34.020 ⇒ 00:44:34.760 Miguel de Veyra: AI!
432 00:44:35.210 ⇒ 00:44:36.940 Uttam Kumaran: Oh, 3, and have it! Have it!
433 00:44:38.300 ⇒ 00:44:38.859 Miguel de Veyra: Hey! Jen.
434 00:44:38.860 ⇒ 00:44:39.340 Uttam Kumaran: Rewrite it.
435 00:44:39.340 ⇒ 00:44:40.360 Miguel de Veyra: Okay.
436 00:44:40.360 ⇒ 00:44:45.650 Uttam Kumaran: But just rewrite as part of like a let’s talk. Let’s let’s say, look, how about we do this rewrite
437 00:44:46.030 ⇒ 00:44:51.990 Uttam Kumaran: ABC Bible to be about a fictional.
438 00:44:52.850 ⇒ 00:44:53.690 Miguel de Veyra: H 5.
439 00:44:53.690 ⇒ 00:44:55.509 Uttam Kumaran: Landscaping company. Yeah.
440 00:44:55.510 ⇒ 00:44:56.190 Miguel de Veyra: Okay.
441 00:44:56.320 ⇒ 00:44:57.760 Miguel de Veyra: Okay, yeah. I’ll do that.
442 00:44:59.930 ⇒ 00:45:06.009 Uttam Kumaran: And then, yeah, just make sure there’s no ABC or anything there. What else for these.
443 00:45:07.115 ⇒ 00:45:09.490 Miguel de Veyra: Yeah. Same for the voice agent.
444 00:45:09.490 ⇒ 00:45:12.150 Uttam Kumaran: And what are what, what are, what are the lists? By the way.
445 00:45:13.570 ⇒ 00:45:14.440 Miguel de Veyra: So do this.
446 00:45:14.440 ⇒ 00:45:14.790 Uttam Kumaran: They’re all.
447 00:45:14.790 ⇒ 00:45:16.630 Miguel de Veyra: Record analysis.
448 00:45:17.470 ⇒ 00:45:18.150 Uttam Kumaran: Okay.
449 00:45:18.731 ⇒ 00:45:22.470 Miguel de Veyra: Patient record search. We probably have to come up with better names.
450 00:45:23.050 ⇒ 00:45:24.709 Uttam Kumaran: Yeah, yeah, yeah, that’s fine.
451 00:45:24.710 ⇒ 00:45:26.470 Miguel de Veyra: Lead research agent.
452 00:45:29.950 ⇒ 00:45:32.630 Miguel de Veyra: the Csr assist copilot
453 00:45:36.590 ⇒ 00:45:38.740 Miguel de Veyra: knowledge creation agent.
454 00:45:41.110 ⇒ 00:45:43.800 Miguel de Veyra: And then there’s voice agent.
455 00:45:44.150 ⇒ 00:45:48.590 Miguel de Veyra: I’m not sure if we wanna oh, there’s a problem with voice agent. By the way, Uta.
456 00:45:49.130 ⇒ 00:45:56.770 Miguel de Veyra: because it uses. If if it’s iframe, we, we can do it via we can do voice via iframe, basically
457 00:45:58.440 ⇒ 00:46:00.880 Miguel de Veyra: because the permission will be on iframe, and there’s no.
458 00:46:00.880 ⇒ 00:46:04.689 Uttam Kumaran: No, no, no, I mean no, no, you just need to be able to call.
459 00:46:06.084 ⇒ 00:46:08.280 Miguel de Veyra: Yeah. But it’s gonna call their phone number okay?
460 00:46:08.670 ⇒ 00:46:15.300 Uttam Kumaran: Well, I guess. Yeah, it’s either the voice agent calls
461 00:46:16.050 ⇒ 00:46:18.839 Uttam Kumaran: and we collect the number. Can we do that?
462 00:46:18.840 ⇒ 00:46:20.010 Miguel de Veyra: Yeah, yeah, of course.
463 00:46:20.910 ⇒ 00:46:24.930 Miguel de Veyra: Okay, then, I would prefer that voice agent.
464 00:46:24.930 ⇒ 00:46:27.160 Miguel de Veyra: I think it’s doing that. But I haven’t touched it in a while
465 00:46:31.110 ⇒ 00:46:34.759 Miguel de Veyra: or basically a a phone number that they can call right.
466 00:46:35.480 ⇒ 00:46:41.890 Uttam Kumaran: Yeah, I would prefer this or that one. But if this works, then that way we collect, we collect their phone number.
467 00:46:42.150 ⇒ 00:46:45.230 Miguel de Veyra: Okay? And then the last one is the sales co-pilot.
468 00:46:48.250 ⇒ 00:46:48.970 Uttam Kumaran: Okay.
469 00:46:53.630 ⇒ 00:46:59.710 Miguel de Veyra: So the 1st 3 I have Iframes, for already, cause. I saw your message earlier today.
470 00:47:00.760 ⇒ 00:47:01.330 Uttam Kumaran: Okay.
471 00:47:01.900 ⇒ 00:47:04.729 Miguel de Veyra: But it’s it’s not deployed yet. Of course I’m still working on it.
472 00:47:06.160 ⇒ 00:47:11.214 Uttam Kumaran: Yeah, yeah, but the iframe. So the Iframes. And then can you put can you put
473 00:47:14.390 ⇒ 00:47:18.619 Uttam Kumaran: Just put the put the put them here as you finish them. So then we can hand this to Helene.
474 00:47:18.620 ⇒ 00:47:19.939 Miguel de Veyra: Okay. Okay. Sure. Sure.
475 00:47:19.940 ⇒ 00:47:23.982 Uttam Kumaran: Or actually, that’s not good practice, can we?
476 00:47:24.580 ⇒ 00:47:29.320 Uttam Kumaran: can we make sure this ends up in there is a demo site here in notion. Right?
477 00:47:30.670 ⇒ 00:47:32.330 Uttam Kumaran: There’s a demos page.
478 00:47:32.690 ⇒ 00:47:33.630 Uttam Kumaran: Let me find it.
479 00:47:33.630 ⇒ 00:47:34.830 Miguel de Veyra: I believe so. Yes.
480 00:47:44.880 ⇒ 00:47:48.220 Uttam Kumaran: Yeah, here, so can you? Add
481 00:47:58.060 ⇒ 00:47:59.720 Miguel de Veyra: Add it there as I’m done.
482 00:48:00.550 ⇒ 00:48:08.100 Uttam Kumaran: Yeah. So add iframe and any related metadata here.
483 00:48:09.070 ⇒ 00:48:10.100 Miguel de Veyra: Okay. Cool.
484 00:48:13.650 ⇒ 00:48:21.442 Uttam Kumaran: Alright great. This is yeah, huge priority. So I wanna see us get that done.
485 00:48:27.110 ⇒ 00:48:30.379 Miguel de Veyra: Out of all that that voice is our lowest priority. Right.
486 00:48:31.460 ⇒ 00:48:33.260 Uttam Kumaran: Yeah. Voice is the lowest.
487 00:48:33.260 ⇒ 00:48:34.340 Miguel de Veyra: Okay, got it?
488 00:48:46.250 ⇒ 00:48:57.250 Uttam Kumaran: Okay, cool. So I’m gonna move these into the cycle. And then so or
489 00:48:57.390 ⇒ 00:49:03.019 Uttam Kumaran: for your awareness, Mustafa, we do 2 week cycles. So this is going to be our final week in the cycle.
490 00:49:03.560 ⇒ 00:49:09.269 Uttam Kumaran: So I’m just gonna move. The things that I feel like are should be fairly accomplishable.
491 00:49:12.020 ⇒ 00:49:13.400 Uttam Kumaran: You know, this week.
492 00:49:14.330 ⇒ 00:49:18.409 Uttam Kumaran: but you can take a look and just make sure that you can. You can do these.
493 00:49:18.650 ⇒ 00:49:19.225 Mustafa Raja: Yeah.
494 00:49:22.370 ⇒ 00:49:25.680 Uttam Kumaran: This one is the same ticket kind of right, Casey.
495 00:49:25.680 ⇒ 00:49:26.500 Casie Aviles: Yeah.
496 00:49:27.630 ⇒ 00:49:28.390 Uttam Kumaran: Okay.
497 00:49:46.340 ⇒ 00:49:48.789 Uttam Kumaran: okay, this one. I’m gonna I’m gonna leave as you.
498 00:50:05.100 ⇒ 00:50:15.999 Uttam Kumaran: Okay, great. And then, does anyone has anyone here use cloudflare pages before.
499 00:50:26.850 ⇒ 00:50:31.739 Uttam Kumaran: Okay, I’m thinking about doing that for this ticket. But it’s fine. I’m just gonna work. That’s
500 00:50:32.040 ⇒ 00:50:33.680 Uttam Kumaran: I’ll just leave this here for now.
501 00:50:36.510 ⇒ 00:50:43.070 Uttam Kumaran: Okay, and then this one, we’re not doing okay.
502 00:50:45.270 ⇒ 00:50:46.579 Miguel de Veyra: What about the Demos?
503 00:50:48.550 ⇒ 00:50:53.990 Uttam Kumaran: I mean, it’s up. It’s we only have a week left. So want to just like
504 00:50:54.280 ⇒ 00:50:57.150 Uttam Kumaran: make like there’s a lot still here to do.
505 00:50:57.150 ⇒ 00:50:57.930 Miguel de Veyra: Yep. Yep.
506 00:50:58.460 ⇒ 00:51:04.150 Uttam Kumaran: So get all that done, and then let’s just take a look at what I mentioned in slack this morning.
507 00:51:04.390 ⇒ 00:51:05.205 Uttam Kumaran: So
508 00:51:06.320 ⇒ 00:51:09.000 Uttam Kumaran: How are we on this backfill of slack messages.
509 00:51:13.890 ⇒ 00:51:20.010 Casie Aviles: I think we have all the slack messages already raw. I just think the next thing is
510 00:51:20.150 ⇒ 00:51:27.749 Casie Aviles: so we we did the I think Miguel did the pre retrieval for Javi, but not the other client. Oh, also for Eden. Okay.
511 00:51:28.100 ⇒ 00:51:30.030 Casie Aviles: by their clients.
512 00:51:32.200 ⇒ 00:51:37.199 Uttam Kumaran: Yeah, my question, Miguel, you’re gonna you’re gonna have you have to do pre renewal for every single client.
513 00:51:37.670 ⇒ 00:51:43.959 Miguel de Veyra: Yes, yes, I mean it’s pretty straightforward once it’s in Github, I mean. Sorry it wasn’t. It’s in super base.
514 00:51:44.780 ⇒ 00:51:49.470 Uttam Kumaran: But like Oh, I guess so. I guess the point here is that we don’t co mingle
515 00:51:52.960 ⇒ 00:51:55.800 Uttam Kumaran: like instead of having it all in one big table.
516 00:51:57.955 ⇒ 00:51:59.730 Miguel de Veyra: Yeah, yeah. It’s separate tables.
517 00:52:01.660 ⇒ 00:52:02.740 Uttam Kumaran: Okay, so.
518 00:52:02.740 ⇒ 00:52:09.910 Miguel de Veyra: I think, yeah, good thing this was mentioned. I think the other thing we wanna do here is basically.
519 00:52:10.780 ⇒ 00:52:14.839 Miguel de Veyra: what do you call this? Because right now it’s not scheduled. I just run the backfill.
520 00:52:15.910 ⇒ 00:52:18.040 Miguel de Veyra: So it stopped getting the latest ones.
521 00:52:20.390 ⇒ 00:52:22.899 Uttam Kumaran: Oh, where is that?
522 00:52:25.560 ⇒ 00:52:26.989 Uttam Kumaran: It’s in polyatomic.
523 00:52:28.148 ⇒ 00:52:39.579 Miguel de Veyra: No, no like the pre retrieval run. I only ran it once, basically. So whatever was in the super base during that time, I think last week, Monday or Tuesday, that’s all it has access to.
524 00:52:42.880 ⇒ 00:52:44.450 Casie Aviles: They have to run it again.
525 00:52:44.450 ⇒ 00:52:45.150 Miguel de Veyra: Yeah, we have.
526 00:52:45.150 ⇒ 00:52:47.409 Miguel de Veyra: Is that something we can run in Daxter?
527 00:52:50.604 ⇒ 00:52:52.490 Miguel de Veyra: Yeah, we’ll coordinate with the wish.
528 00:52:52.670 ⇒ 00:52:56.340 Casie Aviles: If we can schedule it, then I guess it can be on Dogstar right.
529 00:52:56.620 ⇒ 00:52:57.679 Miguel de Veyra: Yes, yes.
530 00:52:57.970 ⇒ 00:53:01.970 Uttam Kumaran: Well, so I’m gonna say, schedule pre receivable jobs in Daxter.
531 00:53:02.380 ⇒ 00:53:02.940 Miguel de Veyra: Yeah.
532 00:53:02.940 ⇒ 00:53:04.365 Uttam Kumaran: And then yes,
533 00:53:05.100 ⇒ 00:53:07.519 Miguel de Veyra: Sorry in Dexter. Can you download the file.
534 00:53:09.830 ⇒ 00:53:10.540 Uttam Kumaran: Yeah.
535 00:53:10.540 ⇒ 00:53:11.400 Miguel de Veyra: Okay, okay?
536 00:53:11.660 ⇒ 00:53:12.639 Miguel de Veyra: Because that’s how I.
537 00:53:12.640 ⇒ 00:53:13.280 Uttam Kumaran: Can you?
538 00:53:13.510 ⇒ 00:53:17.520 Uttam Kumaran: Can you walk through what it is like? What?
539 00:53:18.010 ⇒ 00:53:19.110 Uttam Kumaran: What is the schedule.
540 00:53:20.790 ⇒ 00:53:32.710 Miguel de Veyra: Basically schedule is right now, we’re basically doing it one by like for the past for the existing ones. But schedule is basically at the end of the week. This is what we want to do and then insert it.
541 00:53:33.190 ⇒ 00:53:34.659 Uttam Kumaran: I want to do every day.
542 00:53:35.880 ⇒ 00:53:37.229 Miguel de Veyra: Oh, yeah, we can do everything.
543 00:53:38.170 ⇒ 00:53:39.670 Uttam Kumaran: Like, I want this to be like
544 00:53:40.970 ⇒ 00:53:43.929 Uttam Kumaran: like, like, as fast as we can do this.
545 00:53:45.000 ⇒ 00:53:49.140 Uttam Kumaran: So let’s start with like every 4 h.
546 00:53:50.430 ⇒ 00:53:53.210 Uttam Kumaran: And then what? What are the steps that happen as part of the schedule.
547 00:53:55.797 ⇒ 00:54:06.729 Miguel de Veyra: Yeah, the this, the problem with this is that for us to retreat, to get context for rag is we have to group them all into basically a timeline.
548 00:54:10.970 ⇒ 00:54:14.769 Miguel de Veyra: So, for example, what I do is for every one week, we’re thread.
549 00:54:15.120 ⇒ 00:54:20.050 Miguel de Veyra: right? I get. I chunk them all into one context. And then, of course, if there’s too much, there’s parts.
550 00:54:20.730 ⇒ 00:54:23.989 Uttam Kumaran: So. But yeah, I’ll think about how to do this if it’s every day.
551 00:54:27.930 ⇒ 00:54:29.300 Uttam Kumaran: Yeah, so
552 00:54:30.820 ⇒ 00:54:38.240 Uttam Kumaran: okay, that’s just yeah. I just need like way more information about this. So before we can, or you can loop in a wish. But yeah.
553 00:54:39.275 ⇒ 00:54:47.090 Uttam Kumaran: coordinate with oash on how to get this to Baxter.
554 00:54:50.550 ⇒ 00:54:58.550 Uttam Kumaran: and then I also want, like alerts from Dagster and AI testing.
555 00:55:01.370 ⇒ 00:55:02.359 Miguel de Veyra: Okay. Sure. Sure.
556 00:55:05.460 ⇒ 00:55:07.490 Awaish Kumar: We need for successful runs.
557 00:55:08.780 ⇒ 00:55:13.169 Uttam Kumaran: For now. Yeah, I think just I think, for for all runs now.
558 00:55:13.690 ⇒ 00:55:14.640 Awaish Kumar: Okay.
559 00:55:22.560 ⇒ 00:55:23.260 Uttam Kumaran: Okay.
560 00:55:27.450 ⇒ 00:55:28.279 Miguel de Veyra: Okay. I have some.
561 00:55:28.280 ⇒ 00:55:35.059 Uttam Kumaran: Yeah, I mean, slack linear and Zoom Meetings are gonna be the number one. Most important thing. So
562 00:55:35.810 ⇒ 00:55:37.800 Uttam Kumaran: I wanna make sure that we can get
563 00:55:38.930 ⇒ 00:55:43.269 Uttam Kumaran: all of the slack, and then all of the zoom done as fast as we can.
564 00:55:43.540 ⇒ 00:55:47.250 Uttam Kumaran: Yes, sadly, this one, this one will be important.
565 00:55:48.495 ⇒ 00:55:52.160 Uttam Kumaran: Great cool. We talked about that.
566 00:55:53.190 ⇒ 00:55:55.620 Uttam Kumaran: Talk about this. Okay? And then in terms of
567 00:55:56.450 ⇒ 00:56:03.552 Uttam Kumaran: this item, Mustafa. So this is in the
568 00:56:04.260 ⇒ 00:56:07.110 Uttam Kumaran: Let me just find this ticket, and I can move that
569 00:56:07.680 ⇒ 00:56:09.789 Uttam Kumaran: alright. Let’s see who is here.
570 00:56:35.850 ⇒ 00:56:40.900 Uttam Kumaran: Okay. So, Mustafa, you. You did a lot of work on upwork right?
571 00:56:40.900 ⇒ 00:56:43.050 Mustafa Raja: Yeah, I’ve been hunting on a book.
572 00:56:43.800 ⇒ 00:56:52.590 Uttam Kumaran: Okay. So basically, we’re trying to work on a system that helps us automate, basically.
573 00:56:53.250 ⇒ 00:56:59.699 Uttam Kumaran: couple of things. I’m I’m just, gonna I’ll take some. I’ll take some notes here, and this is part of it. But
574 00:57:00.060 ⇒ 00:57:18.130 Uttam Kumaran: it’s gonna wanna automate finding relevant upwork opportunities, qualifying them, based on criteria, preparing proposal for human review.
575 00:57:18.480 ⇒ 00:57:23.870 Uttam Kumaran: And then manual proposal submission
576 00:57:31.080 ⇒ 00:57:37.369 Uttam Kumaran: is anyone. Have you used the computer use? Or like browser use agents before.
577 00:57:38.400 ⇒ 00:57:38.880 Mustafa Raja: No.
578 00:57:39.960 ⇒ 00:57:41.339 Uttam Kumaran: But you’re familiar with it.
579 00:57:42.478 ⇒ 00:57:46.009 Mustafa Raja: I I believe I I.
580 00:57:46.010 ⇒ 00:57:47.410 Uttam Kumaran: Like you try seeing a video on it.
581 00:57:47.410 ⇒ 00:57:49.240 Mustafa Raja: Yeah, yeah, I’m familiar with.
582 00:57:49.240 ⇒ 00:57:52.240 Uttam Kumaran: So you know, basically, it’s like controlling your browser.
583 00:57:52.240 ⇒ 00:57:52.580 Mustafa Raja: Yeah.
584 00:57:53.540 ⇒ 00:57:55.077 Uttam Kumaran: Okay, okay, so
585 00:57:55.590 ⇒ 00:57:57.939 Mustafa Raja: A menace was one of these.
586 00:58:00.140 ⇒ 00:58:03.169 Mustafa Raja: yeah, Browser base is one of these, and then.
587 00:58:03.170 ⇒ 00:58:05.718 Uttam Kumaran: Computer use. There is
588 00:58:06.590 ⇒ 00:58:09.729 Uttam Kumaran: you can do computer use through Claude also.
589 00:58:10.220 ⇒ 00:58:12.520 Uttam Kumaran: And I think Openai has it, too.
590 00:58:13.149 ⇒ 00:58:20.340 Uttam Kumaran: I guess one thing that I want to show is we have this upwork tracker that we’re like
591 00:58:20.910 ⇒ 00:58:24.989 Uttam Kumaran: that we’re starting to and I’ll just make sure everyone has this
592 00:58:29.430 ⇒ 00:58:32.670 Uttam Kumaran: that we’re starting to use to bid on upwork jobs.
593 00:58:32.900 ⇒ 00:58:40.059 Uttam Kumaran: And so a lot of I think you’re this this sprint and next sprint for you is going to be working closely with me on
594 00:58:40.576 ⇒ 00:58:56.533 Uttam Kumaran: on how we streamline this process? So we want to go from. We just wanna be. And the nice thing is you’ve you’ve been on that work. So you kind of know what the challenge I’m talking about is, we’re basically what we’re doing is
595 00:58:58.030 ⇒ 00:58:59.323 Uttam Kumaran: we have
596 00:59:00.910 ⇒ 00:59:06.293 Uttam Kumaran: like this is an example of something we’ve done manually where we go, and we find like,
597 00:59:06.940 ⇒ 00:59:09.180 Uttam Kumaran: here’s the open item.
598 00:59:09.840 ⇒ 00:59:11.300 Uttam Kumaran: Second.
599 00:59:16.910 ⇒ 00:59:17.910 Uttam Kumaran: find another one.
600 00:59:22.830 ⇒ 00:59:24.840 Uttam Kumaran: So here, here’s an example. Right?
601 00:59:24.980 ⇒ 00:59:28.500 Uttam Kumaran: So let me log in.
602 00:59:54.340 ⇒ 00:59:55.610 Uttam Kumaran: Hold on one second.
603 01:00:03.780 ⇒ 01:00:05.900 Uttam Kumaran: Okay, so
604 01:00:06.900 ⇒ 01:00:12.280 Uttam Kumaran: if I go back and I click on this, and you go to upwork, and you bid on a proposal.
605 01:00:12.390 ⇒ 01:00:14.609 Uttam Kumaran: You can click on. Apply now
606 01:00:14.720 ⇒ 01:00:18.599 Uttam Kumaran: and then. What happens is you have to then fill out a bunch of these
607 01:00:19.608 ⇒ 01:00:25.669 Uttam Kumaran: this information, right? So we we just want to have AI help us both find
608 01:00:25.880 ⇒ 01:00:30.570 Uttam Kumaran: the ones that fit our criteria, which is in that spreadsheet, and then also
609 01:00:30.810 ⇒ 01:00:41.709 Uttam Kumaran: prepare the notes. The lovely thing is, we have all this stuff about our company. Our case studies is all available, so it should be a fairly simple prompt to then take
610 01:00:42.140 ⇒ 01:00:46.710 Uttam Kumaran: this and then basically write a pretty easy case. Study?
611 01:00:47.268 ⇒ 01:00:49.601 Uttam Kumaran: So that’s that’s that’s this goal.
612 01:00:52.770 ⇒ 01:00:55.969 Uttam Kumaran: Any like initial questions. There.
613 01:00:59.450 ⇒ 01:01:03.216 Mustafa Raja: I guess I’d need some
614 01:01:05.897 ⇒ 01:01:15.140 Mustafa Raja: I mean, I guess my criteria that we are filtering these projects in the excel file.
615 01:01:15.960 ⇒ 01:01:20.520 Mustafa Raja: So I think it started there, you know, automating at least that.
616 01:01:22.260 ⇒ 01:01:25.370 Uttam Kumaran: Yeah. So the criteria is in this docs, isn’t is, it’s in here.
617 01:01:25.730 ⇒ 01:01:27.590 Uttam Kumaran: So I will.
618 01:01:28.942 ⇒ 01:01:36.240 Uttam Kumaran: I’m gonna what I’ll do is I’m gonna put another page here called criteria.
619 01:01:39.470 ⇒ 01:01:52.820 Uttam Kumaran: And I’m going to take this this, and then I’ll say, criteria values.
620 01:01:57.890 ⇒ 01:01:59.359 Uttam Kumaran: So I’ll fill that out.
621 01:02:01.100 ⇒ 01:02:07.340 Uttam Kumaran: But yeah, I guess what I want to sort of have you go through is if you can leave any commentary here on like
622 01:02:07.760 ⇒ 01:02:21.387 Uttam Kumaran: before you start working on this or before you sort of go like where very deep. Just leave some comments here that way. The team can give. You can give feedback, and then we can be like cool. This is good to go. Go ahead and and work on it.
623 01:02:21.660 ⇒ 01:02:22.200 Mustafa Raja: Sure.
624 01:02:22.690 ⇒ 01:02:29.969 Uttam Kumaran: Since you’re really familiar upwork, I feel like you know exactly what this is. So just let me know what you think.
625 01:02:30.753 ⇒ 01:02:34.119 Uttam Kumaran: This is. This is like something that would be like.
626 01:02:34.390 ⇒ 01:02:40.390 Uttam Kumaran: extremely, extremely helpful for our sales team. The other item.
627 01:02:42.260 ⇒ 01:02:51.526 Uttam Kumaran: Yeah. So it’s lead list building. And this is something that I will also bring here.
628 01:03:06.990 ⇒ 01:03:09.379 Uttam Kumaran: I’m just gonna move this to our team.
629 01:03:43.390 ⇒ 01:03:44.580 Uttam Kumaran: who is
630 01:04:03.090 ⇒ 01:04:03.780 Uttam Kumaran: not?
631 01:04:08.100 ⇒ 01:04:10.050 Uttam Kumaran: I’ll move to. Okay, there it is.
632 01:04:12.810 ⇒ 01:04:19.520 Uttam Kumaran: Okay. So can you, do you mind taking a look at this project?
633 01:04:20.170 ⇒ 01:04:23.059 Uttam Kumaran: And basically, what we want to do is help.
634 01:04:25.660 ⇒ 01:04:30.980 Uttam Kumaran: Is helped our sales team build lead list, which is basically like
635 01:04:31.110 ⇒ 01:04:40.380 Uttam Kumaran: a couple of different stories. So I want to upload a list of target accounts. And the system can find and return qualified decision makers
636 01:04:40.530 ⇒ 01:04:43.059 Uttam Kumaran: I won to create.
637 01:04:43.220 ⇒ 01:04:47.750 Uttam Kumaran: There’s a couple of different things here. Have. Have you used clay before.
638 01:04:48.278 ⇒ 01:04:51.320 Mustafa Raja: I have just a little, though. I.
639 01:04:51.320 ⇒ 01:04:51.650 Uttam Kumaran: Okay. Okay.
640 01:04:51.650 ⇒ 01:04:53.350 Mustafa Raja: Familiar with the clear and the portal.
641 01:04:54.180 ⇒ 01:05:00.679 Uttam Kumaran: Okay, okay, alright cool. So just take a take a look at this today, and just give me a sense of like.
642 01:05:03.030 ⇒ 01:05:07.650 Uttam Kumaran: what you think of the project I’ll set. I’ll put this. I’ll put this here as well.
643 01:05:16.430 ⇒ 01:05:20.769 Uttam Kumaran: This was part of a different team, and then we moved it to the AI team, so I don’t wanna like
644 01:05:21.420 ⇒ 01:05:24.810 Uttam Kumaran: they may have done in a different way. So I want you to sort of take a look first.st
645 01:05:24.940 ⇒ 01:05:25.520 Mustafa Raja: Okay.
646 01:05:26.220 ⇒ 01:05:26.900 Uttam Kumaran: Okay.
647 01:05:27.350 ⇒ 01:05:34.680 Uttam Kumaran: okay, cool. We’re at 9, 30. Anything else guys. I’m gonna so I’m gonna be out of office
648 01:05:35.583 ⇒ 01:05:37.070 Uttam Kumaran: after Tuesday.
649 01:05:37.794 ⇒ 01:05:43.450 Uttam Kumaran: So I wanna make sure that we just have enough stuff after tomorrow to write out the week.
650 01:05:44.008 ⇒ 01:05:51.539 Uttam Kumaran: I mean, of course, if you need to get a hold of me. I’m I can, you know I’m happy to. I’ll be checking linear and stuff. But
651 01:05:52.025 ⇒ 01:05:54.630 Uttam Kumaran: I just wanna make sure we can get.
652 01:05:55.010 ⇒ 01:05:57.630 Uttam Kumaran: I’ll be, I mean I’ll be back on the
653 01:05:58.700 ⇒ 01:06:00.950 Uttam Kumaran: I’ll be back here on Monday next Monday.
654 01:06:01.100 ⇒ 01:06:01.840 Miguel de Veyra: Okay. Okay.
655 01:06:02.640 ⇒ 01:06:03.550 Uttam Kumaran: But like
656 01:06:03.840 ⇒ 01:06:12.959 Uttam Kumaran: I have my phone, I have my stuff. So just tag me if you need anything. But I I think what we have in our sprint is is definitely like enough work. I just wanna make sure that
657 01:06:13.350 ⇒ 01:06:21.089 Uttam Kumaran: couple of things. So I want us to get get the slack zoom linear combo out for all agents.
658 01:06:21.220 ⇒ 01:06:31.508 Uttam Kumaran: I want us to make sure that Mustafa is up to speed on everything and has access to everything. Away shall get you snowflake. And then
659 01:06:32.990 ⇒ 01:06:49.459 Uttam Kumaran: yeah, I think the biggest item Mustafa is for you to just get get up to going on some of the sales stuff. I don’t think it’s gonna be that hard for you. So but like you’ll be working directly with me on that, and then that way, this team we we sort of covered all our bases. So
660 01:06:49.610 ⇒ 01:06:50.816 Uttam Kumaran: I’m excited.
661 01:06:52.080 ⇒ 01:06:54.799 Uttam Kumaran: Just let me know if I can unblock anything.
662 01:06:55.150 ⇒ 01:07:01.839 Mustafa Raja: Yeah. So so for the slack thing, I might need access to the slack settings for slack app
663 01:07:02.220 ⇒ 01:07:02.800 Mustafa Raja: building.
664 01:07:02.800 ⇒ 01:07:05.399 Uttam Kumaran: Cool, Casey, do you? Can you give him access to that?
665 01:07:06.180 ⇒ 01:07:07.459 Uttam Kumaran: Or do you need me to.
666 01:07:07.790 ⇒ 01:07:08.689 Casie Aviles: Yeah, sure.
667 01:07:09.350 ⇒ 01:07:10.659 Mustafa Raja: Yeah, you can set it up.
668 01:07:10.660 ⇒ 01:07:11.170 Mustafa Raja: Okay, yeah.
669 01:07:12.470 ⇒ 01:07:16.810 Uttam Kumaran: Okay, yeah. Just ping in ping in the AI team channel with any access you need.
670 01:07:16.810 ⇒ 01:07:18.559 Mustafa Raja: Yeah. And then, yeah, let’s
671 01:07:19.040 ⇒ 01:07:22.769 Mustafa Raja: So for any questions, I I’ll just ping in the AI team right.
672 01:07:23.290 ⇒ 01:07:24.519 Uttam Kumaran: Yes, please.
673 01:07:24.520 ⇒ 01:07:25.890 Mustafa Raja: Yeah, okay, thank you so much.
674 01:07:27.050 ⇒ 01:07:27.850 Uttam Kumaran: Okay.
675 01:07:28.460 ⇒ 01:07:29.350 Uttam Kumaran: Alright. Thank you.
676 01:07:29.350 ⇒ 01:07:37.380 Miguel de Veyra: Yeah, I’m sorry. Last question. Since you’re not gonna be here, I think after Monday, can we move the meeting a bit earlier?
677 01:07:38.220 ⇒ 01:07:39.230 Uttam Kumaran: Yes.
678 01:07:40.450 ⇒ 01:07:41.359 Miguel de Veyra: Okay, okay.
679 01:07:46.800 ⇒ 01:07:48.770 Miguel de Veyra: is it just me? Or did autumn cut off.
680 01:08:01.510 ⇒ 01:08:03.269 Uttam Kumaran: Sorry. I think my audio got cut off.
681 01:08:03.270 ⇒ 01:08:03.700 Miguel de Veyra: I said.
682 01:08:03.700 ⇒ 01:08:05.729 Uttam Kumaran: I’ll just Zoom Meeting as is.
683 01:08:06.030 ⇒ 01:08:06.920 Miguel de Veyra: Okay. Okay.
684 01:08:08.160 ⇒ 01:08:09.959 Miguel de Veyra: Okay. Thanks. Everyone.
685 01:08:10.980 ⇒ 01:08:11.460 Mustafa Raja: Thank you.