Meeting Title: Project Review - ForgeLab Date: 2026-04-09 Meeting participants: Rico Rejoso, Miranda Wen, Pranav Narahari, Brylle Girang, Uttam Kumaran
WEBVTT
1 00:00:55.910 ⇒ 00:00:57.190 Pranav Narahari: Hey, Miranda.
2 00:01:44.580 ⇒ 00:01:45.480 Brylle Girang: Hey!
3 00:01:45.950 ⇒ 00:01:46.720 Uttam Kumaran: Guys.
4 00:01:48.920 ⇒ 00:01:49.640 Pranav Narahari: Hey, guys.
5 00:01:54.710 ⇒ 00:01:56.110 Miranda Wen: Hi, hi.
6 00:01:56.110 ⇒ 00:01:57.020 Uttam Kumaran: Hey!
7 00:01:57.200 ⇒ 00:02:00.870 Brylle Girang: Are you ready, Miranda, for your first thesis defense?
8 00:02:00.870 ⇒ 00:02:01.520 Miranda Wen: Yes.
9 00:02:01.520 ⇒ 00:02:02.290 Uttam Kumaran: Yes.
10 00:02:02.290 ⇒ 00:02:06.879 Miranda Wen: Okay, this is causing me nervous, like…
11 00:02:06.880 ⇒ 00:02:09.250 Uttam Kumaran: Good, good.
12 00:02:09.259 ⇒ 00:02:13.159 Brylle Girang: We have a newly grad here, Panam, so we’re supportive of her.
13 00:02:14.010 ⇒ 00:02:15.390 Miranda Wen: That’s cute.
14 00:02:15.390 ⇒ 00:02:16.020 Pranav Narahari: Nope.
15 00:02:16.620 ⇒ 00:02:18.260 Miranda Wen: Okay, thank you for…
16 00:02:18.260 ⇒ 00:02:18.779 Uttam Kumaran: Yeah, maybe.
17 00:02:19.530 ⇒ 00:02:25.970 Uttam Kumaran: Yeah, maybe just to set the stage a little bit, I think B and Pranav would love you guys to sort of put yourself
18 00:02:26.190 ⇒ 00:02:34.810 Uttam Kumaran: as, like, customers, as, like, builders on top of what Miranda’s gonna present, and ask a lot of questions. I really…
19 00:02:34.870 ⇒ 00:02:50.639 Uttam Kumaran: I’m hopeful I’m not gonna ask the majority of the questions today. I want you guys… you guys are folks that are close to AI, and especially close to this product, and, like, this is being built for you and your customers.
20 00:02:50.710 ⇒ 00:02:55.320 Uttam Kumaran: So, like, yeah, I just want to set the stage, and yeah, Miranda, I can hand it back to you.
21 00:02:56.670 ⇒ 00:02:58.420 Miranda Wen: Yes, yes,
22 00:02:59.000 ⇒ 00:03:08.960 Miranda Wen: Yeah, but definitely hoping to get, like, more constructive feedback, because this is the first time I’m doing a product review. Shall we still wait for Rico, or should I just, like, kick it off?
23 00:03:09.120 ⇒ 00:03:10.700 Uttam Kumaran: Yeah, let’s go ahead, yeah.
24 00:03:11.110 ⇒ 00:03:11.840 Miranda Wen: Okay.
25 00:03:20.070 ⇒ 00:03:22.210 Miranda Wen: Yeah, can everybody see my screen?
26 00:03:24.190 ⇒ 00:03:24.870 Pranav Narahari: Yep.
27 00:03:25.460 ⇒ 00:03:26.580 Miranda Wen: Okay, awesome.
28 00:03:26.590 ⇒ 00:03:44.909 Miranda Wen: Yeah, so for this one, I call this, like, Forge Lab for this, platform. So, the goal of Forge Lab is to basically create this, like, unified surface, and start… the main functions, like, to have the AI SKUs becomes, like, very usable product, easy to run, share.
29 00:03:44.910 ⇒ 00:03:59.679 Miranda Wen: governed and reliable, to everyday work across teams, ideally in the future. And, it means to, like, make it more intuitive and less complex for people to execute repeatable workflows faster.
30 00:03:59.680 ⇒ 00:04:08.889 Miranda Wen: more consistently and at a scale. And the scope right now, like, we want to start, as a initial labs.
31 00:04:08.910 ⇒ 00:04:14.120 Miranda Wen: like, open work. It’s built on open work, so we want to, like, start it as an intro…
32 00:04:14.250 ⇒ 00:04:24.120 Miranda Wen: surface for our Forge Lab. We will create some skills, and execute through, like, guided, usable workflows, and,
33 00:04:24.120 ⇒ 00:04:43.929 Miranda Wen: roll out with some internal user group, and develop some features for this platform that can really best to, benefit this, like, internal user groups from here. And then later on, we want to, like, roll out to integrate into the platform, no longer being the open work separately, as well as, like.
34 00:04:43.970 ⇒ 00:04:50.269 Miranda Wen: Be really across the teams, to really move on to, like, other…
35 00:04:50.410 ⇒ 00:05:05.590 Miranda Wen: team members. I will talk more about that in target users and interest strategy, as well as, like, to have, like, a self-serve onboarding, because one of the major things we hope to achieve is to have, like, a very guided platform.
36 00:05:05.590 ⇒ 00:05:07.539 Brylle Girang: Amanda, I just have a question here.
37 00:05:07.540 ⇒ 00:05:08.539 Miranda Wen: Yeah, yeah, yeah.
38 00:05:08.540 ⇒ 00:05:14.870 Brylle Girang: So, what’s going to be the main difference between Forge Lab and what we’re using right now, which is Cursor?
39 00:05:15.400 ⇒ 00:05:23.160 Miranda Wen: Yes, I think, like, for Cursor right now, like, I think the agreement I have with him is, like,
40 00:05:23.470 ⇒ 00:05:29.089 Miranda Wen: Cursor is kind of, like, overkill right now, and we want, like, for cursor, like.
41 00:05:29.310 ⇒ 00:05:38.870 Miranda Wen: Although it’s helpful, like, we can call on skills, but it does not, like, it’s… when we are selling to our…
42 00:05:38.880 ⇒ 00:05:55.969 Miranda Wen: clients, there’s no way, like, we’re selling through Cursor. We need to, like, really package it and make a surface to make it very much presentable. And also, Cursor is, like, very much an IDE platform, and in this surface, we want it to be…
43 00:05:56.180 ⇒ 00:06:14.640 Miranda Wen: intuitive. It’s kind of, like, more, like, chat-native. I will probably talk a little bit more, on the use case catalog system where we’re trying to bet on. That will be different from, like, Cursor, because we benchmark, how we’re gonna be better than some generic AI tooling, like cursor.
44 00:06:16.450 ⇒ 00:06:17.440 Miranda Wen: Okay. Alright.
45 00:06:17.930 ⇒ 00:06:28.990 Miranda Wen: And then moving on to target users, I think, generally speaking, we have, like, 3 broad buckets. One’s, like, workflow operators,
46 00:06:29.070 ⇒ 00:06:46.780 Miranda Wen: And, this could include, like, where we’re considering the first… I’ll talk together with the intro strategy. The first, group we’re targeting, which is a go-to-market team, the business technical sales as a primary intro team. The reason we choose that is that
47 00:06:46.780 ⇒ 00:07:02.139 Miranda Wen: we want them… this… the skills and also this platform to be for them is, like, they don’t need to do much of, like, additional work to tweak to build a skill ground up. It would just, like, deliver us a full package for them. It will be already used for them.
48 00:07:02.140 ⇒ 00:07:11.840 Miranda Wen: And we also consider them as a very ideal, good design partners, because, they having, like, different branches, for example, as I talked to Robert, like,
49 00:07:11.920 ⇒ 00:07:28.429 Miranda Wen: partnership management, as well as, like, demos, could be some repeatable workflows that we first started building, as well as, like, getting more feedback on how to make it more intuitive to everybody on the team. And,
50 00:07:28.690 ⇒ 00:07:30.970 Miranda Wen: Yeah, and moving on, we think, like.
51 00:07:31.230 ⇒ 00:07:35.610 Miranda Wen: like, the strategy CSO delivery next, and we’re…
52 00:07:36.170 ⇒ 00:07:44.389 Miranda Wen: I probably will, like, interview around the group, and maybe select, like, one person from each of this, or,
53 00:07:44.500 ⇒ 00:07:55.409 Miranda Wen: select, like, one user among this to really see if can this… can we, like, basically repeat the process with the first user group to have, like.
54 00:07:55.510 ⇒ 00:08:10.739 Miranda Wen: a set of 2 or 3 repeatable, workflows, skills available, and then getting more feedbacks on, like, features development for the platform. And then, lastly, probably moving on to, like, engineers, builders. And for that.
55 00:08:10.860 ⇒ 00:08:21.879 Miranda Wen: I think it would be kind of very different from what we were considering for the business team and the CSO team, because, like, for them, we do expect them to, like.
56 00:08:22.480 ⇒ 00:08:30.209 Miranda Wen: be, like, the… the beauters, basically, it’s falling to, like, the beauty bucket, where they will…
57 00:08:30.390 ⇒ 00:08:43.109 Miranda Wen: be doing more orchestration work, and doing more adapt and patch skills, and so we will see, like, how the first two feedbacks going, and see if it’s, like, necessary for us to roll out to the engineer’s team.
58 00:08:43.620 ⇒ 00:08:44.570 Miranda Wen: And…
59 00:08:45.200 ⇒ 00:08:59.680 Miranda Wen: For the use case, I think hypothesis, I think this is a core part that can also answer B’s pro- question better. So, for now, like, Forge Lab is being used as validated first for internal brain forge use,
60 00:08:59.680 ⇒ 00:09:11.240 Miranda Wen: as a Chan-native product surface for repeatable workflow, and with the goal for us here to achieve, better, more than curses, like, to be guided, easy to discover.
61 00:09:11.240 ⇒ 00:09:27.669 Miranda Wen: repeated… repeatable by default, and easy to adapt without rebooting, which is aligned with, like, what we wanted to achieve when we are having our first internal user group with the business team, as well later on, rolling on to CSO. So the hypothesis is, like.
62 00:09:27.810 ⇒ 00:09:32.899 Miranda Wen: In our… in Greenforge, we can use
63 00:09:33.060 ⇒ 00:09:44.140 Miranda Wen: we will complete tasks more reliable and reach value faster, so basically the time to value, than generic chat tools, such as, like.
64 00:09:44.260 ⇒ 00:09:56.690 Miranda Wen: chat, Cloud Gemini, or special… specialist AI tooling such as Cursor, Kodak, OpenCode, or remote work, more so for the go-to-market team. And the core business that we want
65 00:09:56.880 ⇒ 00:10:10.880 Miranda Wen: the Forge Lab, users will return to Forge Lab, and makes repeatable work easy to discover, run, reuse, and trust the environment, beautiful outcomes, rather than expert behavior with client-facing use.
66 00:10:11.020 ⇒ 00:10:13.200 Miranda Wen: So, like.
67 00:10:13.240 ⇒ 00:10:30.029 Miranda Wen: later on, after we really prove these things internally, we can potentially roll out to customers. I want to pause for a second, see if we have, like, any questions here, because I feel this is, like, very important for our directions. And, Bia, I saw, like, had a conversation with you on Slack.
68 00:10:30.030 ⇒ 00:10:46.109 Miranda Wen: I asked about the maxim… how to, like, maximize the skills? And I think the ones I highlighted are what I think now, as a general direction on future, how do we maximize the skills? But I want to see, like, if you have, like, more input or, questions about it.
69 00:10:59.730 ⇒ 00:11:09.589 Pranav Narahari: I probably just have, like, the least amount of context to this, too, but yeah, be with your question, and then, Miranda, just like, as you’ve been presenting, I’m getting more context.
70 00:11:10.630 ⇒ 00:11:17.020 Pranav Narahari: So, yeah, I’m kind of interested to know… How, kind of…
71 00:11:18.540 ⇒ 00:11:30.910 Pranav Narahari: Yeah, like, to Bea’s point, we have Cursor that’s doing a lot of, this stuff, and it sounds like, Miranda, what you’re talking about is building a system that is, in some ways, more user-friendly and
72 00:11:31.040 ⇒ 00:11:34.070 Pranav Narahari: More descriptively, like, based on, like, what you just wrote here, like…
73 00:11:34.200 ⇒ 00:11:41.670 Pranav Narahari: Easy to discover, yeah, repeatable, easy to adapt, things of that nature.
74 00:11:43.890 ⇒ 00:11:46.950 Pranav Narahari: I, I guess my, my question is, is,
75 00:11:47.140 ⇒ 00:12:00.859 Pranav Narahari: how much of this would need to be, like, built from, like, scratch? And maybe that’s a question for a different time, versus just, like, built using what we have currently existing with, like, the cursor skills and the platform.
76 00:12:06.510 ⇒ 00:12:16.710 Miranda Wen: So, you mean, like, for this, like, Forge lab? Like, specifically, how is, like, the, how does it look like right now, and how far away from, like, achieving this?
77 00:12:17.660 ⇒ 00:12:31.579 Pranav Narahari: Or I guess, I’m guessing this is, like, just probably just on paper right now, like, we haven’t started building much. I guess my question is, like, how does this take off based on where we’re at currently with just…
78 00:12:31.720 ⇒ 00:12:39.949 Pranav Narahari: our cursor environment, our, cursor skills, and everything else, like, that we’re using from the Brainforge platform.
79 00:12:41.600 ⇒ 00:13:00.970 Miranda Wen: Yeah, so right now, like, we already have, like, a, very basic one built out in OpenWork, but it’s, like, it has, like, some basic functionality where, like, you can chat with it, but in terms of products, like, like, the… for example, the adding skills and this haven’t been really built out.
80 00:13:01.080 ⇒ 00:13:06.169 Miranda Wen: And I think, like, the part where…
81 00:13:06.580 ⇒ 00:13:11.480 Miranda Wen: we will be able to, like, save… save a lot of time, I think.
82 00:13:11.600 ⇒ 00:13:21.110 Miranda Wen: also yesterday, talking with me and was like, I remember, I think, once we got, like, skills ready, and we can start, like.
83 00:13:21.340 ⇒ 00:13:23.819 Miranda Wen: Like, building skills and…
84 00:13:23.960 ⇒ 00:13:30.519 Miranda Wen: Basically, building out skills, and… or even just, like, import skills that we already have that are gonna…
85 00:13:30.680 ⇒ 00:13:45.389 Miranda Wen: best fit for the go-to-market team is going to be fast. And I think, like, where we see more time is really about, like, developing features to achieve what I, the highlighted
86 00:13:45.530 ⇒ 00:13:50.710 Miranda Wen: highlighted goals, highlighted objective or goals here. Whoa.
87 00:13:50.870 ⇒ 00:13:57.410 Miranda Wen: be, like, the main part, but also, it’s like, we’re, well, truly differentiated with, Cursor.
88 00:13:57.790 ⇒ 00:14:01.240 Miranda Wen: And I add on top of that, and I think, like.
89 00:14:01.540 ⇒ 00:14:07.260 Miranda Wen: one… one value I personally think, and what I hope to achieve, it’s like…
90 00:14:07.600 ⇒ 00:14:14.870 Miranda Wen: In the future. For Brainforge, like, everybody, is on Cursor, and everybody’s basically familiar with Cursor.
91 00:14:15.140 ⇒ 00:14:16.380 Miranda Wen: But,
92 00:14:16.590 ⇒ 00:14:25.739 Miranda Wen: we don’t think, like, everybody is necessary to… it’s not every team is necessary to be on cursor. We don’t think that that’s what, like.
93 00:14:26.030 ⇒ 00:14:30.910 Miranda Wen: Cars are really beautiful, or in the future, the development should be like.
94 00:14:31.010 ⇒ 00:14:37.820 Miranda Wen: They’re, curse is definitely for people who are, like, Non-builder, no non-coder.
95 00:14:38.140 ⇒ 00:14:39.590 Miranda Wen: background, it’s…
96 00:14:39.920 ⇒ 00:14:59.679 Miranda Wen: a little bit, like, less intuitive for them, and it takes more time for them to wrap up. So I think, like, ideally, in the future, what we imagine right now, it can, like, for example, time to probably be, like, someone who never used Cursor, who is just, like, going for the sales team, they don’t need to figure out, like, how to build a skill from scratch. They don’t need to figure out, like.
97 00:14:59.970 ⇒ 00:15:13.469 Miranda Wen: the… the giant database of cursors. They just, like, need to… with Forge Lab, they just need to focus on sales strategy and really get what they need, all what they need from Forge Lab.
98 00:15:13.560 ⇒ 00:15:21.469 Miranda Wen: I think that’s our goal here, and they don’t need, like, much, like, wrap-up or, time to educate themselves.
99 00:15:21.690 ⇒ 00:15:32.919 Miranda Wen: In the way that Kurtza does, because for them, it’s, like, the outcome that generates from AI is more important than them, like, to learn how to, like, build a skill from scratch and stuff.
100 00:15:33.680 ⇒ 00:15:35.710 Miranda Wen: If that answers your question.
101 00:15:36.670 ⇒ 00:15:41.180 Pranav Narahari: Got you, yeah, I think I’m getting more, like, kind of context for this, so yeah, that helps.
102 00:15:42.690 ⇒ 00:15:43.820 Miranda Wen: Thank you, yeah.
103 00:15:44.010 ⇒ 00:15:52.029 Brylle Girang: I guess, Miranda, one question that I also have about this is, why… why are we building this now? Like, why is this important now?
104 00:15:53.900 ⇒ 00:15:56.869 Miranda Wen: Mmm… wait, that’s a really good question.
105 00:15:57.270 ⇒ 00:16:03.239 Miranda Wen: I think from my perspective, is that I feel like…
106 00:16:03.540 ⇒ 00:16:15.629 Miranda Wen: skills is, like, a relatively new thing, and people are exploring, like, what it could… could do, obviously. And I saw, like, it can be used to, like.
107 00:16:15.770 ⇒ 00:16:30.500 Miranda Wen: That’s probably using it as, like, like, document console, like, grow me, like, this for document review. And can also… I saw people using it to, like, replace, literally, an ex-colleague with all their experience, and, expertise.
108 00:16:31.720 ⇒ 00:16:45.489 Miranda Wen: And I think, like, the skills had, like, so much potential to be tapped into, but I would say it’s, like, still pretty much… in my personal opinion, maybe really biased,
109 00:16:45.790 ⇒ 00:16:57.170 Miranda Wen: it’s not being, like, widely… so first off, it’s not, like, widely being adopted by organization, and I think a core reason for that is, like, they don’t… they don’t have, like, brain force. Everything’s, like.
110 00:16:57.250 ⇒ 00:17:15.489 Miranda Wen: on database, and everybody’s, like, familiar with Cursor. But secondly, skill is definitely very much useful. It’s definitely, like, way better than them just, like, having an enterprise plan for whatever we say, like, cloud or, or open code here. So, I think, like, this will be, like, an…
111 00:17:15.900 ⇒ 00:17:28.470 Miranda Wen: gap where we could tap into, for organizations that are definitely… most of the world is, you know, less tech-forward than BringForge, and, where
112 00:17:28.680 ⇒ 00:17:45.680 Miranda Wen: the skill… the potential for skills is… haven’t been really explored by them yet, if that makes sense. That’s my opinion, but I also, like, also wouldn’t, like, had another doc which, like, really started this project. I want to see if I’m on the right path for this as well.
113 00:17:50.880 ⇒ 00:17:51.480 Brylle Girang: Okay.
114 00:17:52.550 ⇒ 00:17:53.200 Miranda Wen: Yeah.
115 00:17:55.420 ⇒ 00:18:00.459 Miranda Wen: Cool, then I will move forward to experimentation plan.
116 00:18:00.960 ⇒ 00:18:02.609 Miranda Wen: So,
117 00:18:03.130 ⇒ 00:18:22.129 Miranda Wen: For the experimentation plan, so we basically have a… the very beginning, like, we have this, like, workflow execution, which I mentioned before. So we have, like, a small set, ideally, like, 2 to 4 repeatable high-frequency workflows. For Slack, you kind of use grouping, this case will be the go-to-market team, and measure…
118 00:18:22.240 ⇒ 00:18:26.820 Miranda Wen: First run completion, repeat usage, and drop-off points.
119 00:18:26.910 ⇒ 00:18:40.629 Miranda Wen: for Mashimer in here, like, also, also for later, like, it’s just some… in ideal situations, I need to give more thoughts and see, like, what are some data we can actually collect about, so…
120 00:18:40.630 ⇒ 00:18:53.880 Miranda Wen: this is the part that’s, like, still kind of up in the air. I’ve been trying to schedule more time with Greg and also more learn about, like, what are the things we can collect, so just take those, success metrics as, like, a reference.
121 00:18:53.940 ⇒ 00:19:00.489 Miranda Wen: Yeah. And so we definitely hope, like, to see from the internal user group is about, like, their…
122 00:19:00.670 ⇒ 00:19:12.889 Miranda Wen: task… is the task completion better? Is this more… is the… is this very clear to them? And, is the confidence higher compared to with, like, a generic chat for the same work?
123 00:19:13.090 ⇒ 00:19:26.009 Miranda Wen: And the second thing is, like, faster time to value for new users, especially when the, compare it, like, towards what I mentioned, for example, with,
124 00:19:26.170 ⇒ 00:19:43.419 Miranda Wen: cursor codecs, are they, like, able to set up, execute, and value… validate, like, whether or not it’s, like, really useful for the first-time users? Then moving forwards, it’s more about, like, the governance part, whether, if this can…
125 00:19:44.810 ⇒ 00:19:51.080 Miranda Wen: Really help them with, like, the operational work, in terms of, like, approvals, controls.
126 00:19:51.140 ⇒ 00:20:09.589 Miranda Wen: And, I think also one thing very important here is, like, the knowledge we use. We want to, like, track with expert build workflows, so to get reused across users and teams, so that they can really turn the skills we use for specialist knowledge into, like, repeatable internal products.
127 00:20:09.790 ⇒ 00:20:22.879 Miranda Wen: And, that’s our goal here, and we’re gonna go out to business user adoption, and then internal, then the external. And I think the most important things here are, like, the knowledge.
128 00:20:22.880 ⇒ 00:20:31.940 Miranda Wen: Reuse workflow execution faster time to value will be, like, the top three we’re gonna, focus on during this experiment plan.
129 00:20:32.550 ⇒ 00:20:42.209 Miranda Wen: And I also have this, like, delivery milestone, just, like, to… as a guidance for us when we… later after this call, creating tickets.
130 00:20:42.330 ⇒ 00:21:00.200 Miranda Wen: And definitely hope to get, like, a lot more input on this to make sure, like, it will be helpful for us in creating tickets and starting on executing. So, Milestone wants to make the entry surface ready, so right now we can have, like, as I said, it has, like, the basic functionality for having a chat.
131 00:21:00.200 ⇒ 00:21:05.120 Miranda Wen: But in the future, we want… but the first milestone, definitely, like, make…
132 00:21:05.120 ⇒ 00:21:12.640 Miranda Wen: It’s the workflow fully ready, where we can, like, import, skills, pull skills, and…
133 00:21:12.700 ⇒ 00:21:30.110 Miranda Wen: work on the platform. And also, like, we… I think it’s also, like, a very good starting point to having the landing surface to, achieve the guided and also easy to discover, so to explain what Forge Lab is for, who it’s for for, and what…
134 00:21:30.120 ⇒ 00:21:36.480 Miranda Wen: workflow can be wrong first, and having this kind of basic navigation and workflow so that
135 00:21:36.640 ⇒ 00:21:42.989 Miranda Wen: When we go out to, like, the first internal user group, it wouldn’t be, like, fully blank to them.
136 00:21:43.160 ⇒ 00:21:52.660 Miranda Wen: And the milestone tool is to have, like, the, guided skill MVP run, so we want to test if the…
137 00:21:52.850 ⇒ 00:21:58.410 Miranda Wen: The skills, as I mentioned, 2 to 4 skills, can they run end-to-end without…
138 00:21:58.540 ⇒ 00:22:02.810 Miranda Wen: Very stably, as well as, like, high confidence,
139 00:22:02.880 ⇒ 00:22:19.279 Miranda Wen: And, have, like, reliable outputs consistently. And, and also, like, they don’t need to… like, it’s more… and it also will be, like, a lot more… hopefully, it will be, like, easier than,
140 00:22:19.400 ⇒ 00:22:20.920 Miranda Wen: Then, then cursor.
141 00:22:21.120 ⇒ 00:22:27.079 Miranda Wen: For the… for them to, like, run skill with the output and know what to do next.
142 00:22:27.200 ⇒ 00:22:35.930 Miranda Wen: And the third milestone is, like, the pilot workflow catalog live, so we wanna, like, Launch a usable first.
143 00:22:36.300 ⇒ 00:22:44.429 Miranda Wen: catalog for the initial internal user groups, a set of workflows. And, it’s basically as a…
144 00:22:45.070 ⇒ 00:22:57.520 Miranda Wen: considering, like, what we did before, and we also want to, like, them to have, like, very… each skill for them is, like, easy to discover. I think this would be, like, a very important part, where we know what’s really…
145 00:22:57.520 ⇒ 00:23:08.149 Miranda Wen: available for them, even when we build, like, I don’t know, 10, 20 skills in the future. We want to know, like, exactly what they know, what’s available for them, what’s the…
146 00:23:08.430 ⇒ 00:23:17.429 Miranda Wen: Use cases like, and what does, like, the input… what’s, like, an ideal prompt, input pattern, and what do you expect from the output?
147 00:23:19.480 ⇒ 00:23:38.509 Miranda Wen: So, yeah, that’ll be, like, our third point, and that will be also the time I feel like we are… keep on collecting more feedback from the first user groups. And moving on to the measurement loop life, so we wanted to have, like, the metrics and success criteria to be observed consistently for pilot users and pilot workflows.
148 00:23:38.540 ⇒ 00:23:51.990 Miranda Wen: And, for example, first round completion, return usage, time to first success, and workflow reuse on a regular basis to… as, like, a, like, North Star matrix for us, like, moving forward to see what our…
149 00:23:52.040 ⇒ 00:24:11.760 Miranda Wen: some things we’re lacking, what are the features development that we should lay out in the future? So yeah, so I think the major three decision gates is, like, whether interest surface is stable and clear, and second is, like, the getting skill-run experience reliable enough for us to have, like, this…
150 00:24:11.810 ⇒ 00:24:22.869 Miranda Wen: pilot workflow, catalog available for the first internal user group. And the third is probably the measurement and government… governance, whether or not it’s, like,
151 00:24:23.220 ⇒ 00:24:29.359 Miranda Wen: Live, up-to-date, and ideal for us to moving beyond the first internal cohort.
152 00:24:32.400 ⇒ 00:24:36.449 Uttam Kumaran: It’s like, I just want to pause here, like, I wanna make sure… I mean…
153 00:24:36.720 ⇒ 00:24:41.100 Uttam Kumaran: Pranav and B, and especially you, Pranav, like, you’re gonna be selling
154 00:24:41.300 ⇒ 00:24:43.890 Uttam Kumaran: product, I’m just not hearing that you’re, like.
155 00:24:44.200 ⇒ 00:24:57.149 Uttam Kumaran: I want to make sure you’re totally aware of, like, what this is, and, like, the way this can impact, like, clients, or your work. Like, if those aren’t clear, those are totally fair questions to ask.
156 00:24:57.330 ⇒ 00:25:00.960 Uttam Kumaran: Yeah.
157 00:25:00.960 ⇒ 00:25:07.019 Pranav Narahari: Yeah, I think… It probably was worth, like, me and Miranda, like, syncing before this.
158 00:25:07.020 ⇒ 00:25:10.120 Uttam Kumaran: No, no, I think this is… I think you should ask it here, like, that’s what this is.
159 00:25:10.120 ⇒ 00:25:10.520 Pranav Narahari: Of course.
160 00:25:10.670 ⇒ 00:25:11.490 Uttam Kumaran: You know, cause…
161 00:25:11.490 ⇒ 00:25:11.990 Pranav Narahari: Okay, yeah.
162 00:25:11.990 ⇒ 00:25:16.899 Uttam Kumaran: Think about the project, like, It should be extremely clear to, like, the least
163 00:25:17.110 ⇒ 00:25:25.119 Uttam Kumaran: the least AI-aware person at Brainforge should understand what we’re doing, in addition to the most AI-aware person.
164 00:25:25.220 ⇒ 00:25:29.179 Uttam Kumaran: Right? So that’s the purpose, so your questions are totally valid for this.
165 00:25:30.180 ⇒ 00:25:35.330 Pranav Narahari: Okay, yeah. Yeah, I think…
166 00:25:36.030 ⇒ 00:25:51.619 Pranav Narahari: I’m getting, like, the… the… I see, kind of, like, the problem with Cursor, and I can see, like, why for a customer-facing… or, yeah, for customers using a platform, like, how Cursor could be very…
167 00:25:51.830 ⇒ 00:26:06.409 Pranav Narahari: And kind of, like, kind of your words, like, maybe, like, overkill, and maybe also distracting to, like, what they’re trying to get from it. And so that’s where I see, based on your explanation here of, like, Forge Lab to be, like, fitting in.
168 00:26:06.520 ⇒ 00:26:15.760 Pranav Narahari: I’m… there’s certain things with Cursor, though, that… I think our…
169 00:26:16.370 ⇒ 00:26:22.599 Pranav Narahari: Things to consider, which is, you know, we have these various models, and…
170 00:26:23.220 ⇒ 00:26:29.979 Pranav Narahari: like, let’s say, like, the premium model that… or, you know, the composer model that are built into Cursor.
171 00:26:29.980 ⇒ 00:26:30.680 Miranda Wen: Yeah.
172 00:26:30.680 ⇒ 00:26:38.330 Pranav Narahari: Are we… are we confident that, like, with what we’re building here, that we’re going to be able to replicate that type of…
173 00:26:38.450 ⇒ 00:26:40.330 Pranav Narahari: behavior.
174 00:26:40.530 ⇒ 00:26:59.379 Pranav Narahari: where the output that we get from that, we all know is really good output. Have we been able to measure that same output from Forge Lab, or from, you know, the MVP or POC versions of Forge Lab, to say, yeah, we are meeting that same mark that we’re getting with Cursor?
175 00:27:01.880 ⇒ 00:27:08.439 Miranda Wen: Mmm… Yeah, that’s a really, really good question, and that’s… I think that’s, like, the…
176 00:27:08.770 ⇒ 00:27:17.390 Miranda Wen: top scene, we wanna… we definitely… what we want is, like, it’s because it’s, like, a unified service… surface, so, like, what we want to get is, like.
177 00:27:17.810 ⇒ 00:27:21.550 Miranda Wen: The quality… the quality shouldn’t, like…
178 00:27:21.650 ⇒ 00:27:31.729 Miranda Wen: be seeing a drop there, for sure. If anything, it can only be, like, more accurate and more tailored to, like, the user group we are rolling into.
179 00:27:32.270 ⇒ 00:27:38.869 Miranda Wen: I wonder, like… Yeah, I think… We’ll put that in…
180 00:27:38.870 ⇒ 00:27:46.779 Pranav Narahari: Hopefully the… Yeah, the pro of these models isn’t necessarily just that it has,
181 00:27:47.050 ⇒ 00:28:00.700 Pranav Narahari: the context of the, you know, the customer’s data. That’s another benefit of cursor that, you know, it takes advantage of the… the repos that you’ve had… you’ve added to your workspace.
182 00:28:01.870 ⇒ 00:28:18.710 Pranav Narahari: But… and this is… I don’t know how exactly they do this, too, but they say, at least, and maybe this is all marketing, and it’s not actually that difficult under the hood, that they’re leveraging all these different models to kind of create, like, the best type of,
183 00:28:20.040 ⇒ 00:28:23.540 Pranav Narahari: type of response to the user in chat, like.
184 00:28:23.540 ⇒ 00:28:23.880 Miranda Wen: Yeah.
185 00:28:23.880 ⇒ 00:28:31.349 Pranav Narahari: best type of agentic experience, or the best type of just chat experience. I think it’s worth diving into
186 00:28:31.510 ⇒ 00:28:39.600 Pranav Narahari: what is actually happening under the hood for, like, let’s say the premium model in the chat, or the Composer 2 model.
187 00:28:39.600 ⇒ 00:28:40.700 Miranda Wen: Hmm.
188 00:28:41.820 ⇒ 00:28:58.379 Pranav Narahari: what’s actually happening there. And then there’s probably certain, charts online, too, to, like, assess just, like, the difference in output for different types of tests of just, like, the raw, you know, Gemini model, or the raw…
189 00:28:58.380 ⇒ 00:29:04.949 Pranav Narahari: Codex model, as compared to Composer and Premium. So…
190 00:29:04.970 ⇒ 00:29:13.220 Pranav Narahari: I mean, if there’s a chart out there that shows, yeah, you know, if you just use an API key and you use, like, Gemini, it’s going to perform.
191 00:29:13.570 ⇒ 00:29:23.460 Pranav Narahari: without much difference, any noticeable difference, any significant difference to premium, then, yeah, I guess my question’s already answered there.
192 00:29:24.130 ⇒ 00:29:26.579 Pranav Narahari: But, yeah, I think it’s worth looking into that.
193 00:29:27.050 ⇒ 00:29:39.940 Miranda Wen: Yeah, yeah, yeah, totally. Thank you so much, Perna. I think that’s, like, a very, good direction to go. I think, like, on one case, like, we were saying, like, if there’s data and research to support that.
194 00:29:40.130 ⇒ 00:29:57.430 Miranda Wen: like, we just having, like, one API key for, like, one or two models, and it would have the same result as, like, what Cursor has for their marketed, like, composer model, as you said, the marketing thing. But also, like, if… if there is really a difference,
195 00:29:58.320 ⇒ 00:30:03.719 Miranda Wen: how do we, like, keep the results the same? And how do we measure that to make sure
196 00:30:03.850 ⇒ 00:30:10.170 Miranda Wen: it’s created, like, the same and even better value on that, I think. Yeah, I think I will definitely, like.
197 00:30:10.400 ⇒ 00:30:13.719 Miranda Wen: Know that, because…
198 00:30:14.110 ⇒ 00:30:20.409 Miranda Wen: But also, to answer your question, like, for the part about, like, where they have, like, the…
199 00:30:20.590 ⇒ 00:30:26.009 Miranda Wen: Repo, as well as, like, the database for the customers.
200 00:30:26.370 ⇒ 00:30:30.490 Miranda Wen: I think for open work, what we… we can’t… we…
201 00:30:30.600 ⇒ 00:30:34.059 Miranda Wen: at least for now, what the UI looks like.
202 00:30:34.270 ⇒ 00:30:43.900 Miranda Wen: and if I understand correctly, I think it can do the same thing. It has a worker part, it’s like a worker function, so you can…
203 00:30:44.060 ⇒ 00:30:58.489 Miranda Wen: also connect this to, like, your database and your report, if that’s the thing, like, you choose to do. So there’s a bit of, like, approval permission there, if that’s the thing you’re working for. So I think in that case, like.
204 00:30:58.760 ⇒ 00:31:09.049 Miranda Wen: it is… it shouldn’t be that different from cursor, but definitely, like, the composer model part is the thing I know that I need to, like, check and think about this.
205 00:31:11.460 ⇒ 00:31:12.080 Pranav Narahari: Yeah.
206 00:31:13.780 ⇒ 00:31:26.499 Brylle Girang: From an L&D perspective, Miranda, I think we need to flesh out these three main things. The first one is, like, this is going to be an internal tool, and we need to be clear on, like, why
207 00:31:26.580 ⇒ 00:31:40.100 Brylle Girang: Our people should invest time into actually learning Forge Lab, or the platform. The second is, how can you make sure that the adoption for the platform is going to be as smooth as possible?
208 00:31:40.420 ⇒ 00:31:44.160 Brylle Girang: And the third one is…
209 00:31:45.140 ⇒ 00:31:54.489 Brylle Girang: I think I need more clarity on, like, the why behind the platform, like, what’s the main problem, right, about cursor?
210 00:31:54.570 ⇒ 00:32:07.300 Brylle Girang: across all user groups, what are they experiencing, etc. So, for ILND, I’m focusing more on the internal teams and the people that are going to use this platform internally.
211 00:32:07.960 ⇒ 00:32:18.799 Brylle Girang: I would love to see that on the actual plan. I know that I think these four milestones are just phase one, but it would be really helpful if you could, like, see
212 00:32:19.040 ⇒ 00:32:24.150 Brylle Girang: like, the end-to-end goal for the project plan. I know that it doesn’t need to be that detailed, but
213 00:32:24.900 ⇒ 00:32:29.629 Brylle Girang: like, I want to see where we’re going to end up, like, 3 quarters, 2 quarters.
214 00:32:30.290 ⇒ 00:32:30.880 Miranda Wen: Hmm.
215 00:32:32.180 ⇒ 00:32:42.099 Miranda Wen: Yes, yes, okay, yeah, I can definitely… oops, sorry. I can definitely, like, include that. And also, like, for the first questions about, like,
216 00:32:42.270 ⇒ 00:32:54.090 Miranda Wen: like, for user groups to learn about Forge Lab. So I think the goal here for Forge Lab is, like, like, people don’t need to… really need to learn, like, the product educates people itself.
217 00:32:54.190 ⇒ 00:33:13.870 Miranda Wen: can, like, can, like, to be more chat-native, let’s say, like, when people open up chat, chat GPT, an 80-year-old can just type in questions, and it will just go on. And then they learn how to, like, use this, like, chat. So, that’s also why we wanted, like, to have those things, too.
218 00:33:13.940 ⇒ 00:33:23.329 Miranda Wen: be different from Curtis, so ideally, like, that should be… shouldn’t be… ideally, this product shouldn’t be… we don’t need to educate on people on how to use it. I think…
219 00:33:23.330 ⇒ 00:33:26.000 Brylle Girang: I think you need to put that there. I think that’s really helpful.
220 00:33:26.770 ⇒ 00:33:32.889 Miranda Wen: Okay, okay, okay, cool. Yeah, I will make, like, sure it’s, like, more highlighted and clear on that.
221 00:33:33.170 ⇒ 00:33:38.999 Miranda Wen: Yeah, and I think for the adoption part.
222 00:33:39.200 ⇒ 00:33:49.609 Miranda Wen: I feel like, I think, like, a thing plays an important role here is, like, really about, like, the skills. Potentially. Bea and I, we will collaborate to, like.
223 00:33:49.610 ⇒ 00:34:02.819 Miranda Wen: to the go-to-mark skill. So adoption here, what I’m seeing is, like, the things that… the other day I was talking to Robert, the AI automation part of that, during the workflow, that’s, like, important to his job, right? That’ll be 60% time I’m spending on that.
224 00:34:02.850 ⇒ 00:34:04.979 Miranda Wen: But they don’t have this kind of…
225 00:34:05.020 ⇒ 00:34:09.399 Miranda Wen: AI automation available. So I feel like that would be,
226 00:34:09.530 ⇒ 00:34:14.029 Miranda Wen: The part that ensures the adoptions, like, it’s… because they…
227 00:34:14.100 ⇒ 00:34:23.380 Miranda Wen: probably the first trauma, they need to use this skill, they want to try out this skill. So I feel like that would be, like, a good entry part for us to, like, start with.
228 00:34:23.400 ⇒ 00:34:35.529 Miranda Wen: it will be, more very much specific to what they need, a level up than what we have right now in Cursor for the skills that are more generic and can be used across the teams.
229 00:34:35.590 ⇒ 00:34:39.420 Miranda Wen: Yeah. And, for the third part of the end-to-end, yeah, I was… I think…
230 00:34:39.810 ⇒ 00:34:42.010 Miranda Wen: I definitely need to, like,
231 00:34:42.510 ⇒ 00:34:49.869 Miranda Wen: I really appreciate your question today about, like, why this right now, and what it’s gonna look like in the future.
232 00:34:49.889 ⇒ 00:35:05.229 Miranda Wen: I mean, because we still need to see, like, how it’s gonna play out during the experiment plan, during MVP, but I think that’s something, like, I could picture, so let’s have a direction to go, and, I will put, like, a section for that on this doc, yeah.
233 00:35:11.380 ⇒ 00:35:19.630 Uttam Kumaran: Yeah, I think my feedback, one, Miranda, is ask your question askers more questions.
234 00:35:19.890 ⇒ 00:35:30.569 Uttam Kumaran: I think in this meeting, if you’re finding yourself doing most of the talking, I think it’s not a good signal. So I think that’s probably feedback I’ll be giving for everybody that’s doing project reviews.
235 00:35:30.700 ⇒ 00:35:40.330 Uttam Kumaran: But don’t assume that everybody here is… knows even what the heck you’re talking about. Like, Rico hasn’t said a single word. That’s not good.
236 00:35:40.460 ⇒ 00:35:56.910 Uttam Kumaran: I guarantee Rico is a little bit lost on what’s going on. So, that’s one piece of feedback. I think second is you’re gonna have audience as part of these that are visual versus folks that want to read. Like, I am a big reader.
237 00:35:57.050 ⇒ 00:36:06.939 Uttam Kumaran: I don’t, like, I need to see every detail. I’m not really, like, the visuals… I’m not a visual thinker. Some people are very visual. You know, but also, like.
238 00:36:07.280 ⇒ 00:36:24.270 Uttam Kumaran: you have people on this call who are all, like, kind of using AI, like, think about if you were to go to, you know, someone on the design team, you know? Like, put yourself in their shoes, like, would they… would they, in just a few minutes, be like, oh my gosh, I want this today?
239 00:36:24.440 ⇒ 00:36:25.200 Uttam Kumaran: Right?
240 00:36:25.420 ⇒ 00:36:34.029 Uttam Kumaran: So that’s one piece. And then the last piece is, like, I think everything here, and I think, you know, I’m… I’m… Bea is also thinking about this, is…
241 00:36:34.160 ⇒ 00:36:42.909 Uttam Kumaran: I want to make sure that the Brainforge client is represented really, really heavily. You know, one thing that everybody here will tell you is.
242 00:36:43.010 ⇒ 00:36:53.950 Uttam Kumaran: like, I always mention that the client is number one, and then our team is a close number two, because without clients, there’s no… there’s no money, there’s no team. And…
243 00:36:54.210 ⇒ 00:37:06.729 Uttam Kumaran: So, and put another way, though, our people are our product, right? And so, really, I want to make sure that the Brainforge client is represented, even if it’s not part of this original, like, you know.
244 00:37:06.730 ⇒ 00:37:19.080 Uttam Kumaran: few-month plan, I want to make sure that, okay, it’s clear, like, how the client is going to get impacted, whether it’s, like, Brainforge team members can now move faster, or, hey, Pranav is actually going to go sell this.
245 00:37:19.210 ⇒ 00:37:33.009 Uttam Kumaran: you know, sell open work implementations to clients, or B is actually going to go train external clients as part of his L&D commercial plan, you know? So I want to make sure that that’s… that’s also really clear.
246 00:37:34.930 ⇒ 00:37:50.190 Miranda Wen: Mmm, I see, I see. Yes, yes, thank you for… so much for the feedback. Yeah, yeah, I think that’s the client’s part. I think, like, that kind of tweaked, like, this, doc a little bit. I think, like, where I should position it should be more of, like.
247 00:37:50.290 ⇒ 00:37:58.730 Miranda Wen: like, more pitch, more, commercial value in the future. And yes, I can definitely, like,
248 00:37:58.800 ⇒ 00:38:06.270 Miranda Wen: add those things and take notes on that, and that seems, like, very much valid. But also, I want to, like, get, like, a…
249 00:38:06.340 ⇒ 00:38:22.360 Miranda Wen: first impression, from the people here, like, do you guys think, like, as… for example, the field proposal, like, like, train external client as a part of the LMD commercial plan, or sell this product, and stuff, like, do you guys see where…
250 00:38:22.610 ⇒ 00:38:24.249 Miranda Wen: This project who, like.
251 00:38:24.530 ⇒ 00:38:32.530 Miranda Wen: Where do you guys see, like, fits the most, and has the most, like, promising direction to go? Just wanna, like, collect some thoughts here.
252 00:38:33.950 ⇒ 00:38:38.719 Brylle Girang: Sorry, I didn’t fully understand. Can you… can you… can you rephrase that a bit, Miranda?
253 00:38:38.980 ⇒ 00:38:48.980 Miranda Wen: Yeah, yeah, as Wten mentioned, as a part of the feedback, it’s like, he wanted to under… to see more, like, where… to put more emphasis on the client, like.
254 00:38:48.980 ⇒ 00:39:01.090 Miranda Wen: like, how are we gonna, like, use it at the end? That’s also what you mentioned, like, three quarters down. And, like, maybe we can use, like, to train external clients as a part of the L&A commercial plan for you, or maybe we can…
255 00:39:01.090 ⇒ 00:39:12.240 Miranda Wen: sell this to, for example, if the first internal user group, well, we sell it to, like, the go-to-market team for our client, or sell this as a part of the package for our service to provide.
256 00:39:12.370 ⇒ 00:39:16.159 Miranda Wen: Do you guys have, like, a f- first impression of, like.
257 00:39:16.430 ⇒ 00:39:21.539 Miranda Wen: Where could we see a good fit, or where we can most add value to?
258 00:39:22.140 ⇒ 00:39:27.680 Brylle Girang: Okay, well, for me, if this doesn’t work internally, then no client will want this.
259 00:39:27.910 ⇒ 00:39:28.390 Miranda Wen: Yeah.
260 00:39:28.640 ⇒ 00:39:35.059 Brylle Girang: I guess my main… my main ask here is that you need to make sure that
261 00:39:36.290 ⇒ 00:39:47.539 Brylle Girang: a Forge lab will 100% be adopted by our people, and it will 100% maximize what our people is doing, and that way, we have pretty good case study for our clients once the time.
262 00:39:47.540 ⇒ 00:39:47.930 Miranda Wen: Cool.
263 00:39:47.930 ⇒ 00:39:51.670 Brylle Girang: Especially us being an AI-first company, an AI-native company.
264 00:39:53.670 ⇒ 00:39:54.950 Miranda Wen: Totally, totally.
265 00:39:55.650 ⇒ 00:40:04.970 Pranav Narahari: Yeah, also, I think, even after, like, let’s say we do everything that we’re doing in Cursor right now in Forge Lab,
266 00:40:05.150 ⇒ 00:40:13.079 Pranav Narahari: we can expect that what our clients are gonna want to do within Forgelab is gonna be really similar to how we’re using it to…
267 00:40:13.120 ⇒ 00:40:32.289 Pranav Narahari: communicate with them, or to build things for them, they’re going to want a more self-service and not, like, a human-in-the-loop type of, experience, right? Right now, like, we’re using Cursor to just assist in our communications and our development, to then relay that over to the… to the client.
268 00:40:32.650 ⇒ 00:40:36.659 Pranav Narahari: How… so it’s almost like another level of…
269 00:40:37.000 ⇒ 00:40:45.620 Pranav Narahari: Maybe robustness, or certain guardrails that are just, like, being built out so that… we can…
270 00:40:45.810 ⇒ 00:40:51.769 Pranav Narahari: Feel comfortable about, like, hey, we take ourselves out of the picture, and then now it’s just self-service for them.
271 00:40:53.370 ⇒ 00:41:02.489 Pranav Narahari: Yeah, I’m… as you were saying this, I was trying to think of, like, certain examples where we’re doing that currently with our skills.
272 00:41:03.820 ⇒ 00:41:21.859 Pranav Narahari: But, yeah, I think that’s maybe something I need to think about more, or maybe it’s even something that we need to think about building for the future. How do we think about building something where we’re not in the picture as much? You know, as a CSO, like, things are just getting done for me, for the client.
273 00:41:23.440 ⇒ 00:41:39.719 Pranav Narahari: Yeah, so that’s, that’s going to be a, like, and to Bea’s point, like, it has to work for the CSLs, it has to work for the SLs, it has to work for whoever’s using Cursor. Internally here at Brainforge, before, like, we can feel confident that it’s gonna work for the client.
274 00:41:42.260 ⇒ 00:41:47.150 Miranda Wen: Yes, yes, totally. Yeah, thank you so much for your insight, Pranav. I think,
275 00:41:47.490 ⇒ 00:41:53.789 Miranda Wen: The part you raised is, like, also very, very great about, like, yeah, it definitely, like, a…
276 00:41:53.890 ⇒ 00:42:11.029 Miranda Wen: change of role, for… for job alone, like, for CSO. Yeah, it… the ideal goal is, like, to make it, like, more self-service, but I will say, like, what I picture right now, the CS goal… CSO goal, kind of, like.
277 00:42:11.120 ⇒ 00:42:16.249 Miranda Wen: be changed, like… like, ideally, we all have, like, different,
278 00:42:16.360 ⇒ 00:42:20.459 Miranda Wen: for Forge Lab, we all have, like, different packages, like, we’re… it’s just, like.
279 00:42:20.670 ⇒ 00:42:24.930 Miranda Wen: my sell right now. Different packages we’re gonna sell, and
280 00:42:25.030 ⇒ 00:42:29.089 Miranda Wen: When… so I think, like, CSO would be, like.
281 00:42:29.240 ⇒ 00:42:39.140 Miranda Wen: And the sales team will be the ones who make sure, like, our client gets the package they want the most, and having the package that they can cr…
282 00:42:39.260 ⇒ 00:42:46.990 Miranda Wen: they can maximize their value the most. And, as well as, like, I do believe
283 00:42:47.580 ⇒ 00:42:53.089 Miranda Wen: Like, there’ll still be, like, a little bit, like, twig and, adaptation that…
284 00:42:53.320 ⇒ 00:42:58.820 Miranda Wen: probably I see, like, towards, like, the customization, towards, like, the…
285 00:42:59.100 ⇒ 00:43:07.140 Miranda Wen: client we’re selling to, and I think that’s where gonna be, like, the CSO team, play a big role.
286 00:43:07.140 ⇒ 00:43:14.399 Pranav Narahari: That makes sense. So we’ll be building, if we’re using, like, the cursor terminology, like, skills for them.
287 00:43:14.730 ⇒ 00:43:17.850 Pranav Narahari: That they will then be able to trigger on their own.
288 00:43:18.320 ⇒ 00:43:23.920 Miranda Wen: Yes, yes. Ideally, right now, for example, definitely, like, we’re gonna launch this for 2 to 4.
289 00:43:24.030 ⇒ 00:43:28.309 Miranda Wen: Skills, as a repeatable workflow for the go-to-market team.
290 00:43:28.590 ⇒ 00:43:45.930 Miranda Wen: if these skills work well, definitely great, we’ll definitely pack those things, but also, like, we… building a skills is fast, so we… at the end, we’re, like, more so selling this platform. So, what are some other skills? What are some tweaks that are really tailored to their needs, their current workflows?
291 00:43:45.930 ⇒ 00:43:52.749 Miranda Wen: I think that would be the part where CSO and go-to-market team need to delve into and help with, yeah.
292 00:43:55.420 ⇒ 00:43:56.040 Pranav Narahari: Gotcha.
293 00:43:57.510 ⇒ 00:44:02.070 Miranda Wen: Yeah, yeah, Rico, like, do you have any questions, just to make sure, like.
294 00:44:02.260 ⇒ 00:44:08.279 Miranda Wen: Like, yeah, you can ask anything, so I can also know where I could, like, do better and be more clear about.
295 00:44:11.070 ⇒ 00:44:16.139 Rico Rejoso: Yeah, honestly, I didn’t get how… or how different it is from…
296 00:44:16.360 ⇒ 00:44:23.330 Rico Rejoso: cursor, so I don’t have any questions unless I try or read through this document, so I wasn’t able… wasn’t really participating a lot.
297 00:44:24.860 ⇒ 00:44:27.849 Miranda Wen: Oh, okay, gotcha, gotcha. Yeah, I think, like…
298 00:44:28.090 ⇒ 00:44:36.010 Miranda Wen: Yeah, maybe it would be… oh, that’s a very… also very good feedback. I think maybe I should, like, have, like, a bit more visual, and also, like, show what we…
299 00:44:36.150 ⇒ 00:44:43.379 Miranda Wen: have right now for the platform to have a better idea, for this, like, project review. Yeah, but…
300 00:44:43.500 ⇒ 00:44:46.690 Miranda Wen: Yeah, but basically, like…
301 00:44:47.180 ⇒ 00:45:06.789 Miranda Wen: basically, like, this will be… like, what you’re gonna see is not an IDE, it will be just, like, ChatGBT kind of, like, surface, and you can just directly go chat, pull on the skill set you need, and there’s no, like, a million… there’s no terminal, there’s no million, like,
302 00:45:07.500 ⇒ 00:45:16.860 Miranda Wen: data repos, like, you need to, like, find or think of, so it would be, like, definitely a lot simpler and easier to execute, ideally, yeah.
303 00:45:17.560 ⇒ 00:45:20.320 Rico Rejoso: Okay, and will this also live within the platform?
304 00:45:21.030 ⇒ 00:45:21.590 Rico Rejoso: Or…
305 00:45:21.590 ⇒ 00:45:27.869 Miranda Wen: Yes, right now, it’s… will be, at the very beginning stage, it will be, like, on open work, kind of, like.
306 00:45:28.010 ⇒ 00:45:35.470 Miranda Wen: Separate from the infrastructure, but yeah, after we roll it out, test it out at, like, it will be a part of the…
307 00:45:36.710 ⇒ 00:45:39.280 Miranda Wen: Yeah, it will be a part of the platform, I think.
308 00:45:40.320 ⇒ 00:45:40.980 Rico Rejoso: Okay.
309 00:45:44.590 ⇒ 00:46:03.159 Uttam Kumaran: Okay, great. I think, Miranda, one thing that would be helpful as a follow-up is, like, if you can compile feedback, I mean, this meeting, is recorded, and I think for everybody on this call, I think it would be great to summarize, sort of, like, what their feedback was, and then, like, what your next steps are, but I would love to see some…
310 00:46:03.460 ⇒ 00:46:21.840 Uttam Kumaran: you know, rapid iteration, if we can make updates to this plan with that feedback, and then… yeah, I think you heard some consistent themes from everybody, so I do feel like this is… this is really positive. I do think that, like, everybody here also is really tough to get a hold of, everybody’s pretty busy, so these sort of batch
311 00:46:21.900 ⇒ 00:46:25.859 Uttam Kumaran: ways of collecting feedback, or Slack is gonna be your best bet.
312 00:46:27.630 ⇒ 00:46:32.569 Uttam Kumaran: you know, especially now that I think you… you’ve met everybody at least once, and you have, like…
313 00:46:32.770 ⇒ 00:46:47.079 Uttam Kumaran: you know, you have some action items, so that would be helpful to see, you know, today, and sort of, again, I think one thing, like, I think it could be a good milestone is, like, we can actually run this maybe even tomorrow with another group.
314 00:46:47.630 ⇒ 00:46:53.460 Uttam Kumaran: And… again, their questions are probably gonna be really similar, so… yeah.
315 00:46:54.760 ⇒ 00:47:06.069 Miranda Wen: Okay, totally, yeah, thank you so much. Thank you so much, students. Thank you so much for everybody’s time here. Yeah, it was, like, really, really great to hear about this feedback. Yeah, I really appreciate everybody’s insights and questions here.
316 00:47:07.970 ⇒ 00:47:11.090 Pranav Narahari: I kind of have a… just a quick question on just, like, project reviews.
317 00:47:11.090 ⇒ 00:47:18.820 Uttam Kumaran: Yeah, I’m gonna… I’m gonna leave you… I’m gonna leave you guys, but you stay on, like, I don’t… I don’t… I don’t… you guys can stay on as long as you like, so I’m just gonna go to the next call.
318 00:47:18.820 ⇒ 00:47:20.320 Brylle Girang: Bye-bye. Great job, Rhonda.
319 00:47:23.480 ⇒ 00:47:30.030 Pranav Narahari: Yeah, cool. Miranda, we can actually just, probably talk about it later. I actually had more of a question for, like, Utam and B on this, so…
320 00:47:30.080 ⇒ 00:47:30.750 Miranda Wen: Yeah.
321 00:47:30.750 ⇒ 00:47:40.969 Pranav Narahari: Yeah, but yeah, I actually, I wonder, did you send this file to them beforehand, and did they get to read it, or was this kind of, like, their first time seeing this as well?
322 00:47:41.540 ⇒ 00:47:58.990 Miranda Wen: I think they’re… it’s, like, their first time as well. So I… I asked for a pull request for Uten, but I don’t think he ever, like, merged into this, like, so yeah, it’s, like, still on my badge, but I will ping him again, so, like, at least you guys can see it. And I also, like.
323 00:47:59.000 ⇒ 00:48:03.330 Miranda Wen: Well, slacking the platform about this stock, so… Yeah.
324 00:48:03.330 ⇒ 00:48:07.130 Pranav Narahari: Yeah, so if it’s in GitHub and it’s on your own branch, we should still be able to see it.
325 00:48:07.570 ⇒ 00:48:08.780 Miranda Wen: Oh, okay.
326 00:48:09.140 ⇒ 00:48:13.579 Pranav Narahari: Yeah. So, okay, that’s good to know. I mean, that would be my feedback, probably.
327 00:48:13.770 ⇒ 00:48:16.619 Miranda Wen: Yeah, Utam has seen it before, like, yeah.
328 00:48:16.620 ⇒ 00:48:19.090 Pranav Narahari: Oh, okay, yeah, so my…
329 00:48:19.330 ⇒ 00:48:30.179 Pranav Narahari: I guess… yeah, I would… maybe people are different in how they want to absorb it, but I would definitely want to read it before, and then come in with, like, a little bit more context and, like, questions off the bat.
330 00:48:30.300 ⇒ 00:48:37.180 Pranav Narahari: At least that’s what, you know, Utam and B did with my project review, so…
331 00:48:37.610 ⇒ 00:48:49.610 Pranav Narahari: Yeah, I think if I were to have done that, I would have gotten, like, a lot better questions for you. But yeah, okay, going forward, I think that’s gonna be just, like, feedback that I give to being UTAM on this process, so…
332 00:48:49.830 ⇒ 00:48:51.499 Pranav Narahari: Yeah, I’ll talk to them.
333 00:48:52.420 ⇒ 00:48:55.000 Miranda Wen: Cool, cool, that’s also good to know, yeah, because…
334 00:48:55.040 ⇒ 00:49:14.560 Miranda Wen: Because when… when I had this, call with, like, Utem on Wednesday afternoon, and he basically told me, oh, just edit this and… and pull this, like, Google… Google Calendar invite together, so I… I saw that was it, but also good to know, like, if that… that’s, like, works better, yeah.
335 00:49:15.190 ⇒ 00:49:27.919 Pranav Narahari: Yeah, I mean, maybe it just works better for me, because I think maybe with Utam, he’s more busy, and would just… maybe prefers to have, like, less context going into this call, so he can just, like, hear it all straight from me. Yeah, so…
336 00:49:28.050 ⇒ 00:49:29.560 Pranav Narahari: Yeah, I’ll just,
337 00:49:30.230 ⇒ 00:49:34.309 Pranav Narahari: I’ll sync with Utomo about this, but yeah, I need to hop to another call as well.
338 00:49:34.310 ⇒ 00:49:35.399 Miranda Wen: Okay. But yeah.
339 00:49:35.450 ⇒ 00:49:37.069 Pranav Narahari: This is great. I’ll talk to you soon.
340 00:49:37.070 ⇒ 00:49:41.310 Miranda Wen: Okay, thank you so much, Fernanda. Really appreciate you being here today.
341 00:49:41.940 ⇒ 00:49:43.390 Pranav Narahari: Totally, yeah, see ya.
342 00:49:43.390 ⇒ 00:49:44.790 Miranda Wen: Okay, have a great day.
343 00:49:45.470 ⇒ 00:49:46.279 Pranav Narahari: You too, bye.