Meeting Title: Uttam Kumaran Date: 2025-02-28 Meeting participants: Mariane Cequina, Nicolas Sucari, Michael Weinberg, Uttam Kumaran, Hannah Wang, Miguel De Veyra, Casie Aviles, Sahana Asokan, Payas, Ryan, Caio Velasco
WEBVTT
1 00:00:30.780 ⇒ 00:00:31.629 Nicolas Sucari: Hi team.
2 00:00:31.810 ⇒ 00:00:32.750 Uttam Kumaran: Hey, guys.
3 00:00:34.730 ⇒ 00:00:35.440 Caio Velasco: So.
4 00:00:49.480 ⇒ 00:00:51.080 Uttam Kumaran: Need to find a background
5 00:01:42.670 ⇒ 00:01:44.550 Uttam Kumaran: that’s not a good background.
6 00:01:44.550 ⇒ 00:01:46.869 Ryan: 1st Zoom Meeting Location.
7 00:01:50.070 ⇒ 00:01:50.760 Ryan: Okay.
8 00:02:24.000 ⇒ 00:02:26.089 michael weinberg: Oh, shit! Why is the Titanic sinking.
9 00:02:28.500 ⇒ 00:02:32.439 Nicolas Sucari: Yeah, the botanic is sinking. So we had, like our
10 00:02:33.334 ⇒ 00:02:39.409 Nicolas Sucari: Icebreaker for today is about like what would be the worst Zoom Meeting location.
11 00:02:41.000 ⇒ 00:02:48.979 Nicolas Sucari: so I guess mine will be like if you’re on on the deck on the Titanic, and it’s sinking, and you still need to be at the meeting.
12 00:02:49.160 ⇒ 00:02:50.260 Nicolas Sucari: It’ll be funny.
13 00:02:52.730 ⇒ 00:02:56.150 michael weinberg: I have to think about this. Now, this is this is a really hard problem.
14 00:02:57.210 ⇒ 00:02:58.140 Nicolas Sucari: Yeah.
15 00:02:58.680 ⇒ 00:03:02.889 Nicolas Sucari: Just imagine people running everywhere and jumping off board
16 00:03:03.080 ⇒ 00:03:09.359 Nicolas Sucari: water down your feet, and you’re still there like showing off some dashboards. Nice.
17 00:03:11.470 ⇒ 00:03:15.669 michael weinberg: I I see what you’re I see what you’re you’re saying. I’ve got some ideas now.
18 00:03:20.430 ⇒ 00:03:35.449 Hannah Wang: I remember when I was in school. Cause I did like Covid started when I was still a senior, like a lot of people or not a lot of people, but a background that I saw. Commonly it’s not the worst background, but they would just put like a screenshot of
19 00:03:35.560 ⇒ 00:03:41.969 Hannah Wang: themselves as their background, and just leave the screen. And like, do other stuff. So it looks like you’re attending class. But
20 00:03:42.110 ⇒ 00:03:43.580 Hannah Wang: you’re not there.
21 00:03:44.120 ⇒ 00:03:46.880 Hannah Wang: I feel like I’ve seen that a couple of times.
22 00:03:48.050 ⇒ 00:03:48.910 Uttam Kumaran: That’s funny.
23 00:03:58.770 ⇒ 00:03:59.910 Nicolas Sucari: What’s that? Utah?
24 00:04:00.820 ⇒ 00:04:03.589 Uttam Kumaran: This is like a desert, and it was like.
25 00:04:03.590 ⇒ 00:04:04.250 Nicolas Sucari: Okay.
26 00:04:04.250 ⇒ 00:04:09.680 Uttam Kumaran: Yeah, it’s it was just so windy. And there was basically like sand blowing everywhere. And you have to like wear.
27 00:04:10.050 ⇒ 00:04:14.529 Uttam Kumaran: It’s like, wear a bandana, and it’s it would be a horrible place to take a zoom call from.
28 00:04:17.529 ⇒ 00:04:22.300 Nicolas Sucari: All the sun coming through your eyes. Yeah, yeah, I can feel it.
29 00:04:22.550 ⇒ 00:04:23.570 Nicolas Sucari: It’s painful.
30 00:04:23.770 ⇒ 00:04:24.355 Uttam Kumaran: Yes.
31 00:04:26.920 ⇒ 00:04:27.480 Ryan: Mindslip.
32 00:04:27.480 ⇒ 00:04:27.830 Nicolas Sucari: Like a little.
33 00:04:27.830 ⇒ 00:04:29.150 Ryan: Graphic concert.
34 00:04:40.610 ⇒ 00:04:47.150 Nicolas Sucari: What about the the other ones? Come on, guys, just think of a a worse place for you.
35 00:04:47.340 ⇒ 00:04:51.429 Nicolas Sucari: If you needed to go into a zoom like you will suffer.
36 00:04:52.620 ⇒ 00:04:55.990 michael weinberg: Has anyone seen the movie inglorious bastards.
37 00:04:56.990 ⇒ 00:04:57.670 Nicolas Sucari: Yeah.
38 00:04:57.870 ⇒ 00:04:58.680 michael weinberg: Okay.
39 00:04:59.040 ⇒ 00:05:00.180 michael weinberg: Can you guess?
40 00:05:03.850 ⇒ 00:05:06.270 michael weinberg: The final scene. Are you in the theater?
41 00:05:07.910 ⇒ 00:05:10.640 michael weinberg: That excellent guess?
42 00:05:21.950 ⇒ 00:05:24.100 michael weinberg: Let’s see if Zoom actually decides to work.
43 00:05:26.000 ⇒ 00:05:30.949 michael weinberg: I’m on. I so I recently set up Linux on like an old windows computer.
44 00:05:31.570 ⇒ 00:05:36.610 michael weinberg: and most things work. But occasionally things just don’t.
45 00:05:49.680 ⇒ 00:05:55.739 Nicolas Sucari: Are we waiting for someone else to come? I don’t know if I think Roberts was not gonna join Bo either.
46 00:05:57.080 ⇒ 00:05:59.889 Uttam Kumaran: Did you end up sending this to Demo a.
47 00:06:01.260 ⇒ 00:06:06.089 Nicolas Sucari: Yes, you added in there, I think it wasn’t there.
48 00:06:06.090 ⇒ 00:06:08.289 Uttam Kumaran: Doesn’t look like he’s on the invite. So.
49 00:06:09.730 ⇒ 00:06:14.349 Nicolas Sucari: Yeah, but he was there when I tried to send it. It was already in.
50 00:06:14.700 ⇒ 00:06:15.360 Uttam Kumaran: Oh!
51 00:06:15.680 ⇒ 00:06:21.649 Mariane Cequina: I think he sent a message in the slack, Robert, he said, that Boo and him will miss the meeting.
52 00:06:29.190 ⇒ 00:06:31.860 Uttam Kumaran: Yeah, I think we’re just missing still, the AI team, though.
53 00:06:32.830 ⇒ 00:06:39.220 Nicolas Sucari: Yeah, let me ping them a lot of maybe I I can’t send like add new ones. I don’t know what’s going on there, but.
54 00:07:03.540 ⇒ 00:07:08.420 Uttam Kumaran: Okay. I mean, we can get started. I don’t know, Nico, if you’re I don’t know if you’re messaging other people or
55 00:07:09.790 ⇒ 00:07:10.730 Uttam Kumaran: or what.
56 00:07:13.570 ⇒ 00:07:15.060 Nicolas Sucari: Yeah, I got started.
57 00:07:15.190 ⇒ 00:07:18.690 Nicolas Sucari: I don’t know if anyone else want to share the worst location.
58 00:07:19.070 ⇒ 00:07:21.410 Nicolas Sucari: Come on, Casey, at a background. There.
59 00:07:26.063 ⇒ 00:07:26.969 Nicolas Sucari: what’s that.
60 00:07:27.350 ⇒ 00:07:28.600 Uttam Kumaran: Like a land party.
61 00:07:29.170 ⇒ 00:07:33.700 Casie Aviles: Yeah, yeah, in the in the Philippines, we have these computer shops and
62 00:07:34.230 ⇒ 00:07:38.494 Casie Aviles: people tend to get carried away playing band games and stuff. So.
63 00:07:39.480 ⇒ 00:07:40.509 michael weinberg: That makes sense.
64 00:07:40.760 ⇒ 00:07:41.309 Miguel de Veyra: Hey, everyone.
65 00:07:41.310 ⇒ 00:07:41.990 Casie Aviles: Yeah.
66 00:07:46.570 ⇒ 00:07:50.250 Uttam Kumaran: My God, you gotta choose a background of the worst place to take a Zoom Meeting.
67 00:07:52.530 ⇒ 00:07:55.200 Miguel de Veyra: That’s where I grew up, Casey. Computer shops.
68 00:07:56.120 ⇒ 00:07:56.770 Casie Aviles: Yeah.
69 00:07:57.780 ⇒ 00:07:58.430 Miguel de Veyra: They!
70 00:08:00.930 ⇒ 00:08:04.502 Ryan: I guess everybody did, and I guess in the Philippines.
71 00:08:04.860 ⇒ 00:08:06.480 Miguel de Veyra: Yeah. Mute.
72 00:08:07.150 ⇒ 00:08:07.669 Miguel de Veyra: Thank you.
73 00:08:07.670 ⇒ 00:08:12.290 Ryan: I wouldn’t eat lunch, so I can save money to to play.
74 00:08:12.770 ⇒ 00:08:13.530 Casie Aviles: Yeah, true, for sure.
75 00:08:13.530 ⇒ 00:08:14.070 Ryan: Shop.
76 00:08:14.440 ⇒ 00:08:17.730 Miguel de Veyra: I used to wake up like 4 am. Just escape from my parents.
77 00:08:17.960 ⇒ 00:08:20.039 Miguel de Veyra: Tell them going football practice.
78 00:08:25.060 ⇒ 00:08:27.869 michael weinberg: Esports are a thing now. Did it pay off.
79 00:08:28.050 ⇒ 00:08:29.175 Miguel de Veyra: No, no. Bro.
80 00:08:34.750 ⇒ 00:08:36.640 Miguel de Veyra: Where? How do you pick one?
81 00:08:37.020 ⇒ 00:08:38.609 Miguel de Veyra: I have presets.
82 00:08:41.549 ⇒ 00:08:45.009 Hannah Wang: Yeah, I forgot how to add a background, too. I’m trying to figure that out.
83 00:08:45.010 ⇒ 00:08:51.039 Uttam Kumaran: If you go to things, and then you go to background, and then you click on virtual background.
84 00:08:52.290 ⇒ 00:08:54.500 Uttam Kumaran: And then there’s a little plus button.
85 00:08:55.770 ⇒ 00:08:58.040 Hannah Wang: Oh, that’s tiny. Okay.
86 00:08:58.040 ⇒ 00:08:58.820 Uttam Kumaran: Yes.
87 00:08:59.360 ⇒ 00:09:00.420 Hannah Wang: God, he works.
88 00:09:00.420 ⇒ 00:09:02.069 Miguel de Veyra: There you go. Yeah, I wouldn’t.
89 00:09:03.540 ⇒ 00:09:08.969 michael weinberg: I. So I’ve been pressing this plus button over and over, and then uploading a file, and then nothing.
90 00:09:10.840 ⇒ 00:09:11.650 michael weinberg: I’m like.
91 00:09:12.090 ⇒ 00:09:15.896 Uttam Kumaran: Just print it out and like hold it, hold it behind your head, or something.
92 00:09:16.310 ⇒ 00:09:19.540 michael weinberg: Yeah, that’s true. I’ll I’ll just share my screen
93 00:09:27.460 ⇒ 00:09:31.079 michael weinberg: for 1 min, and then that gives everyone a sense.
94 00:09:34.450 ⇒ 00:09:39.250 michael weinberg: This is this is the scene when, like the
95 00:09:39.390 ⇒ 00:09:44.369 michael weinberg: the Jewish French women, the Americans, and everyone all kind of just let loose at once.
96 00:09:45.440 ⇒ 00:09:49.949 michael weinberg: And the entire leadership is there. It’s like the most glorious scene of all time.
97 00:09:52.220 ⇒ 00:09:54.159 Miguel de Veyra: Which from which movie is this.
98 00:09:54.160 ⇒ 00:09:56.559 michael weinberg: So this is from the movie inglorious bastards.
99 00:09:56.560 ⇒ 00:09:57.550 Miguel de Veyra: I’m Loris pastors.
100 00:09:57.550 ⇒ 00:09:58.440 michael weinberg: Yeah, it’s so good.
101 00:09:58.900 ⇒ 00:10:00.320 Miguel de Veyra: Going to Tarantino.
102 00:10:00.490 ⇒ 00:10:02.910 michael weinberg: So yeah, it’s it’s it’s definitely like
103 00:10:03.750 ⇒ 00:10:07.720 michael weinberg: the dumbest Tarantino movie in like, the best way.
104 00:10:14.770 ⇒ 00:10:23.719 Uttam Kumaran: Okay, cool. I guess, Nicole, I’ll get started. Yeah, I don’t see Demo A on the invite. Oh, actually. Oh, I do see him. But I okay, I messaged him. Let’s see
105 00:10:24.578 ⇒ 00:10:28.519 Uttam Kumaran: but let’s get started. I’ll share slides for today.
106 00:10:29.600 ⇒ 00:10:31.270 Uttam Kumaran: Great.
107 00:10:54.470 ⇒ 00:11:03.269 Uttam Kumaran: great. So I think once we start off. Typically all of these meetings again, looking at like our mission statement and values.
108 00:11:03.655 ⇒ 00:11:18.750 Uttam Kumaran: I think we did a better job, I think, last 2 weeks, 3 weeks. As we’ve been presenting this, we’re doing a better job. I think we did a lot on Item one, I think this week especially, which was, we have a lot of new team members that join. I think everyone is leveling up a lot. So
109 00:11:19.030 ⇒ 00:11:22.119 Uttam Kumaran: I’m I’m really excited there.
110 00:11:22.540 ⇒ 00:11:25.050 Uttam Kumaran: But for folks that are new. I think.
111 00:11:25.710 ⇒ 00:11:33.089 Uttam Kumaran: These are sort of principled ways to, as you guys make decisions on behalf of yourself or the business or clients to think
112 00:11:33.430 ⇒ 00:11:37.080 Uttam Kumaran: about these things. And hopefully, these are good ways to
113 00:11:37.200 ⇒ 00:11:51.689 Uttam Kumaran: sort of consider when we’re making decisions. I think we’ve had a lot of good conversations across all teams about each of these. And and how do we do more with less? And how do we move faster? And so I think this is a good place to to as a good reminder.
114 00:11:52.157 ⇒ 00:12:01.329 Uttam Kumaran: And then, yeah, I guess we do have Mike on this call. And Demo Lade, I don’t think is on this call but maybe, Mike.
115 00:12:01.460 ⇒ 00:12:09.568 Uttam Kumaran: I will maybe let. I’ll just let you do an intro I just did another intro on another call, so I may let you just
116 00:12:10.170 ⇒ 00:12:18.719 Uttam Kumaran: give this feel, and and would love to I could fill in any any blanks but very excited to to have you and have you meet the rest of the team.
117 00:12:19.560 ⇒ 00:12:21.992 michael weinberg: Sweet. Yeah, I’ll do the fast version.
118 00:12:23.116 ⇒ 00:12:35.909 michael weinberg: so I have done sort of an around the world and data I’ve not been an official product manager, but I’ve done my fair share of it. But I started as a
119 00:12:36.632 ⇒ 00:12:41.730 michael weinberg: sort of a data scientist doing some like predictive work for some consultancy
120 00:12:42.245 ⇒ 00:12:52.119 michael weinberg: you know, ended up doing a lot of work databases, and then got progressively deeper into what would basically become data, engineering and
121 00:12:52.360 ⇒ 00:12:56.090 michael weinberg: kind of the modern day, like analytics, engineering, plus a lot of the like
122 00:12:56.350 ⇒ 00:13:01.700 michael weinberg: system administration of warehouses. Performance. Kind of leads you into
123 00:13:01.880 ⇒ 00:13:16.129 michael weinberg: data governance. Thinking about stuff for bigger enterprises. I worked with Utam at. We work at the moment when masoshun decided that we work, maybe is worth 110 billion dollars.
124 00:13:16.380 ⇒ 00:13:21.490 michael weinberg: Guy pulled a number out of a hat, I’m sure. But we were growing super fast. So
125 00:13:22.012 ⇒ 00:13:28.007 michael weinberg: ended up getting experience with, you know, kind of crazy stuff, hyper growth, all the weird things.
126 00:13:28.670 ⇒ 00:13:32.210 michael weinberg: I worked for 2 presidential campaigns.
127 00:13:32.910 ⇒ 00:13:39.154 michael weinberg: I’m I’m anti-fascist, so I won’t even I don’t think I need to say who you can guess.
128 00:13:40.150 ⇒ 00:13:48.919 michael weinberg: I’m I’ve been a manager and a staff engineer in various roles in the last couple of years.
129 00:13:49.514 ⇒ 00:13:54.470 michael weinberg: And here in Brain Forge, I’m happy to kind of help out, and
130 00:13:55.100 ⇒ 00:13:59.170 michael weinberg: particularly an internal capacity, and see how I can help make the team grow.
131 00:13:59.951 ⇒ 00:14:04.309 michael weinberg: If you guys have any questions, I mean, just like, you know, reach out
132 00:14:04.520 ⇒ 00:14:06.600 michael weinberg: schedule time. Happy to chat.
133 00:14:07.460 ⇒ 00:14:10.900 Uttam Kumaran: Yeah, I see Mike as a really one great resource.
134 00:14:11.301 ⇒ 00:14:23.969 Uttam Kumaran: For our clients, as sort of as we take on harder and harder projects that involve tons of stakeholders and really complicated technology. I also see Mike is a really great resource for everybody on engineering. On how to go from
135 00:14:24.020 ⇒ 00:14:47.680 Uttam Kumaran: just, you know, one engineer on one team to to leveling up, not only in a technical capacity, but, more importantly, on how to work with others, and sort of have your voice be heard. So I will start connecting Mike directly with a few people. To sort of just spend time, you know, on on whatever anyone needs support. I think it’s a great resource on how to become a better engineer and a better team member.
136 00:14:48.040 ⇒ 00:14:54.639 Uttam Kumaran: and Mike is also a really really incredible friend of mine. So 1st of all, so I’m really happy to have him help us out.
137 00:14:56.480 ⇒ 00:15:01.700 Uttam Kumaran: And Michael find ways to help across the company, so I’m excited for him to sort of see everything in slack and
138 00:15:02.191 ⇒ 00:15:19.659 Uttam Kumaran: also for him to see this AI side as well, so great. And then, yeah, I don’t think Demo Lade is on here, but I will probably wait for Monday then for him to give his intro. But Demo Lade will be joining us on the analytics engineering side. Actually previously worked. With Mike
139 00:15:20.072 ⇒ 00:15:28.260 Uttam Kumaran: at another company. But it’s also coming in as a really great analytics engineer, and sort of excited to welcome him onto the team as well.
140 00:15:29.670 ⇒ 00:15:39.559 Uttam Kumaran: Yeah, and then moving on to next. So in terms of update overall so very excited to close out this month. Month end is sort of
141 00:15:40.380 ⇒ 00:15:48.150 Uttam Kumaran: usually just as stressful as sort of month start. We have to close books and collect payments and pay everyone out. So I’m sort of
142 00:15:48.560 ⇒ 00:16:12.279 Uttam Kumaran: focus on that a little bit. This weekend this will be our biggest month in the history of the company. We are currently servicing the most clients we ever had. We are upselling clients, and taking on more work for them. We’ve also grown. This is the at the moment the largest the company has ever been. So a lot of 1st this month, which I’m really really excited, for
143 00:16:12.380 ⇒ 00:16:23.630 Uttam Kumaran: all that helps me do is set more audacious goals and ask that we move faster, and that we execute, you know more on behalf of clients. We are.
144 00:16:23.750 ⇒ 00:16:41.095 Uttam Kumaran: We are basically running, you know, 6 or 7 full engagements. Which is really really amazing. I will talk pretty candidly about how I think some of those are going but I feel really, really happy that we have the opportunity to to do that, and also the opportunity to build a team around that
145 00:16:41.520 ⇒ 00:16:46.229 Uttam Kumaran: we are constantly recruiting. So again, I’ll remind everyone that if there’s someone in your
146 00:16:46.400 ⇒ 00:17:03.670 Uttam Kumaran: life that you you think you would describe as very smart. Please introduce them to me, no matter what they’re doing. I’m happy to talk to them and see if there’s opportunity to work together. And yeah, I think we are starting to to finally get much more
147 00:17:04.105 ⇒ 00:17:25.239 Uttam Kumaran: organized as well on the marketing and design side. We had a really great meeting yesterday about our growth on how we can start to build out content and campaigns excited to continue that we had a lot of. We had a great data retro right before this, a lot of really great learnings. And the AI team definitely crushed it this week. We have some really, really cool Demos to share
148 00:17:25.930 ⇒ 00:17:29.000 Uttam Kumaran: any questions on any of that that I can answer.
149 00:17:32.350 ⇒ 00:17:36.615 Uttam Kumaran: Those are the things that are on my mind. I’m happy to, of course, to talk about anything but
150 00:17:37.550 ⇒ 00:17:38.529 Uttam Kumaran: Let me know.
151 00:17:39.100 ⇒ 00:17:56.379 Nicolas Sucari: Yeah, apart from the feedback that you gave on the retro board. If anyone has any other feedback on, if it was easy or difficult to use that. Let me know. And yeah, we can work on improving that board. So it’s easier for everyone on each retro to use the tool. Okay.
152 00:17:58.350 ⇒ 00:18:24.184 Uttam Kumaran: Yeah. And I think the retro process I I really love. I think it’s a really good space for everybody to sort of reflect. I think we will honestly start to expand that just from engineering to other parts of the company operations. Marketing sales is we? Just? It’s so small. So Robert is doing a lot of retro himself. But it’s a great process for us to take a look at what we’re doing and and how we’re growing so
153 00:18:25.430 ⇒ 00:18:29.359 Uttam Kumaran: great terms of client health. So
154 00:18:29.510 ⇒ 00:18:36.120 Uttam Kumaran: I wanna start with this. I wanna start with talking about why we’re all here is to serve our clients and the work we’re doing there.
155 00:18:36.677 ⇒ 00:18:41.139 Uttam Kumaran: I also want one thing that we heard in the retro is just sharing a little bit about
156 00:18:41.510 ⇒ 00:18:55.180 Uttam Kumaran: we’re working with and how it’s going. I know I don’t think there will ever be a point where everything here is is like amazing. We will always be. Catch it feeling like we’re catching up. We will always feel like there’s room for improvement.
157 00:18:55.590 ⇒ 00:19:06.429 Uttam Kumaran: But I do think it’s really helpful to both celebrate things we’re doing well, but also understand that there are tons of room for improvement when we’re when we’re working, when we’re talking about the work we’re delivering for clients.
158 00:19:06.985 ⇒ 00:19:14.579 Uttam Kumaran: for Urban SIM. This is one of our newest clients. We are just right now planning sort of the work that we’re gonna do over the next 6 months for them.
159 00:19:14.770 ⇒ 00:19:17.979 Uttam Kumaran: We’re great at planning. So in terms of just that
160 00:19:18.629 ⇒ 00:19:20.869 Uttam Kumaran: so I feel like that’s going well
161 00:19:21.000 ⇒ 00:19:29.480 Uttam Kumaran: in terms of planning and executing. We definitely have some work to do. I think Eden and and Javi both have a lot of upward mobility.
162 00:19:29.905 ⇒ 00:19:51.039 Uttam Kumaran: A lot of work has gotten done in just this week since Monday. I think we realigned a lot of things. The output rates on both clients are going well. Eden, we’re continuing to work on process from going from dashboard requirements to ae work back up to creating dashboards. And how do we move them? Also to a point where we’re proactively finding insights.
163 00:19:51.160 ⇒ 00:20:16.366 Uttam Kumaran: Javi. I think we’re still stuck in this like the dashboard world where we got a bunch of dashboard workout. We’re up for renewal next week. I think we did what we could this week to get out all the core dashboards. We’ll see. I I think we we did what we could this week. I don’t know what I I still think. We’re probably like 50%. If they’re gonna renew or not
164 00:20:17.110 ⇒ 00:20:21.940 Uttam Kumaran: so we’ll see how this goes. I I think we we came together a lot among the engineering team to get stuff done
165 00:20:22.361 ⇒ 00:20:29.959 Uttam Kumaran: but if they. If they renew, we definitely have to continue to change things and get this more on a place where we’re proactively giving them insights.
166 00:20:30.583 ⇒ 00:20:37.270 Uttam Kumaran: Cool parts. I think we did a better job this week of listening and getting requirements. However.
167 00:20:37.460 ⇒ 00:20:41.260 Uttam Kumaran: we really haven’t done much for them over the past 2, 3 months.
168 00:20:41.922 ⇒ 00:20:42.680 Uttam Kumaran: And I’m like.
169 00:20:42.820 ⇒ 00:21:06.720 Uttam Kumaran: I would say, like this again, this is our 1st client that the company has had. So they. We have a huge amount of trust with them, and so they don’t get nervous when we sort of slow down, because they’ve seen us grow. However, I’m I’m pretty disappointed in in us overall, and that we haven’t been able to move this from a place that we’re collecting requirements, executing and actively engaging them.
170 00:21:07.292 ⇒ 00:21:15.629 Uttam Kumaran: This is probably our toughest client. We have had almost 3 or 4 data people go through and churn out of this client.
171 00:21:17.200 ⇒ 00:21:43.159 Uttam Kumaran: I am, of course, the only this is the 1st client I brought into the company. So they really trust this, though, and pious and Bo are working on sort of how do we move them from this phase onto the phase where we’re actively delivering insights. I think we had a really good week and really good meeting with their CEO yesterday. We have some action items. I won’t flip this to. Okay, until I get a good sense of like, okay, we have a, we have a damn good process about meeting with them every week.
172 00:21:43.300 ⇒ 00:21:46.599 Uttam Kumaran: and sort of delivering. So we’ll see
173 00:21:46.870 ⇒ 00:21:52.139 Uttam Kumaran: on ABC. Home and stack Blitz. I am sort of probably the bottleneck on both of these.
174 00:21:54.150 ⇒ 00:22:00.069 Uttam Kumaran: I am pming both, and both, I think, need a lot more love and setting a 3 month roadmap
175 00:22:00.525 ⇒ 00:22:04.650 Uttam Kumaran: ABC, that AI team has crushed it in terms of deliverables, and a sort of
176 00:22:06.110 ⇒ 00:22:23.619 Uttam Kumaran: had my back, and sort of just like we deliver. So much for them that they sort of are looking past a little bit of the things we’re missing on the planning side on stack lists, though Ryan is really running this solo and I haven’t been much help in terms of helping on the development side, but also setting a forward roadmap.
177 00:22:24.166 ⇒ 00:22:32.820 Uttam Kumaran: Again. We are the kind of the overarching theme is, I’m basically pm, in all of these right now, and it’s a struggle.
178 00:22:33.330 ⇒ 00:22:53.979 Uttam Kumaran: My, my time goes to focus on the biggest fires. We had a huge fire on Eden and Javi, and that we’re sort of putting out stack. Blitz definitely needs some love to sort of tee this up for next week. We are bringing on a lead. Pm, that’s starting sort of part time next week.
179 00:22:54.398 ⇒ 00:23:11.179 Uttam Kumaran: On a on a trial basis, just to sort of understand if he can work with us, and we can work with him. His name is Steven. He’ll be here on Monday, and sort of joining some of the Monday meetings. So I’m very, very excited. As everybody on the data team and the AI team knows if I can go spend time
180 00:23:11.500 ⇒ 00:23:25.109 Uttam Kumaran: on engineering. And with engineers things move really quickly. And so that’s where I need to be. And so I’m very excited to offload some of that and and put put me in a better position to support everyone on the engineering side.
181 00:23:25.555 ⇒ 00:23:30.289 Uttam Kumaran: And then also begin to work on 3 month roadmaps for each of these clients.
182 00:23:30.440 ⇒ 00:23:31.730 Uttam Kumaran: The nice thing is.
183 00:23:32.400 ⇒ 00:23:53.620 Uttam Kumaran: I know this has been a little bit of a gloomy conversation so far. The great thing is even with the fact that we’re not 100 on each of these clients still love us and really appreciate the work we’re doing, which shows that the way we deliver, and the way we empathize with clients is still really, really strong. Of course our expectations are to be the best. And so, until we hit that
184 00:23:53.740 ⇒ 00:23:54.940 Uttam Kumaran: I’m not happy.
185 00:23:55.495 ⇒ 00:24:20.710 Uttam Kumaran: We I don’t. I don’t look left to right at other consultancies and other agencies on. If we’re doing good. I look at a couple of factors, our clients extremely extremely happy, almost like thinking that we’re cheap for what we deliver. Second, I look to see that we are constantly taking on more work, and we are able to raise our rates and raise the amount of money we can charge.
186 00:24:21.297 ⇒ 00:24:37.910 Uttam Kumaran: And I. Those are the really the things that I look at. We have had some wins on that for a couple of clients definitely. For a couple of these we could take on way more and are really, we’re held back by our our ability to plan and our ability to execute consistently.
187 00:24:38.418 ⇒ 00:25:04.659 Uttam Kumaran: So that’s the 4 1 1 on how we’re doing. Any questions there. I know this is probably the 1st time where we’ve done like a full overview of everything. This list will increase. We will be taking on more clients. I’ll talk a little bit how sales is going. But any questions here I sort of want to hear feedback from anybody on any of these accounts. If you disagree with me, if you agree with me, do you have ideas?
188 00:25:05.237 ⇒ 00:25:12.989 Uttam Kumaran: I’m just one opinion, I sort of have a different angle. I’m seeing sort of things from my level. But please, if you have any thoughts, let me know
189 00:25:22.450 ⇒ 00:25:34.509 Uttam Kumaran: in particular, I think I’ll probably call on yeah, like Miguel or Ryan, or any of the core, or away for any of the core engineers that are that are on or leading these clients. You guys have.
190 00:25:35.170 ⇒ 00:25:38.970 Miguel de Veyra: Mean, Frank, I think, if to say one more thing.
191 00:25:39.610 ⇒ 00:25:57.840 Uttam Kumaran: I don’t. I don’t mean this to. I’ll take kind of take the edge off. This conversation is, even if you’re on this account, and it’s a not good, or it’s okay. It’s not an attack on you. It’s not an attack. It’s a we’re all working on this together as a Brain Force team. However, my expectations should match your expectations, and
192 00:25:58.200 ⇒ 00:26:02.559 Uttam Kumaran: if we have a disagreement on the health, and I would love to have a discussion on that.
193 00:26:04.340 ⇒ 00:26:07.930 Miguel de Veyra: Yeah, I mean, ABC is pretty accurate, I guess.
194 00:26:09.230 ⇒ 00:26:16.170 Miguel de Veyra: we have, like we had like a couple of meetings with them this week, and most of the ball is on their hands. I would say.
195 00:26:16.750 ⇒ 00:26:19.199 Miguel de Veyra: right, so yeah.
196 00:26:19.740 ⇒ 00:26:24.860 Miguel de Veyra: But yeah, it’s pretty much going all according to plan for you.
197 00:26:24.860 ⇒ 00:26:29.510 Uttam Kumaran: You guys have done a good job at like filling like, just by just executing. We’ve saved a lot of.
198 00:26:29.990 ⇒ 00:26:32.349 Miguel de Veyra: Yeah, sometimes you just have to do. And.
199 00:26:32.350 ⇒ 00:26:35.460 Uttam Kumaran: But you saw today’s meeting. There’s still things that like I missed.
200 00:26:35.590 ⇒ 00:26:38.190 Uttam Kumaran: I should have asked them last week about some of those questions.
201 00:26:38.530 ⇒ 00:26:39.240 Miguel de Veyra: Yeah.
202 00:26:39.700 ⇒ 00:26:40.350 Uttam Kumaran: So.
203 00:26:42.370 ⇒ 00:26:43.250 Ryan: Yeah, in terms of.
204 00:26:43.250 ⇒ 00:26:44.210 Miguel de Veyra: What you can do.
205 00:26:45.920 ⇒ 00:26:47.199 Miguel de Veyra: Okay, sorry. Go ahead.
206 00:26:48.180 ⇒ 00:26:51.609 Ryan: Yeah, in terms of stack bits as well. I yeah. I agree with the
207 00:26:51.720 ⇒ 00:26:53.620 Ryan: the state current state like,
208 00:26:54.550 ⇒ 00:27:02.460 Ryan: there’s I would say, there’s some influx in like the need to create models for the bare metrics stuff.
209 00:27:02.640 ⇒ 00:27:03.800 Ryan: And like,
210 00:27:04.970 ⇒ 00:27:13.180 Ryan: yeah. So I’m I’m the only one doing it at the moment. And like, sometimes I’m I get stuck with like data validation. Like
211 00:27:13.780 ⇒ 00:27:22.440 Ryan: I I I, I create the models. But the data doesn’t match with the the ui, basically. So
212 00:27:22.700 ⇒ 00:27:29.549 Ryan: yeah, it would be great to have someone have us like a different point of view as well.
213 00:27:29.800 ⇒ 00:27:30.380 Uttam Kumaran: Yeah.
214 00:27:34.910 ⇒ 00:27:38.370 Uttam Kumaran: anything else from anyone who’s on one of these teams.
215 00:27:44.150 ⇒ 00:27:44.890 Uttam Kumaran: Cool.
216 00:27:47.100 ⇒ 00:27:53.950 Uttam Kumaran: Okay, let’s talk a little bit about some of our progress this week on our like okrs, I think, for folks
217 00:27:54.300 ⇒ 00:27:59.420 Uttam Kumaran: that are new. I think, Mike, you may be the only one who hasn’t seen this but we sort of have
218 00:27:59.780 ⇒ 00:28:04.110 Uttam Kumaran: our okrs, what we’re trying to. Yeah, Kyle, go ahead.
219 00:28:04.970 ⇒ 00:28:06.269 Uttam Kumaran: Oh, if you I don’t know if you’ve raised.
220 00:28:06.270 ⇒ 00:28:07.400 Caio Velasco: Don’t worry about.
221 00:28:07.710 ⇒ 00:28:24.069 Uttam Kumaran: Okay, yeah. So basically, our objective is to accelerate revenue goals. And we have a couple of key results increase in terms of increased. Mrr, I would say, this really goes to what I just talked about for our clients, which is
222 00:28:25.400 ⇒ 00:28:40.019 Uttam Kumaran: We have a few clients that we can increase our prices on. However, in order for Robert to own that conversation and go into that with tons of evidence to go for it, we need to execute right? So
223 00:28:40.750 ⇒ 00:28:43.769 Uttam Kumaran: at least we need to be at okay, if not great
224 00:28:44.080 ⇒ 00:28:53.299 Uttam Kumaran: for him to go. Say, Hey, we’re crushing it. We could take more. Do you want us to do more? And then for their answer to be like 100? Why didn’t you ask me sooner.
225 00:28:53.570 ⇒ 00:29:01.790 Uttam Kumaran: like I think we’re probably sitting on an extra 30 grand a month worth of those conversations that’s purely blocked by our ability to execute.
226 00:29:02.150 ⇒ 00:29:10.049 Uttam Kumaran: which is great, which which shows that, like okay, the money’s there, right? And so we just have to get closer to to it.
227 00:29:10.475 ⇒ 00:29:27.930 Uttam Kumaran: In terms of 400 k. Mr. And pipeline. This is also growing. So again, Robert is owning this. A lot of this has been unlocked by his time to go work on sales, so we do have pretty good sales process, not only on outbound. We have inbound. We have partnerships, so we’re working on this.
228 00:29:29.280 ⇒ 00:29:32.380 Uttam Kumaran: I feel pretty good about us getting this
229 00:29:32.640 ⇒ 00:29:34.809 Uttam Kumaran: up back to this level by March.
230 00:29:35.040 ⇒ 00:29:46.990 Uttam Kumaran: Pipeline revenue means that for companies that we’re that are currently a lead meaning we’ve either had an initial conversation with, or we are in proposal, for we have roughly 400 grand worth.
231 00:29:47.414 ⇒ 00:29:59.300 Uttam Kumaran: Of course what gets closed is will be a subset of this, but we always want to know that we’re in conversations to give you a sense of where we’re at. We send probably 6 proposals out this week.
232 00:29:59.908 ⇒ 00:30:06.249 Uttam Kumaran: Which is a huge week we need to. Our conversion rate on those proposals is probably anywhere from 5 to 10%
233 00:30:06.920 ⇒ 00:30:11.729 Uttam Kumaran: meaning for every 100 proposals we send out, we get 10 that close.
234 00:30:12.030 ⇒ 00:30:17.750 Uttam Kumaran: So that’s the thing I want us to keep in mind is these numbers need to be very high for it to actually come down to close
235 00:30:17.940 ⇒ 00:30:19.520 Uttam Kumaran: to close clients.
236 00:30:20.720 ⇒ 00:30:21.790 Uttam Kumaran: And so
237 00:30:22.180 ⇒ 00:30:30.009 Uttam Kumaran: really, this is a function of his time. But also all our engineering team’s ability to execute gives the sales team confidence to actually go get this
238 00:30:30.445 ⇒ 00:30:46.234 Uttam Kumaran: and then expanding to into legal services. Yeah, we we have 2 legal companies that are in proposal phases with. And we’re continuing to sort of talk to a couple of other like platform players. On, how do we implement their tools? This is our bet on the sales side.
239 00:30:46.760 ⇒ 00:30:50.800 Uttam Kumaran: for this quarter, which is, is legal, a good place for us to expand into
240 00:30:54.590 ⇒ 00:30:57.489 Uttam Kumaran: great let’s talk about
241 00:30:58.440 ⇒ 00:31:03.599 Uttam Kumaran: delightful service delivery. So one, every client needs to receive a high quality message. We did a
242 00:31:03.710 ⇒ 00:31:11.209 Uttam Kumaran: much better job this week of talking to almost every client every day. I think in particular, the ones that were
243 00:31:11.340 ⇒ 00:31:13.660 Uttam Kumaran: lagging the most were pool parts.
244 00:31:14.194 ⇒ 00:31:21.309 Uttam Kumaran: Javi Eden we talk to every day. Urban sends. I talk to every day. ABC, so for those 3,
245 00:31:22.240 ⇒ 00:31:37.910 Uttam Kumaran: we still have a little bit of room to grow where they get a message from us, an update or question every single day. That’s a good sort of volume barometer to indicate again, these are companies spending anywhere from 10 to 30 grand on us every month.
246 00:31:38.150 ⇒ 00:31:42.630 Uttam Kumaran: Imagine, if you, your company, that you worked really hard for.
247 00:31:42.750 ⇒ 00:31:46.310 Uttam Kumaran: and you were spending 30 grand on a consultant, and he didn’t message you for 2 days.
248 00:31:47.160 ⇒ 00:31:58.090 Uttam Kumaran: Not a good feeling. And so that’s what this really, this key result goes to. And we’re doing a much better job. I would probably change this to yellow. This is still a red.
249 00:31:58.320 ⇒ 00:32:04.070 Uttam Kumaran: I think my time spent on client work is probably honestly like 80%
250 00:32:04.220 ⇒ 00:32:06.299 Uttam Kumaran: and I think I’m probably like.
251 00:32:06.430 ⇒ 00:32:09.280 Uttam Kumaran: I would say, 9 to 5. It’s like a hundred percent.
252 00:32:10.990 ⇒ 00:32:20.940 Uttam Kumaran: meaning, I’m I’m just like we’re, I’m still working like 1215 h days. So this really needs to get affected. I think our higher on the Pm. Side is gonna help this a bunch.
253 00:32:21.100 ⇒ 00:32:27.280 Uttam Kumaran: And I think Damalade and some of our engineers continuing to level up and take on more work.
254 00:32:27.510 ⇒ 00:32:29.639 Uttam Kumaran: should continue to affect this.
255 00:32:30.379 ⇒ 00:32:36.269 Uttam Kumaran: Our ability to do everything perfectly hinges on this alone.
256 00:32:36.947 ⇒ 00:32:50.589 Uttam Kumaran: Right for for us to be able for me and Robert to be able to go outside and work on the machine, spend time with marketing, spend time with sales, spend time with operations. This is the core thing that needs to get affected. We are still not there right now.
257 00:32:51.166 ⇒ 00:33:11.040 Uttam Kumaran: Junior Pm. Agent is getting a lot better, though. I think the one place that we could totally improve this, though, is on its usage. I think we are still not 100 there on this getting leveraged in the best way. But I think Casey has a really great demo of how this is going. And sort of the impact we’re making there.
258 00:33:12.830 ⇒ 00:33:18.600 Uttam Kumaran: And then yes. So for Brand to content. So maybe, Hannah, I’ll just.
259 00:33:18.710 ⇒ 00:33:29.330 Uttam Kumaran: I can just steal your thunder and sort of walk through this really quickly. I would love for us to even share. If we have anything we want to share from the design team.
260 00:33:29.811 ⇒ 00:33:51.229 Uttam Kumaran: It’s always great to this stuff looks really really amazing. But 2 things. So one across the board on these 2 items, we’re working on a broad, content strategy sort of looking at everything we’ve done over the last year and deciding on couple of things. One. Which channels do we want to start putting content. On which do we want to turn off? We sort of came to a decision that we’re really just gonna focus on linkedin
261 00:33:51.949 ⇒ 00:34:16.780 Uttam Kumaran: and so we’re basically turning off X Instagram, and tick tock. The second piece is, we have a pretty cohesive. We have at least the workings of a strategy on how to get Robert and myself to post more and post more content on a recurring basis. Again, this is really a function of time and a function of like spending time on Linkedin, which again, I don’t
262 00:34:17.040 ⇒ 00:34:24.280 Uttam Kumaran: do much at all, maybe like once a day, or once every other day. So again I’ll go back to the previous result that
263 00:34:24.650 ⇒ 00:34:28.880 Uttam Kumaran: much of our of our time can get freed up. We can go work on these other initiatives.
264 00:34:29.256 ⇒ 00:34:42.069 Uttam Kumaran: But we’ve done a great job of finishing up some of our capabilities, deck some other assets. Hannah is leading meeting between the sales team, which is one of the clients of this org and sort of getting requirements from them on new decks or one pages that they need
265 00:34:42.480 ⇒ 00:34:55.840 Uttam Kumaran: and then, yeah, we’re basically gonna measure 2 things, one increasing our engagement, second, driving people to the website. And then 3, rd driving them to book calls. So I know Nico is owning sort of the measurement of post hog. And how do we actually measure
266 00:34:56.238 ⇒ 00:35:03.210 Uttam Kumaran: how people are coming onto the site and where they’re coming from, and then, of course, getting them to click and book a call with us.
267 00:35:05.610 ⇒ 00:35:18.519 Uttam Kumaran: anything I missed there ideally. I think we we were able to finish the content strategy, Doc, by early next week, and I would love to share that with the whole team. So everybody has a good idea of what? We’re what we’re working on here.
268 00:35:22.010 ⇒ 00:35:22.890 Uttam Kumaran: Cool?
269 00:35:23.522 ⇒ 00:35:29.979 Uttam Kumaran: Demos. Yeah, I guess. Nico, what’s the best way to to run this? Should I
270 00:35:31.100 ⇒ 00:35:33.509 Uttam Kumaran: press press play? Is there audio.
271 00:35:33.510 ⇒ 00:35:34.200 Nicolas Sucari: Yeah.
272 00:35:34.690 ⇒ 00:35:40.469 Nicolas Sucari: the the video has audio. So if you’re sharing with the audio there, we should be able to hear
273 00:35:40.770 ⇒ 00:35:42.980 Nicolas Sucari: what Casey recorded.
274 00:35:43.390 ⇒ 00:35:44.690 Nicolas Sucari: Yep, let’s try.
275 00:35:44.690 ⇒ 00:35:46.900 Uttam Kumaran: Can you just let me know if there’s audio?
276 00:35:47.100 ⇒ 00:35:49.030 Uttam Kumaran: And yeah, we just want to introduce.
277 00:35:49.860 ⇒ 00:35:58.149 Nicolas Sucari: Agent. So as the name the name suggests. So this is an AI agent specifically fed with context about
278 00:35:58.360 ⇒ 00:35:59.609 Nicolas Sucari: client Javi coffee.
279 00:36:00.750 ⇒ 00:36:04.940 Nicolas Sucari: So just to give some context on the idea behind this agent
280 00:36:06.140 ⇒ 00:36:08.239 Nicolas Sucari: know that we have a lot of client data.
281 00:36:08.340 ⇒ 00:36:14.400 Nicolas Sucari: So right now, it’s scattered across multiple platforms. So we have like data on the notion side.
282 00:36:14.969 ⇒ 00:36:19.210 Nicolas Sucari: And then we also have, like Zoom Meeting transcripts related to the client or with the client.
283 00:36:19.810 ⇒ 00:36:22.160 Nicolas Sucari: We have, of course, our slack communications.
284 00:36:22.540 ⇒ 00:36:28.909 Nicolas Sucari: And then we also have, like, our git repositories. So yeah, there’s a lot of data that we have.
285 00:36:29.510 ⇒ 00:36:38.819 Nicolas Sucari: And yeah, searching across all these platforms may not be the most efficient. And I guess sometimes you just want to.
286 00:36:39.010 ⇒ 00:36:43.719 Nicolas Sucari: you know, ask someone and ask someone and get an answer really quickly about something.
287 00:36:45.140 ⇒ 00:36:47.320 Nicolas Sucari: So I guess the next.
288 00:36:47.810 ⇒ 00:36:53.240 Nicolas Sucari: the next thing you would think about is okay. Maybe if you take some of these data. And just, you know.
289 00:36:53.380 ⇒ 00:37:07.439 Nicolas Sucari: manually paste them onto like chat, Gpt and query, or like a ask questions over the over those data. And yeah, that’s 1 way to do it, although still you have to do it. Do do things manually, and feed the context manually to chatgpt.
290 00:37:08.930 ⇒ 00:37:14.170 Nicolas Sucari: And that’s where the coffee Javi coffee client, hub, agent, comes into play.
291 00:37:14.660 ⇒ 00:37:16.470 Uttam Kumaran: That work did the full screen work.
292 00:37:19.370 ⇒ 00:37:22.209 Nicolas Sucari: Yeah, so let’s see the agent in action. So
293 00:37:22.588 ⇒ 00:37:26.459 Nicolas Sucari: so basically, how you start a chat with this agent is, you just tag it.
294 00:37:28.250 ⇒ 00:37:31.589 Sahana Asokan: So yeah, let me ask a basic question.
295 00:37:32.220 ⇒ 00:37:34.690 Sahana Asokan: So I’m gonna ask for their email addresses.
296 00:37:37.690 ⇒ 00:37:42.730 Sahana Asokan: And to make sure that the agent is receiving your messages.
297 00:37:43.130 ⇒ 00:37:45.440 Sahana Asokan: It’s going to leave a reaction.
298 00:37:46.040 ⇒ 00:37:48.590 Sahana Asokan: Okay? Great. So yeah, we have this.
299 00:37:49.990 ⇒ 00:37:53.470 Sahana Asokan: So we have these email addresses for the clients.
300 00:37:53.640 ⇒ 00:37:59.620 Sahana Asokan: So if you want to double check. Yeah, we have the same email addresses over here on notion.
301 00:38:00.160 ⇒ 00:38:05.610 Sahana Asokan: So right now let me try asking question the status of some tickets.
302 00:38:08.240 ⇒ 00:38:09.120 Uttam Kumaran: Let’s go.
303 00:38:09.900 ⇒ 00:38:15.170 Uttam Kumaran: So yeah, some responses will take some time due to the contact size.
304 00:38:17.590 ⇒ 00:38:21.559 Uttam Kumaran: But yeah, so here, for example. So yeah, it it
305 00:38:21.860 ⇒ 00:38:25.840 Uttam Kumaran: answered regarding the tickets that are blocked. So let’s double check.
306 00:38:26.400 ⇒ 00:38:32.300 Uttam Kumaran: Yeah. So this is one of the blocked tickets. And yeah, it also reads the status reason which is
307 00:38:32.630 ⇒ 00:38:35.649 Uttam Kumaran: blocked by portable team. No connector developed yet. Okay.
308 00:38:36.220 ⇒ 00:38:44.709 Uttam Kumaran: so let’s try to ask it a question on the Git Repository. So right now, it only has context on the Dbt project, folder or directory.
309 00:38:45.200 ⇒ 00:38:46.080 Uttam Kumaran: So, yeah.
310 00:38:50.670 ⇒ 00:38:56.749 Uttam Kumaran: okay, great. Yes. So we have. So the agent gave us this structure.
311 00:38:58.270 ⇒ 00:39:04.279 Uttam Kumaran: So yeah, this shows that we could ask questions on or yeah pertaining to the Git repository.
312 00:39:05.160 ⇒ 00:39:10.979 Uttam Kumaran: And I guess one last question to ask the agent. So let’s ask for a summary on the work we’ve done so far.
313 00:39:11.600 ⇒ 00:39:15.129 Uttam Kumaran: So yeah, let’s see how the bot is going to answer,
314 00:39:17.780 ⇒ 00:39:22.080 Uttam Kumaran: okay, so yeah, we have this pretty good summer. You.
315 00:39:23.840 ⇒ 00:39:24.469 Uttam Kumaran: So yeah.
316 00:39:24.470 ⇒ 00:39:29.670 Nicolas Sucari: Those are just some examples of the questions that we could ask the agent.
317 00:39:30.200 ⇒ 00:39:36.450 Nicolas Sucari: So yeah, definitely, not. There are some edge cases that we’ve not yet tested for. And
318 00:39:36.830 ⇒ 00:39:40.399 Nicolas Sucari: yeah, as for the next steps, definitely, more tests and
319 00:39:40.740 ⇒ 00:39:43.889 Nicolas Sucari: feedback from the team would be very much appreciated.
320 00:39:44.520 ⇒ 00:39:48.399 Nicolas Sucari: we want to make this work for the team. And yeah, so
321 00:39:48.860 ⇒ 00:39:55.459 Nicolas Sucari: would be glad if you could test it out, provide any feedback like missing missing data or hallucinations.
322 00:39:56.110 ⇒ 00:40:00.863 Nicolas Sucari: or like any feature requests that could help improve the experience definitely. Let us know
323 00:40:01.370 ⇒ 00:40:04.819 Nicolas Sucari: And also ideally, we want to expand this to other clients and
324 00:40:05.500 ⇒ 00:40:09.750 Nicolas Sucari: and maybe like, for example, whenever we have a new client we already have, like a client hub
325 00:40:10.140 ⇒ 00:40:12.970 Nicolas Sucari: set up for them. So yeah, that’s all.
326 00:40:16.230 ⇒ 00:40:17.419 Uttam Kumaran: Yeah, that’s amazing.
327 00:40:18.300 ⇒ 00:40:25.290 Uttam Kumaran: Where can people access this in slack right now?
328 00:40:26.580 ⇒ 00:40:29.210 Casie Aviles: Over at the AI Agents Channel.
329 00:40:31.350 ⇒ 00:40:33.060 Casie Aviles: Yeah, I added it. There.
330 00:40:36.580 ⇒ 00:40:40.010 Uttam Kumaran: I’m gonna I’m just gonna send the
331 00:40:40.870 ⇒ 00:40:43.190 Uttam Kumaran: the link to the brain forge team.
332 00:40:45.310 ⇒ 00:40:46.140 Uttam Kumaran: Yeah.
333 00:40:50.030 ⇒ 00:40:54.199 Nicolas Sucari: Yeah. In order to use the Javi coffee client. Hub, you need to tag
334 00:40:54.420 ⇒ 00:41:02.339 Nicolas Sucari: Jabbie coffee brain forge agent. I think it’s called Casey, and every time you need to ask a question you need to tag that
335 00:41:02.680 ⇒ 00:41:03.990 Nicolas Sucari: that bot.
336 00:41:06.070 ⇒ 00:41:06.750 Casie Aviles: Yes.
337 00:41:09.710 ⇒ 00:41:10.429 Uttam Kumaran: Yeah, I think.
338 00:41:10.430 ⇒ 00:41:11.759 Nicolas Sucari: And yeah, and and
339 00:41:12.050 ⇒ 00:41:21.970 Nicolas Sucari: yes, you, you just go like that and ask the question. But then, Casey, we’re just using the Dbt folder from. Git right. We’re not using any other folder yet.
340 00:41:22.570 ⇒ 00:41:25.369 Casie Aviles: Yeah, yeah, we’ve we’ve limited it to that for now.
341 00:41:27.010 ⇒ 00:41:37.090 Nicolas Sucari: That’s perfect. Yeah, if you try to ask something from the code or, yeah, ask some questions or a file or something, it should be able to answer that. So yeah, just test it out.
342 00:41:38.380 ⇒ 00:41:40.690 Nicolas Sucari: It will be great to have that feedback.
343 00:41:43.930 ⇒ 00:41:44.930 Uttam Kumaran: Yeah. Let’s see.
344 00:41:47.500 ⇒ 00:41:50.389 Nicolas Sucari: Live Demos. We love this, we love this. Yeah.
345 00:41:50.390 ⇒ 00:41:56.000 Uttam Kumaran: Yeah, if it works, this is really great.
346 00:41:59.110 ⇒ 00:42:00.569 Uttam Kumaran: Maybe it’ll say, No.
347 00:42:01.510 ⇒ 00:42:07.450 Uttam Kumaran: okay, let’s see. I’ll I’ll leave it, and we can come back and see whether it it’s right.
348 00:42:08.490 ⇒ 00:42:12.119 Uttam Kumaran: Okay, cool. What? What’s next on Demos, you know.
349 00:42:12.400 ⇒ 00:42:13.960 Uttam Kumaran: because, Kaya, if you want to go.
350 00:42:16.390 ⇒ 00:42:20.290 Caio Velasco: So, yeah, so I started
351 00:42:20.470 ⇒ 00:42:28.820 Caio Velasco: more or less 2 weeks ago. And I think my 1st test was to work on gorgeous and understand a bit of what is happening and
352 00:42:29.160 ⇒ 00:42:29.970 Caio Velasco: what
353 00:42:30.808 ⇒ 00:42:43.439 Caio Velasco: what we, what we need to do in terms of dashboarding metrics, and all of that. So in the beginning was, of course, quite difficult to me to understand everything, so it took took me some time to just
354 00:42:44.299 ⇒ 00:42:53.019 Caio Velasco: understand, like the database Dbt. Install it on my machine, and all like understand, like that state fraud, and how things are done.
355 00:42:53.538 ⇒ 00:42:58.470 Caio Velasco: But then was really good, because I I was able to learn like quite a lot
356 00:42:59.034 ⇒ 00:43:09.419 Caio Velasco: and well for this specific ticket. We had some business questions that usually come well come from the client or from the analyst.
357 00:43:09.560 ⇒ 00:43:21.409 Caio Velasco: and at least on my end. I felt the need to understand a little bit more about well, some details some specific terms, or terminology, or or things like that.
358 00:43:21.905 ⇒ 00:43:30.530 Caio Velasco: Then I think that moved also to to a bit of like some tabs on a spreadsheet that we’re using now.
359 00:43:31.031 ⇒ 00:43:52.180 Caio Velasco: It was guiding a bit of my work and and trying to be more specific about some of those terms. So I learned about macros ticket fields. And and also I had. I had work a bit with unstructured data. But also I I learned a bit about those things in in more details.
360 00:43:52.591 ⇒ 00:44:01.460 Caio Velasco: And at the end of the day I understood that most of the things were brought by portable we had already a database for that.
361 00:44:01.974 ⇒ 00:44:18.055 Caio Velasco: So I was working through them and understanding each of them, and well getting some new information also regarding probably a history that was not all there. For for example, the message tables or something.
362 00:44:18.882 ⇒ 00:44:26.440 Caio Velasco: And then along that process I moved to the Dbt and started also building the models over there, learning a bit more about data modeling
363 00:44:27.519 ⇒ 00:44:37.259 Caio Velasco: and while we have been having very good meetings in the a ae team, and I was learning a lot also to to
364 00:44:37.350 ⇒ 00:44:58.630 Caio Velasco: to make some changes and and put things in a in a in a sufficient way for the analysts, so that we don’t have to do more than we need or do less than we should so, to be honest. At the end of those 2 weeks I learned like quite a lot, and that made me happy because I like to be challenged. And and for me that was quite helpful.
365 00:44:59.066 ⇒ 00:45:11.139 Caio Velasco: And I think now we would we are. I just did. The Pr. Yesterday was approved today, and I think now we are waiting for for feedback. But it was was quite a a nice journey.
366 00:45:12.790 ⇒ 00:45:14.006 Uttam Kumaran: Yeah. Pass. Go ahead.
367 00:45:14.310 ⇒ 00:45:30.309 payas: Just want to give a huge shout out to like just like, really really well done here, like asking so many good questions about the requirements like in your 1st 2 weeks, getting a ticket for like, build a like, build a data like, clean up this data for a thing you’ve literally never heard of.
368 00:45:30.340 ⇒ 00:45:52.150 payas: It’s so challenging. And he, like asked so many good questions like was really detailed about it, built a good back and forth, and even like we’re like, he’s like, yeah, the organization is on a mess is a mess. And I’m like, I just like he ripped a spreadsheet to like, this is like a better way. We should do this in the 1st 2 weeks. Just like huge! Shout out Kyle. I know you like you take a lot of pride in your work and just let just know it’s appreciated.
369 00:45:52.930 ⇒ 00:45:55.029 Caio Velasco: Thank you. I really appreciate that. Thank you.
370 00:45:59.640 ⇒ 00:46:03.140 Uttam Kumaran: Great and yeah, Kyle’s proposal for the
371 00:46:03.866 ⇒ 00:46:12.069 Uttam Kumaran: dashboard to ae process. We’re that’s basically what we’re gonna start running with and deploying across all our clients so that didn’t exist.
372 00:46:12.250 ⇒ 00:46:16.320 Uttam Kumaran: And that’s like one of the big things we even talked about in the retro meeting. So
373 00:46:17.500 ⇒ 00:46:18.869 Uttam Kumaran: really excited for that.
374 00:46:21.370 ⇒ 00:46:21.929 Caio Velasco: Appreciate it.
375 00:46:21.930 ⇒ 00:46:22.460 Caio Velasco: Yes.
376 00:46:23.020 ⇒ 00:46:23.980 Uttam Kumaran: And then
377 00:46:24.680 ⇒ 00:46:44.420 Uttam Kumaran: I don’t know if Bo is on. I think he’s probably working on this actually, actively but maybe I will just present a little bit of his work here. So we are working well, primarily, Sahana and Bo are working on tableau dashboards for Eden. Basically a
378 00:46:46.020 ⇒ 00:46:47.650 Uttam Kumaran: basically a.
379 00:46:47.920 ⇒ 00:47:01.740 Uttam Kumaran: we’re moving them from looker studio, which is really really hard to develop on. And kind of a bad product overall to tableau for their executive reporting, as well as reporting across pharmacy, customer service, and a few other divisions.
380 00:47:02.309 ⇒ 00:47:23.740 Uttam Kumaran: I think it’s just helpful to kind of see what dashboarding work looks like for a lot of people that aren’t on the data side. I’m sort of stealing a little bit of his thunder but really one. The way we get dashboard requirements from clients is a set of questions. I want to understand Xyz about my team, our clients, our sales
381 00:47:24.478 ⇒ 00:47:28.420 Uttam Kumaran: and we take that, and we produce mockups.
382 00:47:28.630 ⇒ 00:47:32.059 Uttam Kumaran: I think maybe, Sahana, I’ll maybe just pull up your
383 00:47:32.610 ⇒ 00:47:38.480 Uttam Kumaran: dashboard, mock up for Eden because I do think that it’s pretty good
384 00:47:39.069 ⇒ 00:47:52.259 Uttam Kumaran: and maybe I’ll I’m I’m happy to walk through it unless you have any specific things to say. But this is a really great example of a dashboard mock up that we are doing. And I think I’ll let me pick up.
385 00:47:52.600 ⇒ 00:48:09.339 Uttam Kumaran: Yeah, this is a good example. I think hopefully, everyone can I can walk through is basically dashboard mockup is exactly what what it is and for the folks in the design side. This is like Lo-fi wireframes for dashboards. And basically dashboards go through
386 00:48:09.890 ⇒ 00:48:32.090 Uttam Kumaran: sort of a similar design phase. But then also an execution phase. And what we’re looking for from the analyst team is to tell for the A side to look at. What metrics do we need? What are the filters you guys need. What is the data marks that needs to be built? And so this is a dashboard on agent performance agent, performance and agent is a customer service agent. And they are using Zendesk which is their customer service software.
387 00:48:32.498 ⇒ 00:48:44.329 Uttam Kumaran: Basically, what we’re working on is taking a look at how the health of that agent performance is for this business. How many tickets are assigned? What is the resolution? Time? How many tickets are escalated?
388 00:48:44.748 ⇒ 00:49:12.799 Uttam Kumaran: Scorecards for each agent performance? Metrics across the entire team. And so Sahana sort of works with the business teams to sort of get approval on, hey? This has everything we need to work on. Once this is confirmed, this gets passed off and sort of built with the Ae team on what metrics are needed. So what data do we need to get from Zendesk? How does that need to get modeled? And then, ultimately, where is that available in the data warehouse to query.
389 00:49:13.120 ⇒ 00:49:32.300 Uttam Kumaran: so the for the AI team, they’ve been doing a lot of stuff in Snowflake recently. This is basically the process of what data needs to land in snowflake. How does it get queried, and what metrics need to be made available? So we are most likely gonna go through with a mock up process for any future dashboard that we do but again, like
390 00:49:32.300 ⇒ 00:49:44.879 Uttam Kumaran: shout out to Sahana for setting this stage again, we’ll take this and sort of build a more robust process to do across all dashboards. And I, if I heard today, the feedback was really good from the client. So that’s
391 00:49:45.100 ⇒ 00:49:46.000 Uttam Kumaran: that’s awesome.
392 00:49:48.755 ⇒ 00:49:55.139 Uttam Kumaran: Cool, I think that’s it. Any other topics?
393 00:49:56.018 ⇒ 00:49:59.420 Uttam Kumaran: Anything else that we wanted to talk about today?
394 00:50:02.850 ⇒ 00:50:04.546 Uttam Kumaran: I feel like, maybe
395 00:50:05.280 ⇒ 00:50:19.350 Uttam Kumaran: Hannah and team next week we could do something around the contents strategy. And then also any new items from design. I think it’s always great to you guys produce a really great work. So it’s always awesome to see that and then, yeah, please, please,
396 00:50:19.900 ⇒ 00:50:22.809 Uttam Kumaran: use the agent in slack.
397 00:50:23.436 ⇒ 00:50:28.030 Uttam Kumaran: The only way that has feedback is if we’re able to actually get real questions.
398 00:50:28.465 ⇒ 00:50:32.560 Uttam Kumaran: I I haven’t been using it either. So I’m at fault for that as well. But.
399 00:50:33.201 ⇒ 00:50:38.479 Uttam Kumaran: we want to start picking off the low, easy questions that we would have maybe
400 00:50:38.610 ⇒ 00:50:45.621 Uttam Kumaran: looked and spent 30 min looking for, or asked one another, and start hitting the agent with those questions.
401 00:50:46.280 ⇒ 00:50:49.590 Uttam Kumaran: probably the way we’re gonna yeah, go ahead.
402 00:50:49.590 ⇒ 00:50:53.029 Nicolas Sucari: Go check the response to them the questions that you already did. It’s there.
403 00:50:53.030 ⇒ 00:50:53.520 Uttam Kumaran: Okay.
404 00:50:53.520 ⇒ 00:51:03.719 payas: I was. Gonna say, dude like this is actually like one of the most insane things I’ve ever seen, like no joke like my day job might buy this like I’m not even joking like I’d love to like.
405 00:51:04.010 ⇒ 00:51:04.510 Uttam Kumaran: Yeah. Nobody.
406 00:51:04.510 ⇒ 00:51:04.840 payas: I don’t know.
407 00:51:04.840 ⇒ 00:51:05.370 payas: But you guys
408 00:51:05.370 ⇒ 00:51:14.730 payas: built here. This itself is like you could sell this, and it’s like millions to be made sorry. I was like, kind of Casey, this is you, Casey and Nico, you guys.
409 00:51:15.910 ⇒ 00:51:17.440 Nicolas Sucari: Okay, it’s okay for you. Yeah.
410 00:51:17.440 ⇒ 00:51:20.190 payas: Dude, amazing work. Dude what the fuck.
411 00:51:20.190 ⇒ 00:51:20.589 Ryan: Like you said.
412 00:51:20.590 ⇒ 00:51:22.579 payas: Mind is blown like I’m I’m.
413 00:51:22.580 ⇒ 00:51:26.860 Uttam Kumaran: Yeah, we presented it. Nobody said anything. And I’m like dude. We’ve been working on this for like.
414 00:51:27.230 ⇒ 00:51:27.800 payas: Yeah. But like
415 00:51:28.330 ⇒ 00:51:40.789 payas: you gotta see the magic for yourself, you know. Like, give it a rip, give it a rip, and you’ll see the magic for yourself. I mean, if you work on Javi, and you’re like, I literally would have spent like 2020 min in Snowflake to do that. And now I don’t have to.
416 00:51:41.810 ⇒ 00:51:49.289 michael weinberg: I love that it. It tells me like it’s using a hash function for a particular column. I I wasn’t expecting that much detail.
417 00:51:52.380 ⇒ 00:52:06.799 Uttam Kumaran: Okay, yeah. Every. No. Everybody stayed silent. And I was like, okay, is this not impressive? Because this is insane, like, we’ve been working on this, for we’ve been working on some version of this. In fact, Miguel’s 1st task when he came on was to build this a version of this
418 00:52:07.275 ⇒ 00:52:10.719 Uttam Kumaran: and so we’ve been working on this some version of this since September.
419 00:52:11.322 ⇒ 00:52:14.250 Uttam Kumaran: On and off, on and off, as we’ve gotten clients.
420 00:52:14.380 ⇒ 00:52:23.869 Uttam Kumaran: This is the best we. This is the best sort of version of this, where we’ve shoving everything we have about a client and using a lot of the AI techniques. We’ve learned.
421 00:52:25.220 ⇒ 00:52:37.919 Uttam Kumaran: and yeah, I mean, I think this is insane. I’ve never heard of a company that’s doing this right now and built it from scratch like there are some tools getting sold that could do this. But to give you a sense, this probably cost us like
422 00:52:38.130 ⇒ 00:52:42.539 Uttam Kumaran: several cents, maybe like one or 2 cents, or maybe a fraction of a cent
423 00:52:42.660 ⇒ 00:52:44.840 Uttam Kumaran: like we built this all in house.
424 00:52:44.840 ⇒ 00:52:48.529 payas: And it. And it’s connected to Snowflake. Is that is it just directly connected to Snowflake? And it’s.
425 00:52:48.530 ⇒ 00:52:56.700 Uttam Kumaran: No, this, this just it just infers this from it’s inferring this from the repo and documentation we have about code
426 00:52:57.215 ⇒ 00:53:00.360 Uttam Kumaran: level we could do is probably give it like the top.
427 00:53:00.850 ⇒ 00:53:02.820 Uttam Kumaran: 20 rows from each table, but
428 00:53:03.170 ⇒ 00:53:08.400 Uttam Kumaran: it’s purely bit built on our documentation, and and it’s inferring a lot from the repo.
429 00:53:08.620 ⇒ 00:53:10.960 michael weinberg: Can I? Can I ask a question about this.
430 00:53:10.960 ⇒ 00:53:13.910 Uttam Kumaran: Yeah, please, and ask Casey, because that’s it.
431 00:53:14.520 ⇒ 00:53:19.090 michael weinberg: Yeah. This question is for sure, for Casey, not for you, Tom. Come on.
432 00:53:19.662 ⇒ 00:53:26.337 michael weinberg: okay. So where it says, int review tags, it says, this model extracts tags associated with reviews.
433 00:53:27.410 ⇒ 00:53:35.380 michael weinberg: I’m wondering, is it? Is that? Is it getting that from a doc string? Or is it able to understand the abstraction of what the code is doing.
434 00:53:37.753 ⇒ 00:53:39.769 Casie Aviles: I’m not really sure yet.
435 00:53:39.770 ⇒ 00:53:41.860 Uttam Kumaran: This may be a question for me. It is.
436 00:53:41.860 ⇒ 00:53:44.170 michael weinberg: Oh, I see. Cause you’re using it. Okay, that’s.
437 00:53:44.170 ⇒ 00:53:46.540 Ryan: Well, I thought it gets in the sequel.
438 00:53:46.540 ⇒ 00:53:47.689 Uttam Kumaran: It’s reading the sequel.
439 00:53:47.910 ⇒ 00:53:48.580 Ryan: Yeah.
440 00:53:48.730 ⇒ 00:53:55.600 Uttam Kumaran: And it’s inferring. And then it’s also taking our notion. So in case we got requirements from the client about this.
441 00:53:55.770 ⇒ 00:54:03.090 Uttam Kumaran: then it’ll take that and sort of build it again. The equivalent is like, let’s say you were able to copy, paste all this into a chat. Gpt, prompt.
442 00:54:03.230 ⇒ 00:54:09.130 Uttam Kumaran: You’d be like, okay, it’s working. We’re doing that automatically on a recurring basis
443 00:54:09.500 ⇒ 00:54:20.320 Uttam Kumaran: with everything. So Zoom Meetings. So if we talked about okendo in a Zoom Meeting, that’s in knowledge. It has a repo. It has all the messages we’ve ever messaged in slack.
444 00:54:20.580 ⇒ 00:54:26.820 Uttam Kumaran: So it’s gonna be able to infer most. It doesn’t actually even need to know the table.
445 00:54:27.260 ⇒ 00:54:29.150 Uttam Kumaran: Really, you know, I’m like.
446 00:54:29.150 ⇒ 00:54:34.660 michael weinberg: I’m holding my breath a little bit because I’m like, okay, like the
447 00:54:34.910 ⇒ 00:54:37.110 michael weinberg: let’s say you sent a person to do this.
448 00:54:37.460 ⇒ 00:54:43.299 michael weinberg: Please answer this question, and you set an expectation of this level of detail for starters
449 00:54:43.680 ⇒ 00:54:50.271 michael weinberg: actually just parsing through stuff and collating. It is like it’s like an hour of work.
450 00:54:50.840 ⇒ 00:54:57.970 michael weinberg: But then, maybe, like, you would think like, Okay, well, the advantage of a human is that they can
451 00:54:58.413 ⇒ 00:55:06.759 michael weinberg: consolidate the information from the various sources. So there’s a lot of content in the sequel. There’s also a lot of content in like notion.
452 00:55:06.870 ⇒ 00:55:13.979 michael weinberg: And, like, you know, like a dumb AI bot like is gonna just try and spit back what it found as is.
453 00:55:14.460 ⇒ 00:55:32.889 michael weinberg: But since all of this is a single piece of context, if it’s able to understand the relationships between those 2 things. So, for example, if it answered that question on the basis of both the sequel and what it found in notion in one shot while doing that for all of these other fields and doing all of those
454 00:55:33.520 ⇒ 00:55:41.459 michael weinberg: that like that actually is like a really big deal. So I’m holding my because I’m like, I want to see where it hallucinates and doesn’t. But already this is this is wild.
455 00:55:41.460 ⇒ 00:55:50.230 Uttam Kumaran: Yeah, to think about like, kind of probably how that’s happening. We use ginger templating for Dbt, so it has references. It also, metadata has the table name.
456 00:55:51.020 ⇒ 00:56:02.239 Uttam Kumaran: We haven’t even given it like the the get, like, commit history. So we’ll just keep layering on more and more stuff. Basically, ideally. Also, this is, gonna have all the meetings we’ve had.
457 00:56:02.872 ⇒ 00:56:08.720 Uttam Kumaran: And so I want this to take care of 20% of questions like point blank
458 00:56:09.358 ⇒ 00:56:18.289 Uttam Kumaran: and like anytime, anyone’s awake wherever they can answer what what was talked about. What’s the work? Where is it? Right? And so?
459 00:56:18.960 ⇒ 00:56:26.720 Uttam Kumaran: And then we’re gonna basically, as soon as we nail it for Javi, it’ll take like, basically an hour to turn this on for every client. Basically.
460 00:56:28.900 ⇒ 00:56:34.840 Uttam Kumaran: So the more structured our work. The more documentation we have on how we do things. Playbooks
461 00:56:35.000 ⇒ 00:57:02.729 Uttam Kumaran: so ultimately, like. Part of the reason we use notion, in fact, is because I wanted to make a really sincere decision on where documentation lives. If it lived in Google, docs and notion and other things, we could have never accomplished this. Everything is in notion, and everything is in Github, and everything’s on zoom, and everything’s in slack. So we have 4 core sources which allows us to really do this very intentional decision. By the way, to like, do that for this reason alone, sort of thought when we
462 00:57:02.910 ⇒ 00:57:09.379 Uttam Kumaran: last summer that, like something like this, was possible. It’s taken us longer than I hoped, but
463 00:57:09.820 ⇒ 00:57:22.760 Uttam Kumaran: we had to build. We have to get really Rock Star AI people, because I couldn’t. I don’t know how to do this end to end. And I don’t have the time. So the AI team really took a really complicated problem. This also requires knowledge of the data side
464 00:57:23.010 ⇒ 00:57:30.510 Uttam Kumaran: to do this right. And you gotta think the AI team. They’re not data engineers, right? And so they’ve had to empathize and learn a lot about what we do.
465 00:57:30.810 ⇒ 00:57:38.390 Uttam Kumaran: But I agree, like, I think this is, this is really really incredible. So a lot of culmination from the last 6 months to kind of get to this point.
466 00:57:41.510 ⇒ 00:57:44.965 Uttam Kumaran: But great. Okay, Nico, thanks for telling me to go back.
467 00:57:45.280 ⇒ 00:57:45.950 Nicolas Sucari: Yeah.
468 00:57:45.950 ⇒ 00:57:47.839 Uttam Kumaran: I wouldn’t have had this moment otherwise.
469 00:57:49.330 ⇒ 00:57:59.300 Nicolas Sucari: Yeah. And you can. You can keep asking questions. It’s really good. You can ask, what Michael is saying, like, you can ask, what? What are the sources? What that this data is coming from? And it should be able to
470 00:57:59.470 ⇒ 00:58:01.649 Nicolas Sucari: to let you know. So yeah, it’s pretty cool.
471 00:58:04.860 ⇒ 00:58:08.959 Uttam Kumaran: Cool. And then, yeah, if you have a if you have an AI bot that you want built
472 00:58:09.120 ⇒ 00:58:12.070 Uttam Kumaran: 1 800 brain forge AI.
473 00:58:12.430 ⇒ 00:58:12.930 Nicolas Sucari: Yes.
474 00:58:13.430 ⇒ 00:58:14.429 Uttam Kumaran: Mentioned it in the AI.
475 00:58:14.430 ⇒ 00:58:14.830 Nicolas Sucari: Haven’t.
476 00:58:14.830 ⇒ 00:58:16.749 Uttam Kumaran: Channel. I’m currently.
477 00:58:16.750 ⇒ 00:58:17.380 Nicolas Sucari: Showing you.
478 00:58:17.380 ⇒ 00:58:44.710 Uttam Kumaran: Yeah, I’m currently the one telling all these ideas. But another piece I want to work on is how everybody in the team can leverage AI more. And our AI team is here to not only serve clients but to serve us. Similarly, we’re doing data stuff on our own data. So we. This is where we work on the machine, you know. And so any ideas anyone has even like, if you have ideas on how to work on this like, if you want to go build your own agents. We have all the tools we have, all the infra.
479 00:58:45.130 ⇒ 00:58:51.100 Uttam Kumaran: We’ll pay for it like it’s all out there, so don’t feel overwhelmed. In fact.
480 00:58:51.240 ⇒ 00:59:00.109 Uttam Kumaran: if you ask anyone on the AI team, we’ve all sort of started doing this just over the last 2 years, and mainly I would say, I only did this over the last like year. So
481 00:59:00.350 ⇒ 00:59:04.339 Uttam Kumaran: it’s a there’s definitely a lot to lot of value here.
482 00:59:05.450 ⇒ 00:59:14.559 Nicolas Sucari: Yeah, also what I think. It’s a good time to mention that we have all other agents in there that we can use. We have ticket here. We have lead researcher.
483 00:59:15.127 ⇒ 00:59:23.450 Nicolas Sucari: Maybe we should do like an AI demo on all of the agents that we have at some point so that everyone knows them and starts using them.
484 00:59:23.610 ⇒ 00:59:28.989 Uttam Kumaran: Yeah, because I don’t think yeah, definitely next week to do and do some sort of office hours on
485 00:59:29.220 ⇒ 00:59:33.519 Uttam Kumaran: folks that want to build more agents, and how to actually deploy that in slack and things like that.
486 00:59:34.900 ⇒ 00:59:35.470 Nicolas Sucari: Yeah.
487 00:59:36.460 ⇒ 00:59:37.070 Uttam Kumaran: Okay.
488 00:59:38.290 ⇒ 00:59:44.009 Uttam Kumaran: okay, cool. Alright. We needed the whole hour. I thought, this is gonna end early. This is really great. So thanks everyone for
489 00:59:44.642 ⇒ 00:59:55.560 Uttam Kumaran: the feedback. Thanks for listening to my gloomy thing about clients, what? We’re getting better. But we need to be honest about where we are. And then, yeah, to Kyle.
490 00:59:56.009 ⇒ 01:00:01.040 Uttam Kumaran: And I know Beau is in here. But Sahana, for present, for letting me present our work.
491 01:00:01.210 ⇒ 01:00:14.010 Uttam Kumaran: and to Casey for this awesome work in the AI team. So appreciate it. And then, yeah, we’ll have a team meeting again on Monday. If any questions I’ll be on rest of the day today. Just please message me. And yeah.
492 01:00:14.440 ⇒ 01:00:15.719 Uttam Kumaran: thanks for taking the time.
493 01:00:16.880 ⇒ 01:00:17.950 Miguel de Veyra: Thanks. Everyone have a.
494 01:00:17.950 ⇒ 01:00:18.580 Caio Velasco: Thank you.
495 01:00:18.580 ⇒ 01:00:19.939 Uttam Kumaran: Thank you. Guys. Thank you.