Meeting Title: L&D Check-Ins Date: 2026-04-22 Meeting participants: Brylle Girang, Uttam Kumaran, Mustafa Raja, Amber Lin
WEBVTT
1 00:00:09.560 ⇒ 00:00:10.660 Brylle Girang: Hello!
2 00:00:11.170 ⇒ 00:00:12.150 Uttam Kumaran: Ayy.
3 00:00:29.470 ⇒ 00:00:31.060 Brylle Girang: Hi, Amber. Hi, Mustafo.
4 00:00:32.080 ⇒ 00:00:33.080 Amber Lin: Hi there!
5 00:00:36.740 ⇒ 00:00:49.780 Brylle Girang: Alright, so, I just wanted to get a call with the team, and then just try to unblock some blockers, and also just, just, just going to walk you through the courses, and explain why we’re not, we’re not launching it yet, but…
6 00:00:50.150 ⇒ 00:00:57.019 Brylle Girang: I think one thing that I wanted to get out of the way immediately is when it comes to, like, the Snowflake certifications.
7 00:00:57.330 ⇒ 00:00:59.370 Brylle Girang: Mustafa, let me know if…
8 00:00:59.540 ⇒ 00:01:14.410 Brylle Girang: There’s something wrong with what I’m going to say, but Otam, right now, the Snowflake certifications need the examinees to, like, pay up front before going through the exam. That’s $175 USD payment.
9 00:01:14.410 ⇒ 00:01:14.810 Uttam Kumaran: Okay.
10 00:01:14.810 ⇒ 00:01:21.010 Brylle Girang: My question is… I don’t feel comfortable, like, asking them to, like, bathe first.
11 00:01:21.570 ⇒ 00:01:25.770 Uttam Kumaran: Okay, I’ll pay, but I… but actually, like, this is where… let me rephrase it, like.
12 00:01:26.400 ⇒ 00:01:32.939 Uttam Kumaran: I think what would be helpful, one, you could just… we don’t have to wait for this meeting, you could just tell me in Slack. Second.
13 00:01:33.190 ⇒ 00:01:36.279 Uttam Kumaran: Like, you could just be like, hey, they’re asking us to pay up front.
14 00:01:36.420 ⇒ 00:01:50.279 Uttam Kumaran: can you pay up front, or can you give me a solution? I’m gonna just message Snowflake, and I’ll get us some vouchers. Or we’ll pay for it. So, if you need that today, just create a ticket for the ops team.
15 00:01:50.530 ⇒ 00:01:52.130 Uttam Kumaran: And then me and Rico will…
16 00:01:52.500 ⇒ 00:01:55.670 Uttam Kumaran: We’ll handle it. If you can also draft me…
17 00:01:55.830 ⇒ 00:02:03.950 Uttam Kumaran: a message, like, an email, that’s just like, hey, these people on my team are going for Snowflake certs.
18 00:02:04.270 ⇒ 00:02:10.280 Uttam Kumaran: Can I get… can we inquire about free, vouchers?
19 00:02:10.490 ⇒ 00:02:14.650 Uttam Kumaran: I’ll send that to, like, Snowflake’s partner team today.
20 00:02:14.950 ⇒ 00:02:18.650 Brylle Girang: Okay, that was not, like, in my mind, but thank you.
21 00:02:18.650 ⇒ 00:02:21.540 Uttam Kumaran: Yeah, that’s why, like, don’t think about the solution.
22 00:02:21.960 ⇒ 00:02:27.110 Uttam Kumaran: I have, like, 100,000 solutions for you. Think about, just… give me the problem.
23 00:02:27.110 ⇒ 00:02:27.480 Brylle Girang: Yeah.
24 00:02:27.480 ⇒ 00:02:28.200 Uttam Kumaran: And… Yeah.
25 00:02:28.550 ⇒ 00:02:36.589 Uttam Kumaran: Yeah, send it to me in Slack, because I’ll just… I’ll solve this, so, yeah. Okay, great. Yeah, I don’t want anybody paying for certs, so…
26 00:02:36.750 ⇒ 00:02:41.029 Uttam Kumaran: even if Snowflake doesn’t come through, I’ll pay for it. Yeah.
27 00:02:41.590 ⇒ 00:02:52.369 Brylle Girang: Okay. Alright, Mustafa, is there anything else? Like, I have updated the certification roadmap for Snowflake, so the blocker is out, we should be able to, like, push them to proceed there.
28 00:02:52.370 ⇒ 00:02:57.520 Mustafa Raja: Yeah, then I’m going to take a look at the roadmap itself.
29 00:02:57.730 ⇒ 00:03:01.810 Mustafa Raja: If it looks good to me, I’m going to start pushing people.
30 00:03:02.670 ⇒ 00:03:06.810 Brylle Girang: Okay, perfect. Yeah, I’m just… then just going through…
31 00:03:07.500 ⇒ 00:03:19.919 Brylle Girang: the current state of our LMS right now. I’m calling this the LMS, but that’s to be finalized. So I think, Amber, one of your main curiosities is when it comes to, like, the foundation course.
32 00:03:19.920 ⇒ 00:03:30.419 Brylle Girang: And over the course of the few weeks, past few weeks, there’s no longer a foundation course, and I have divided it into two main courses that everyone should take.
33 00:03:30.440 ⇒ 00:03:39.359 Brylle Girang: There’s a quick start course, which would focus more on, like, onboarding the new hires, what do they need to do and follow in order to, like.
34 00:03:39.530 ⇒ 00:03:52.410 Brylle Girang: meet the baseline for people on their week one, on their day one. And then there’s an AI accelerator course which focuses more on how AI can help them meet their standards.
35 00:03:52.570 ⇒ 00:03:59.310 Brylle Girang: They should be launched this week, and the main… the main blocker here is I want to make sure that everyone gets to
36 00:03:59.500 ⇒ 00:04:02.169 Brylle Girang: Set up their open code first.
37 00:04:02.290 ⇒ 00:04:04.090 Brylle Girang: And not try to, like.
38 00:04:04.240 ⇒ 00:04:09.540 Brylle Girang: barrage them with so many… with so many tasks to do when it comes to L&D, but…
39 00:04:09.940 ⇒ 00:04:17.099 Brylle Girang: Since the foundation module is basically divided into two pieces, our OKR should also be updated.
40 00:04:17.339 ⇒ 00:04:25.780 Brylle Girang: And I think we’re going to focus on the AI Accelerator course since then, because this gives us a better way to quantify this.
41 00:04:26.800 ⇒ 00:04:28.270 Brylle Girang: The… is that clear?
42 00:04:29.360 ⇒ 00:04:40.169 Amber Lin: Yeah, once we launch it, I’ll ask my team to complete it, but, I know you say we’re launching this week, so I’ll wait for that.
43 00:04:43.320 ⇒ 00:05:02.090 Brylle Girang: Okay, I think the main blocker that I’m solving here is when it comes to, like, these canonical terms, like cursor automations, we’re moving out of cursor, so we’re making sure that we phrase this in a way that’s going to be more scalable as we change tools, and I don’t want, like, to confuse people, like, which tool should I use, etc.
44 00:05:02.580 ⇒ 00:05:12.599 Brylle Girang: So here is the LMS. Also, here are, like, the enablement guides. Here is where we’re going to dump things. My main ask for this team is that
45 00:05:12.750 ⇒ 00:05:19.990 Brylle Girang: Me, myself, I am fairly limited when it comes to building enablement guides for service line-specific stuff.
46 00:05:20.110 ⇒ 00:05:28.599 Brylle Girang: I guess one example that would be amazing here are the skills that you have built, Amber, when it comes to, like, the golden data sets, etc.
47 00:05:28.660 ⇒ 00:05:40.180 Brylle Girang: If I had your expertise, I would create those guides, but since I do not, I would be leading more on you into, like, helping me build this stuff out and posting it here.
48 00:05:40.280 ⇒ 00:05:56.709 Brylle Girang: one thing that I wanted to avoid is just us sending out announcements or messages to Slack, and then no way to, like, document it for future people, or people that’s going to join, and then just have one page that they can see what… whatever we have built in the past.
49 00:05:56.990 ⇒ 00:05:58.990 Brylle Girang: I…
50 00:05:59.640 ⇒ 00:06:08.649 Brylle Girang: Right now, I think we have all the standards when it comes to, like, building these guides, building courses, building everything when it comes to the LMS.
51 00:06:08.770 ⇒ 00:06:16.779 Brylle Girang: I would love you to, like, go through open code or cursor, try to build enablement guides for your skills that you have just shipped.
52 00:06:16.940 ⇒ 00:06:20.390 Brylle Girang: And then we can review and launch it with the team.
53 00:06:20.500 ⇒ 00:06:25.929 Brylle Girang: And that’s going to be the same for data and AI, Mustafa.
54 00:06:26.820 ⇒ 00:06:28.169 Amber Lin: Sounds good.
55 00:06:29.650 ⇒ 00:06:34.689 Brylle Girang: Okay, perfect. And then, lastly, the certifications. So…
56 00:06:34.940 ⇒ 00:06:44.699 Brylle Girang: still Omni, Snowflake, and then we’re going to be adding DPT here, but before DPT, we need to make sure that everyone’s alright with Omni and Snowflake first.
57 00:06:44.810 ⇒ 00:06:45.650 Brylle Girang: Yeah.
58 00:06:45.650 ⇒ 00:06:54.890 Amber Lin: For Omni, most people did the exam. I need to check in with Eva and Jasmine, but then we need to do the demo. Once the demo’s done, we’re done.
59 00:06:56.090 ⇒ 00:07:01.219 Brylle Girang: Yeah, and, like, is there a blocker for them when it comes to, like, making the demos as app?
60 00:07:01.470 ⇒ 00:07:04.660 Amber Lin: Well…
61 00:07:05.380 ⇒ 00:07:22.080 Amber Lin: Not really… I’ll check in. For me, I need to hear back to get the demo submission link, but otherwise, I think all the other folks have heard back, and they just haven’t done the demo yet, I think? So, I’m gonna nudge them today.
62 00:07:22.280 ⇒ 00:07:27.489 Brylle Girang: Okay, and just to confirm, have you heard back from Jasmine on her certification?
63 00:07:27.790 ⇒ 00:07:30.410 Amber Lin: No, I’m gonna check in.
64 00:07:31.200 ⇒ 00:07:32.680 Brylle Girang: Gotcha, okay.
65 00:07:32.680 ⇒ 00:07:52.209 Uttam Kumaran: Give her until, like, give her until Friday, Amber. I mean, she’s probably slammed, but I think this is where, like, part of… you guys should try other tools versus, like, hey, where are we? Like, hey, do we want to do, like, a working session on Friday? Like, Friday afternoon? Like, you know, I think…
66 00:07:52.770 ⇒ 00:07:58.040 Uttam Kumaran: One tool, of course, is, like, where is this? But see if you guys can try other…
67 00:07:58.440 ⇒ 00:08:05.990 Uttam Kumaran: you know, other opportunities to help folks, you know? It’s like, we’re almost like… School counselors, you know?
68 00:08:08.810 ⇒ 00:08:10.470 Brylle Girang: Okay, and then,
69 00:08:10.580 ⇒ 00:08:28.830 Brylle Girang: since that’s only Omni platform, Amber, once everyone’s all good here, we will be moving into Omni Quick Start for select folks. Like, this doesn’t need to be, like, everyone, and I’m guessing for strategy, this would mainly be Greg and Jasmine? Is that correct, Otam?
70 00:08:30.730 ⇒ 00:08:33.049 Uttam Kumaran: For which one?
71 00:08:33.059 ⇒ 00:08:38.069 Brylle Girang: For a quick start, so this focuses more on, like, training people on how to use Omni, right?
72 00:08:40.330 ⇒ 00:08:45.679 Uttam Kumaran: Yeah, I mean, I would… any… as many people as want to go for it, I’m game.
73 00:08:46.050 ⇒ 00:08:46.820 Brylle Girang: Okay.
74 00:08:46.820 ⇒ 00:08:52.949 Uttam Kumaran: It’s heavier, But I would prefer at least Greg and Jasmine and Robert
75 00:08:53.680 ⇒ 00:08:55.850 Uttam Kumaran: Or at least Greg and Jasmine, but, like.
76 00:08:56.260 ⇒ 00:09:00.809 Uttam Kumaran: I don’t know, I think Amber, Advait, you guys can go for it, too. It’s not gonna hurt.
77 00:09:02.360 ⇒ 00:09:07.039 Uttam Kumaran: It’s heavier, though, but I, like, I would love if you guys want to go for it as well.
78 00:09:08.860 ⇒ 00:09:09.430 Brylle Girang: Okay.
79 00:09:09.990 ⇒ 00:09:10.510 Amber Lin: Yep.
80 00:09:11.820 ⇒ 00:09:22.229 Brylle Girang: Perfect, okay. And yeah, Mustafa, once we review the Snowflake certifications that should be unlocked today, let’s move into pushing the data in the AI team.
81 00:09:22.400 ⇒ 00:09:23.820 Brylle Girang: They’re moving fast here.
82 00:09:23.820 ⇒ 00:09:24.390 Mustafa Raja: Yep.
83 00:09:25.820 ⇒ 00:09:31.009 Brylle Girang: Alright, that’s everything on my end. Do you have any questions, Mustafa, Amber?
84 00:09:34.570 ⇒ 00:09:44.379 Amber Lin: No, just to summarize what we talked about, I’ll check in on OmniCerts, I’ll check our…
85 00:09:44.430 ⇒ 00:09:58.700 Amber Lin: foundation, and then open code courses myself, and then see what skills we can develop. And then I’ll wait for your announcement, because I know you’re changing some stuff.
86 00:09:58.760 ⇒ 00:10:03.519 Amber Lin: And then I’ll push the team to complete those foundational courses.
87 00:10:05.730 ⇒ 00:10:18.429 Brylle Girang: Yeah, yeah, cool. Also, one last thing, so I talked with Otem about L&D as a service, and I know that I have discussed this with you when we were, like, building out the Q2 plan for learning and development.
88 00:10:18.590 ⇒ 00:10:26.770 Brylle Girang: We might move faster there, compared to, like, the initial plan on, like, launching it around Q3.
89 00:10:26.910 ⇒ 00:10:45.840 Brylle Girang: After VixelCon, we have had overwhelming responses into people really wanting to adopt AI, and really wanting people to help them learn AI into their companies, into their businesses. So, I am focusing on building out, like, our offers, our playbooks.
90 00:10:46.020 ⇒ 00:10:49.840 Brylle Girang: And, it’s not finalized yet, but I might need
91 00:10:49.950 ⇒ 00:11:07.949 Brylle Girang: everyone’s help into making sure that we cover all bases, and also your feedback, and then your comments on what we can offer to our clients in the future. So it might involve, you know, training people on how to use AI tools, training people how to use cursor, the basic stuff, what is AI, etc.
92 00:11:10.250 ⇒ 00:11:20.790 Brylle Girang: I am not 100% sure yet what specific help I need from the team, but I just want to make sure that you should be expecting that I will be leaning more.
93 00:11:20.860 ⇒ 00:11:31.499 Brylle Girang: Or I will be leaning on you two to help me out with building out the courses, and then the materials that we would need to make sure that we launch a successful L&D service.
94 00:11:33.020 ⇒ 00:11:36.949 Brylle Girang: Yeah, I think that’s it, that’s it on my end.
95 00:11:38.170 ⇒ 00:11:41.329 Uttam Kumaran: Yeah, my, couple things for my side, so…
96 00:11:41.470 ⇒ 00:11:47.110 Uttam Kumaran: I guess, like, how is it going with, like, OpenCode Desktop or Orca? Like, are you guys…
97 00:11:47.860 ⇒ 00:11:55.549 Uttam Kumaran: part of the reason I want to come to this call is, like, are you guys, like, feeling like you’re moving faster because of some of the stuff that came out in, like, the last few weeks?
98 00:11:58.740 ⇒ 00:11:59.380 Uttam Kumaran: Mustafa.
99 00:11:59.470 ⇒ 00:11:59.960 Amber Lin: Whatever.
100 00:11:59.960 ⇒ 00:12:01.409 Uttam Kumaran: Any questions for you?
101 00:12:01.410 ⇒ 00:12:07.199 Amber Lin: Yeah, I’m thinking about, the OpenCo agents. I have…
102 00:12:07.800 ⇒ 00:12:18.230 Amber Lin: not used it after I set it up. I’ve tested a few questions, but I will use that today on my CTA stuff, and I’ll get back to you on,
103 00:12:18.440 ⇒ 00:12:19.759 Amber Lin: How it helped.
104 00:12:20.020 ⇒ 00:12:31.210 Mustafa Raja: Yeah, so for me, I’ve been using Orca with OpenCode CRI, and, it’s been pretty good. I’ve been using, the Kimi 2.5 model with it.
105 00:12:31.720 ⇒ 00:12:43.430 Mustafa Raja: With OpenCodeZen, and it’s been pretty good. I’ve been, I’ve been pushing out stuff with it. So it’s work… it’s working with me. It’s working for me. Pretty good.
106 00:12:44.260 ⇒ 00:12:48.999 Uttam Kumaran: Okay, like, for context, like, this is kind of, like, my workflow right now. I have Orca.
107 00:12:49.180 ⇒ 00:12:54.560 Uttam Kumaran: Which I’m using for, like, platform tasks.
108 00:12:54.670 ⇒ 00:12:57.739 Uttam Kumaran: I don’t even look at a single line of code, to be honest.
109 00:12:57.850 ⇒ 00:13:01.259 Uttam Kumaran: for the most part, I’m like, these are all work trees.
110 00:13:01.540 ⇒ 00:13:06.690 Uttam Kumaran: Honestly, what I’ve been doing is I create the ticket, and then I just hand it to it, and I go CE…
111 00:13:07.430 ⇒ 00:13:10.389 Uttam Kumaran: like, plan, or CE debunk.
112 00:13:10.470 ⇒ 00:13:22.500 Uttam Kumaran: And I kind of let things run. I’m using a mix of codecs until I get basically run out, which is, like, every, like, hour, and then I switch to open code. They’re both just CLI coding.
113 00:13:22.500 ⇒ 00:13:32.250 Uttam Kumaran: So, I have, like, 3 or 4 of these running at all times, sort of fixing things, and then I use the cursor agents. This is mainly, like, small things, like…
114 00:13:32.280 ⇒ 00:13:35.540 Uttam Kumaran: for example, I was working on,
115 00:13:35.880 ⇒ 00:13:38.289 Uttam Kumaran: I’m working on some stuff from, you know.
116 00:13:38.510 ⇒ 00:13:41.750 Uttam Kumaran: Amber, we have our, like, data platform documentation.
117 00:13:43.530 ⇒ 00:13:45.470 Uttam Kumaran: I’ve tried to ask people to just, like.
118 00:13:46.550 ⇒ 00:13:51.820 Uttam Kumaran: get that cleaned up and stuff like that, and I don’t know, it’s not happening, so I’m just gonna rip…
119 00:13:52.140 ⇒ 00:13:56.250 Uttam Kumaran: Template guides, skills, and… shift that. So
120 00:13:56.480 ⇒ 00:13:59.360 Uttam Kumaran: I don’t know, sometimes the interface is nicer here.
121 00:13:59.700 ⇒ 00:14:05.849 Uttam Kumaran: Basically, most likely, I’m gonna replace cursor Agents with our work platform.
122 00:14:06.360 ⇒ 00:14:10.969 Uttam Kumaran: And so I’ll probably be out of cursor entirely here shortly as well.
123 00:14:12.390 ⇒ 00:14:19.070 Uttam Kumaran: So yeah, I mean, I guess my ask is, like, this group is, like, 2 or 3 steps behind me.
124 00:14:19.590 ⇒ 00:14:29.009 Uttam Kumaran: If you guys don’t push it, and if the team is ahead of you, you’re gonna be jammed trying to teach anybody. It’s painful, and yes, like.
125 00:14:29.110 ⇒ 00:14:36.330 Uttam Kumaran: It’s gonna change, like, every week, but try it. Like, this stuff is… I’m moving, like, really lightning fast right now.
126 00:14:36.980 ⇒ 00:14:45.720 Uttam Kumaran: Does that make sense? And then just ping if you’re like, this isn’t working, or, like, this is annoying. I probably have dealt with it, and I’m, like, figuring it out.
127 00:14:48.090 ⇒ 00:14:48.630 Amber Lin: Okay.
128 00:14:50.220 ⇒ 00:14:56.430 Brylle Girang: I will be making, like, an Orca desktop guide today, and then shipping it out to the team.
129 00:14:56.620 ⇒ 00:15:00.530 Brylle Girang: I think it’s amazing, I think… There are people who.
130 00:15:00.530 ⇒ 00:15:01.030 Mustafa Raja: coin.
131 00:15:01.210 ⇒ 00:15:02.720 Brylle Girang: benefit from, from how.
132 00:15:02.720 ⇒ 00:15:03.450 Uttam Kumaran: Yeah.
133 00:15:03.840 ⇒ 00:15:04.880 Brylle Girang: uses this.
134 00:15:06.910 ⇒ 00:15:07.650 Brylle Girang: Yeah.
135 00:15:11.610 ⇒ 00:15:25.000 Uttam Kumaran: The last, the last piece, is I’m gonna also work on… now that we have open code, and we’re gonna have a few other surfaces, I’m gonna work on a dashboard for, like, skill use tracking.
136 00:15:25.130 ⇒ 00:15:32.490 Uttam Kumaran: And, like, general usage? I don’t know, it’s maybe next week, I don’t know.
137 00:15:32.630 ⇒ 00:15:41.229 Uttam Kumaran: So, that’s one thing. The other thing is, I worked on plugins, so… for example,
138 00:15:41.470 ⇒ 00:15:46.690 Uttam Kumaran: We now have a plugins repo, BrainForge Plugin Marketplace.
139 00:15:46.960 ⇒ 00:15:53.609 Uttam Kumaran: You guys know, like, in Kersher, you have plugins, like Amber, Mustafa, you guys, like, I said some stuff about that, you guys probably clipped around that.
140 00:15:54.030 ⇒ 00:15:54.650 Mustafa Raja: Yep.
141 00:15:55.740 ⇒ 00:16:15.520 Uttam Kumaran: Like, you can now install Brainforge Plugin Marketplace. One of the things that’s probably difficult for y’all and everybody is, like, my MCP is broken, or it’s uninstalled, or, like, it’s not in my other repo, this is gonna fix that. Like, you can just tell your agent, install these plugins, or basically, in cursor.
142 00:16:16.010 ⇒ 00:16:24.300 Uttam Kumaran: dashboard plugins. You literally can go into here, and there’s a team marketplace.
143 00:16:24.780 ⇒ 00:16:28.359 Uttam Kumaran: And in here, all of our common plugins.
144 00:16:28.580 ⇒ 00:16:35.309 Uttam Kumaran: And then also, this includes all of the skills, rules, and MCPs. So you no longer have to configure this on your own.
145 00:16:35.620 ⇒ 00:16:38.540 Uttam Kumaran: You could literally just install all of these.
146 00:16:39.070 ⇒ 00:16:45.600 Uttam Kumaran: And I’m gonna move a lot of the skills out of the BrainForge platform repo, so that you can use them across any
147 00:16:46.120 ⇒ 00:16:47.399 Uttam Kumaran: Repo, you’re in.
148 00:16:47.700 ⇒ 00:17:05.740 Uttam Kumaran: Because the skills are more tied to your IDE, or your Agentic workspace, less tied to the repo. Yeah. So, Amber, like, you’re using Cursor here, here, here, it doesn’t matter. Cursor, your local machine will have an understanding of the plugins that are available everywhere.
149 00:17:05.890 ⇒ 00:17:18.209 Uttam Kumaran: And so… this is the first step that’s in the direction of a skills, like, marketplace that I’m, like, working on. Unfortunately, there is not a… there’s not a product for this right now, like, that exists.
150 00:17:18.410 ⇒ 00:17:25.089 Uttam Kumaran: And it’s… so it’s not something I can just, like, find and figure out, so I’ll have to build something for us.
151 00:17:25.310 ⇒ 00:17:30.300 Uttam Kumaran: That basically allows people to, like, search, install, upload, edit.
152 00:17:30.510 ⇒ 00:17:34.009 Uttam Kumaran: proposed changes to Brave Forge Skills.
153 00:17:34.750 ⇒ 00:17:42.310 Uttam Kumaran: The other thing that this allows for, and Amber, you saw this, is, like, CT… your client will have specific skills, you may have Amber skills.
154 00:17:42.820 ⇒ 00:17:49.139 Uttam Kumaran: the data service may have skills, and the company may have skills, right? So I want to allow for all of that.
155 00:17:49.970 ⇒ 00:17:57.579 Uttam Kumaran: And… Again, this may be something I hand off to Davis or something, I’m not sure yet, so…
156 00:17:57.840 ⇒ 00:18:01.560 Uttam Kumaran: Try this out today and uninstall your other stuff.
157 00:18:01.680 ⇒ 00:18:07.029 Uttam Kumaran: And then, if it’s working really well, tell everybody about this. I’m not gonna have time.
158 00:18:07.030 ⇒ 00:18:07.410 Amber Lin: Okay.
159 00:18:07.410 ⇒ 00:18:11.389 Uttam Kumaran: But, yeah, this should be much, much easier for you.
160 00:18:11.700 ⇒ 00:18:12.240 Brylle Girang: I am…
161 00:18:12.240 ⇒ 00:18:12.860 Amber Lin: Okay.
162 00:18:12.860 ⇒ 00:18:14.399 Brylle Girang: Those things here with them.
163 00:18:14.400 ⇒ 00:18:14.730 Uttam Kumaran: Yeah.
164 00:18:14.730 ⇒ 00:18:22.720 Brylle Girang: The first one is that I understand that only… I think… correct me if I’m wrong, but only Cursor, like, supports these plugins.
165 00:18:23.020 ⇒ 00:18:23.880 Uttam Kumaran: Not true.
166 00:18:24.170 ⇒ 00:18:25.010 Uttam Kumaran: Not true.
167 00:18:25.010 ⇒ 00:18:25.430 Brylle Girang: Okay.
168 00:18:25.430 ⇒ 00:18:32.090 Uttam Kumaran: So, NPX, like, plugins, basically, there is a…
169 00:18:33.650 ⇒ 00:18:40.610 Uttam Kumaran: There is basically, like, a plugins… Well, okay, let’s… basically,
170 00:18:43.310 ⇒ 00:18:46.469 Uttam Kumaran: Across every single surface, you can use plugins.
171 00:18:46.760 ⇒ 00:18:58.460 Uttam Kumaran: So, for example, there are plugins for open code, for codecs, you can install plugins, and so plugins is a first-class citizen across all of…
172 00:18:58.570 ⇒ 00:19:02.010 Uttam Kumaran: Across every surface, actually.
173 00:19:02.100 ⇒ 00:19:18.759 Uttam Kumaran: So it’s not just a cursor-specific thing. I am going to move to nothing is gonna be cursor-specific anymore. Okay. In fact, you’ll see that you may not have noticed, but a lot of the stuff in the repo is gonna be… you’re able to use it across open code.
174 00:19:18.780 ⇒ 00:19:23.099 Uttam Kumaran: Claude, Codex, 4Cursor.
175 00:19:23.270 ⇒ 00:19:32.710 Uttam Kumaran: or whatever the hell you… whatever CLI agent editor you bring to the table. So, Plugins is, like, a first-class citizen. Plugins is basically…
176 00:19:32.820 ⇒ 00:19:39.520 Uttam Kumaran: MCP, hooks, rules, skills, all, like, in one.
177 00:19:39.830 ⇒ 00:19:40.250 Brylle Girang: Yeah.
178 00:19:40.280 ⇒ 00:19:43.400 Uttam Kumaran: And it’s recognized, kind of, globally, so…
179 00:19:43.610 ⇒ 00:19:47.049 Uttam Kumaran: Sorry to cut you off, but yeah, it is like a… it’s a more global thing.
180 00:19:47.390 ⇒ 00:19:53.500 Brylle Girang: Yeah, that’s one down, that’s… That’s good. The second piece is that,
181 00:19:53.970 ⇒ 00:19:58.020 Brylle Girang: What was it again? Oh yeah, so how does this, like, affect…
182 00:19:58.470 ⇒ 00:20:10.460 Brylle Girang: how people can contribute to, like, our skills. Like, since it’s going to be a different repo, it’s going to be a plugin, can people, like, easily contribute to our skills, create skills, edit skills, etc?
183 00:20:17.200 ⇒ 00:20:21.440 Uttam Kumaran: There’s gonna be fewer and fewer people,
184 00:20:25.190 ⇒ 00:20:30.259 Uttam Kumaran: There’s gonna be fewer and fewer people that are gonna be using IDEs for their work.
185 00:20:30.610 ⇒ 00:20:36.200 Uttam Kumaran: A lot of people, like the CSOs, for example, probably just gonna move to using work.
186 00:20:36.460 ⇒ 00:20:42.320 Uttam Kumaran: So we’re gonna… so I think some of us will propose PRs to edit skills.
187 00:20:42.420 ⇒ 00:20:46.310 Uttam Kumaran: But some people are going to propose changes to skills via UI.
188 00:20:46.640 ⇒ 00:20:50.649 Uttam Kumaran: all of those have to go through an approval, and frankly, I think this group
189 00:20:50.800 ⇒ 00:20:54.810 Uttam Kumaran: It’s probably gonna be, like, the owners of skills across the company.
190 00:20:54.940 ⇒ 00:20:56.899 Uttam Kumaran: Skills are gonna be like products.
191 00:20:57.830 ⇒ 00:21:06.860 Uttam Kumaran: So, the fact that the skill works, the integration works, I don’t… I’m actually seeing, like, how fast…
192 00:21:07.350 ⇒ 00:21:11.930 Uttam Kumaran: we’re moving, which is not… like, I think this group… I mean, like, I think…
193 00:21:12.180 ⇒ 00:21:24.489 Uttam Kumaran: I think Amber and Bea, you’re creating skills. Mustafa, I don’t see you creating skills. I’m creating, like, a shitload of skills. I don’t see many other people in the company creating skills, so my hunch there is that, like.
194 00:21:25.400 ⇒ 00:21:36.160 Uttam Kumaran: I would like them to, I don’t think people are getting it, I’m not gonna wait, so I’m just gonna keep working on it, and then this group is probably gonna own the first version of every skill. As people make updates.
195 00:21:36.290 ⇒ 00:21:45.560 Uttam Kumaran: they’ll propose a PR, and then we will basically approve that. Whether that PR comes through the UI, or whatever, I think I’ll have to figure that out.
196 00:21:46.140 ⇒ 00:21:46.760 Brylle Girang: Okay.
197 00:21:47.670 ⇒ 00:21:48.240 Brylle Girang: What’s…
198 00:21:48.240 ⇒ 00:22:01.280 Uttam Kumaran: I’m telling you, like, this is where, like, I’m finding that my ability to, like, accelerate… after every major piece of work, I try to come up with a skill so I can log that. I’m also halfway through a skill cleanup
199 00:22:01.440 ⇒ 00:22:03.520 Uttam Kumaran: Workstream. But…
200 00:22:03.740 ⇒ 00:22:13.869 Uttam Kumaran: Honestly, what I’m finding is the compound engineering skills are sick. I’m using it, like, for the most. I’m basically, like, using that for everything. I have an idea, I run ID8,
201 00:22:14.140 ⇒ 00:22:17.379 Uttam Kumaran: I then take a couple of those ideas, I run Brainstorm.
202 00:22:17.780 ⇒ 00:22:20.230 Uttam Kumaran: take a couple of those, I run Plan.
203 00:22:20.830 ⇒ 00:22:24.140 Uttam Kumaran: And then I create a ticket with the plan, and then I exit, and then I do work.
204 00:22:24.650 ⇒ 00:22:32.500 Uttam Kumaran: And it’s, like, it’s super good. Like, try it, like, today. It’s… I wouldn’t say it if it was, like, even, like.
205 00:22:32.680 ⇒ 00:22:46.789 Uttam Kumaran: okay, it’s really, really good. It will save you a lot of time. Like, Amber, if you have a package of work for a semantic layer, or… you could even… you should even try giving it, like, hey, figure out the roadmap for Snowflake agents.
206 00:22:47.070 ⇒ 00:22:57.200 Uttam Kumaran: So what I would do in that case is I would literally go into Orca, or whatever, I would say, I would say, great, like, let’s just say…
207 00:22:57.700 ⇒ 00:23:06.180 Uttam Kumaran: Oh, I have workspace here. Where is, Where do I go?
208 00:23:06.610 ⇒ 00:23:07.790 Uttam Kumaran: Wait, what?
209 00:23:13.020 ⇒ 00:23:18.160 Uttam Kumaran: Okay, well, let’s just say I was in… I was in Kirscher. The way I would do this…
210 00:23:18.400 ⇒ 00:23:27.269 Uttam Kumaran: you know, you could just be the platform or whatever. I would basically say, help me create a roadmap for
211 00:23:27.400 ⇒ 00:23:43.579 Uttam Kumaran: Evolving from using just Cortex, Sidebar, and Analyst to using Cortex agents in Snowflake for CTA.
212 00:23:43.820 ⇒ 00:23:46.649 Uttam Kumaran: And then I would use this, I would say CDID8.
213 00:23:47.900 ⇒ 00:23:51.339 Uttam Kumaran: And, like, If I just hit enter.
214 00:23:52.360 ⇒ 00:23:58.829 Uttam Kumaran: Gosh. That’s basically, like, what I would… what I would try to do. You’re gonna find that it’s, like.
215 00:23:59.190 ⇒ 00:24:09.369 Uttam Kumaran: pretty… pretty effective. It’ll come up with a plan, it’ll, like, hit the plan a bunch of times, and then it’ll be, like, which ones do you want to do? And then you can just say, go for it, and it’ll work.
216 00:24:09.710 ⇒ 00:24:10.210 Brylle Girang: Yeah.
217 00:24:12.120 ⇒ 00:24:13.640 Amber Lin: Exciting. Okay.
218 00:24:14.740 ⇒ 00:24:26.330 Brylle Girang: I think one piece for L&D here is making sure that people understand that compound engineering is not just for engineers, so the skill name might… might… might be a bit misleading, but yeah.
219 00:24:37.970 ⇒ 00:24:42.730 Uttam Kumaran: I just wanna, yeah, if we have 2 more minutes, I just wanna show you guys this process. So, like…
220 00:24:44.410 ⇒ 00:25:01.239 Uttam Kumaran: stop slacking for, like, 2 seconds and just watch, because this is gonna… I think this is what I’ve been doing for the last 2 weeks. It’s been absolutely gold mine right now. So, basically, what it’s doing is it’s going… it’s literally, like, looking through… it’s gonna use, EXO Web Search, look through the Snowflake docs.
221 00:25:01.310 ⇒ 00:25:12.429 Uttam Kumaran: find a bunch of stuff, right? It’s gonna write the ideation plan. Basically, raw candidates are, like, candidates for, basically, like, candidates for ideas.
222 00:25:12.780 ⇒ 00:25:23.310 Uttam Kumaran: Power users gets agents, keep analysts in the sidebar, and then it’s basically gonna, like, go in and look at all those ideas and find what they’re calling survivors.
223 00:25:23.850 ⇒ 00:25:26.310 Uttam Kumaran: Which is pretty good.
224 00:25:27.000 ⇒ 00:25:34.950 Uttam Kumaran: And… great. So it says roadmap, foundation, strengthen the semantic model, build a golden question set.
225 00:25:35.140 ⇒ 00:25:48.459 Uttam Kumaran: run a two-speed period, sidebar analyst, and agents, full agent value. Here’s the ideas. One semantic model for analysts is a tool. Operating model. Okay, here’s a search readiness, golden questions, governance.
226 00:25:48.840 ⇒ 00:25:53.279 Uttam Kumaran: And then it sort of looks through, like, okay, which one should we go after? So…
227 00:25:53.500 ⇒ 00:26:03.600 Uttam Kumaran: basically, this is great. It’s like, agents call the analyst. So one thing you can basically do is say, cool, these are, like, 5 ideas I want to work on. So, let’s just say, like.
228 00:26:03.870 ⇒ 00:26:12.960 Uttam Kumaran: Great. I… I love 1, 2, 3, 4, I’ll say CE Brainstorm.
229 00:26:14.890 ⇒ 00:26:20.469 Uttam Kumaran: And what it’s gonna do is literally gonna go one step further and expand each of those ideas.
230 00:26:20.660 ⇒ 00:26:23.340 Uttam Kumaran: And build requirements.
231 00:26:32.820 ⇒ 00:26:34.530 Uttam Kumaran: And I find this, like.
232 00:26:34.940 ⇒ 00:26:45.220 Uttam Kumaran: it creates really, really robust plans that, like, you can actually just quickly move to execution. And so, as soon as this kind of starts writing.
233 00:27:01.660 ⇒ 00:27:07.909 Uttam Kumaran: Cool. So, if I… if we look at this, and I go to, like, a non-markdown…
234 00:27:08.690 ⇒ 00:27:10.549 Uttam Kumaran: Alright, it’s ugly.
235 00:27:15.350 ⇒ 00:27:33.739 Uttam Kumaran: Blah blah blah, the organization today gets Cortex and Snowflake, Cortex Analyst, the next step is to do Cortex agents with orchestration work. Yeah, basically, all production Cortex Analyst comes from the semantic view. For the first production agent’s experience, should Cortex search be in scope, or should Wave 1 be analyst-only agents?
236 00:27:34.030 ⇒ 00:27:38.480 Uttam Kumaran: I’m gonna say both, blah blah blah. So you see how it’s, like, gonna question me?
237 00:27:38.710 ⇒ 00:27:52.010 Uttam Kumaran: And it’s gonna keep going, it’s gonna say, this needs more research, I need a user decision, it’s… it’s gonna say, I need a user decision, and it’s gonna make decisions wherever else it can make a decision. It’s gonna basically say, cool, we’re in a good spot, do you want to go plan?
238 00:27:52.280 ⇒ 00:27:55.449 Uttam Kumaran: like… You guys get the picture here?
239 00:27:55.450 ⇒ 00:27:55.950 Brylle Girang: Yep.
240 00:27:55.950 ⇒ 00:28:00.190 Uttam Kumaran: And so, immediately after this, you can… you can run CEPlan.
241 00:28:01.380 ⇒ 00:28:03.630 Uttam Kumaran: And then you literally can run CE work.
242 00:28:03.790 ⇒ 00:28:13.329 Uttam Kumaran: And CE work, it forces the agent to work, like, really, really long running, so, like, you don’t have to, like, check in on it. It’ll go as far as it can go.
243 00:28:14.560 ⇒ 00:28:28.320 Uttam Kumaran: And I think Amber and Mustafa, what you’re gonna find is many of your tasks, you can literally just hand to it, run through CE debug, CE plan, CE work, and it finishes it. Like, Mustafa, the, the,
244 00:28:28.440 ⇒ 00:28:31.070 Uttam Kumaran: the JSON issue with Eden.
245 00:28:31.730 ⇒ 00:28:36.659 Uttam Kumaran: Yesterday, I literally did it this morning, I said, CE debug, I handed it a ticket, and then I…
246 00:28:37.050 ⇒ 00:28:38.830 Uttam Kumaran: I didn’t even look at it for, like.
247 00:28:38.990 ⇒ 00:28:41.080 Uttam Kumaran: I just finished it, like, 10 minutes.
248 00:28:41.940 ⇒ 00:28:42.730 Mustafa Raja: Yeah.
249 00:28:43.250 ⇒ 00:28:46.619 Uttam Kumaran: See how that’s, like, kind of in… it’s, like, kind of crazy?
250 00:28:48.250 ⇒ 00:29:02.359 Uttam Kumaran: So that’s what I want you guys to try this, because I feel like you’re gonna see that, actually, most of your things, you’re limited by the fact that the agent didn’t… that you didn’t plan earlier, and that the agent isn’t working longer.
251 00:29:02.510 ⇒ 00:29:08.019 Uttam Kumaran: Right? And so both of those are, like, an us problem. We can use the planning features to plan.
252 00:29:08.170 ⇒ 00:29:11.929 Uttam Kumaran: Send it to work, and then where it gets jammed, you can tweak it, but…
253 00:29:12.330 ⇒ 00:29:16.590 Uttam Kumaran: I’m finding it’s getting 90% of the way there on most tasks that I have.
254 00:29:19.550 ⇒ 00:29:21.849 Uttam Kumaran: So that’s my pitch for you today. So, hopefully at these meetings.
255 00:29:21.850 ⇒ 00:29:22.310 Amber Lin: Because I’ll come.
256 00:29:22.310 ⇒ 00:29:26.259 Uttam Kumaran: With a little bit of a pitch, for you guys to try something.
257 00:29:27.000 ⇒ 00:29:31.029 Brylle Girang: I’m actually going to, like, get the recordings here and then trim it down.
258 00:29:31.250 ⇒ 00:29:32.440 Uttam Kumaran: Okay.
259 00:29:32.440 ⇒ 00:29:34.010 Brylle Girang: As about turrets.
260 00:29:34.320 ⇒ 00:29:36.119 Uttam Kumaran: Yeah, good. Hi.
261 00:29:39.880 ⇒ 00:29:42.150 Uttam Kumaran: Okay, it’s 11, so…
262 00:29:42.150 ⇒ 00:29:51.309 Brylle Girang: Yeah, I will be creating, like, a dedicated enablement guide for compound engineering. Mustafa, Amber, let’s work together. I would love for us to, like.
263 00:29:51.540 ⇒ 00:29:59.059 Brylle Girang: Lay down examples for each work stream there. That would… that would give people more ideas on how to use this workflow.
264 00:29:59.410 ⇒ 00:30:00.919 Brylle Girang: I’ll slack you on that.
265 00:30:01.810 ⇒ 00:30:02.500 Mustafa Raja: Okay.
266 00:30:05.430 ⇒ 00:30:06.700 Brylle Girang: Thank you, everyone. Bye-bye.
267 00:30:06.700 ⇒ 00:30:07.729 Uttam Kumaran: Thank you. Bye.