Meeting Title: SpringForge Project Check-in Date: 2026-04-07 Meeting participants: Uttam Kumaran, Miranda Wen
WEBVTT
1 00:00:20.960 ⇒ 00:00:21.850 Uttam Kumaran: Hey!
2 00:00:21.930 ⇒ 00:00:23.020 Miranda Wen: Hang.
3 00:00:23.480 ⇒ 00:00:24.570 Uttam Kumaran: Hey, how’s it going?
4 00:00:24.870 ⇒ 00:00:28.660 Miranda Wen: Hi, been good, been good. Are you back in Austin right now?
5 00:00:28.960 ⇒ 00:00:29.740 Uttam Kumaran: Yes.
6 00:00:30.810 ⇒ 00:00:33.889 Miranda Wen: Awesome! How’s back home? Like, we’ve been traveling?
7 00:00:33.890 ⇒ 00:00:40.030 Uttam Kumaran: Yeah, it’s good. I love it here. I don’t like going anywhere else these days. Yeah, it’s just like…
8 00:00:40.460 ⇒ 00:00:43.150 Uttam Kumaran: Really chill here, so yeah, it’s good.
9 00:00:43.420 ⇒ 00:00:44.910 Miranda Wen: Awesome, awesome.
10 00:00:47.130 ⇒ 00:00:54.709 Uttam Kumaran: Well, yeah, I’m looking forward to kind of seeing what you’ve been working on. I mean, I think probably just to start is, like, I still feel like…
11 00:00:54.840 ⇒ 00:01:13.270 Uttam Kumaran: I’m not hearing enough from you in terms of, like, blockers and where you’re getting stuck on things. It’s been about a week, you know, since last time. I think we sort of had a formal kickoff, so looking forward to seeing, you know, how far you’ve gone today, but I think just a reminder from my side is that, like.
12 00:01:13.400 ⇒ 00:01:28.589 Uttam Kumaran: you see how vocal some of the people on the team are, and we’re not a very big, like, meeting culture, so as many things where I can see you working and your feedback and, like, help unblock you is gonna really be…
13 00:01:28.810 ⇒ 00:01:47.519 Uttam Kumaran: the way, you know, forward here. So, I’m excited to kind of see what you’ve produced, but even just to start there, I’m kind of wonder… I kind of wanted to ask you, like, how has the last week been? Like, what is… what have you been working on? And, like, what have you been learning? And then yet, like, even just reflecting on just that feedback, like, what do you think?
14 00:01:48.050 ⇒ 00:02:02.860 Miranda Wen: Yes, I think, like, I’ve been learning a lot about, like, onboarding on different tools. I know, like, before, I think I joined SpringForge, I never used cursor, but now, like, I feel like cursor is, like, very important to my work, and I learned, like, how to plug in.
15 00:02:03.020 ⇒ 00:02:11.109 Miranda Wen: different snack channels, different sources, and how to, like, select the models and ask questions. This is, like, definitely a very, very much powerful tool.
16 00:02:11.130 ⇒ 00:02:26.269 Miranda Wen: And also, I think I have, like, a better understanding about the project since we last time kicked off, and from there, like, I wrote this product plan. I call it, like, Forge Lab right now. Yeah, I don’t know what you think about it.
17 00:02:26.570 ⇒ 00:02:43.650 Miranda Wen: Yeah, and I think, like, I had, like, a pretty much, like… I tried to, like, include as much possible as thinking I was there, and also use the skill of the document console helped me to, like, polish the document, and also had the chance to talk to Robert and get to know what’s going on on the go-to-market side.
18 00:02:43.720 ⇒ 00:02:50.549 Miranda Wen: To initially validate my sauce. And I think, right now, I think I’m going
19 00:02:50.550 ⇒ 00:03:05.529 Miranda Wen: pretty well, but I will take your words about, like, I think that’s very good to know, to know, like, we’re more so, like, Slack culture, instead of meeting culture, so I will, try to update you for the questions I have, if you could answer, unless necessary, we will take meeting, if I understood it correctly.
20 00:03:06.000 ⇒ 00:03:24.310 Uttam Kumaran: Yeah, no, that’s… that’s helpful. I mean, I think I’m… you’re gonna see that I’m just gonna start becoming… I just am in so many things daily, so for me, to be able to visually see things getting shipped to the platform, or questions in Slack is… is, like, gonna be the… the biggest path forward.
21 00:03:24.590 ⇒ 00:03:38.499 Uttam Kumaran: Like, I think you can tell that this is a little bit of a… it’s a complicated product, and I’m totally aware of that, but what… one thing that makes people here successful is just, like, rapid iteration of, like, okay, I’ve made it this far, I’m jammed near.
22 00:03:38.620 ⇒ 00:03:49.149 Uttam Kumaran: I need some help, or like, even the stupidest question, whatever it is, you know? Because you’re gonna see, especially on this team, we’re gonna be the ones sort of pioneering, like.
23 00:03:49.220 ⇒ 00:04:03.090 Uttam Kumaran: how each of us can develop on the platform and build, sort of, like, this really great async culture, you know? And so that’s really, like, what I’d like to see, and so I think it’s good reflection, I think, to keep that in mind. I mean, simple ways is, like.
24 00:04:03.150 ⇒ 00:04:11.379 Uttam Kumaran: when you log in, just be like, hey, I’m working on this stuff today, or when you log off, I’m like, hey, this is what I kind of completed, or this is where I’m at.
25 00:04:11.500 ⇒ 00:04:16.079 Uttam Kumaran: Because I’m sort of, like, all over the place, so…
26 00:04:16.360 ⇒ 00:04:20.080 Uttam Kumaran: You know, think about managing up a little bit, like, how can you make sure that
27 00:04:20.230 ⇒ 00:04:28.280 Uttam Kumaran: it’s… there’s not a question in my mind on, like, what is Miranda working on, or how is Miranda’s day-to-day for laddering up to our objective, you know?
28 00:04:28.560 ⇒ 00:04:33.780 Miranda Wen: Mmm, I see. Okay, that’s good to know. Yeah, I will start doing that from now on. I think I…
29 00:04:33.780 ⇒ 00:04:34.470 Uttam Kumaran: Perfect.
30 00:04:34.470 ⇒ 00:04:44.139 Miranda Wen: Okay, like, a misunderstanding. I felt like it’s more so, like, oh, we checking on the meeting and pushing things forward. I think that makes… that makes sense to have, like, a daily update in that way.
31 00:04:44.140 ⇒ 00:04:56.260 Uttam Kumaran: Yeah, you’re gonna see that, like, I’m… the meetings I have are either the places in the company that are on fire, or the… wherever, like, the most money is, and, like, outside of that, it’s sort of just, like.
32 00:04:56.370 ⇒ 00:05:11.479 Uttam Kumaran: I’m pushing out so many things that I need to do, so as much as, like, exactly, you’re gonna find that it’s very different than a typical corporate environment here, where, like, we may kick things off, and there’ll be a meeting to get people together, but then the rest sort of just runs
33 00:05:11.480 ⇒ 00:05:25.880 Uttam Kumaran: through Slack and async, and so what is really helpful is if I’m able to see, like, and I’m able to give you feedback, right? Because I can Slack and message all day, and that’s really helpful, like, hey, I see you’re stuck here, you should go talk to this person, or take a look at this doc.
34 00:05:25.910 ⇒ 00:05:32.720 Uttam Kumaran: But… like, what I’m looking for is, like, we’re iterating on, like, a daily and hourly basis, you know?
35 00:05:33.500 ⇒ 00:05:38.119 Uttam Kumaran: to get feedback once every week. Think about, we only have… there’s only 52 cycles then.
36 00:05:38.590 ⇒ 00:05:39.100 Miranda Wen: Yeah.
37 00:05:39.840 ⇒ 00:05:44.549 Uttam Kumaran: And so that’s what I want to see, sort of, like, our feedback loop, you know, really increasing.
38 00:05:44.990 ⇒ 00:05:59.570 Miranda Wen: Okay, yes, yes, yes. Okay, yeah, I think that’s… that’s very good to know, I appreciate you sharing that, and yeah, I think I’ll… yeah, and I also saw, like, you have, like, so many… your calendar, like, so… so packed, and I think that makes sense, yeah. Yeah.
39 00:05:59.690 ⇒ 00:06:00.560 Miranda Wen: Okay, cool.
40 00:06:00.560 ⇒ 00:06:01.860 Uttam Kumaran: Cool, yeah, so…
41 00:06:01.860 ⇒ 00:06:02.260 Miranda Wen: First of all.
42 00:06:02.260 ⇒ 00:06:13.650 Uttam Kumaran: Yeah, feel free to share what you prepared today. I just sort of been locked in on a couple other things, so I can hand the floor over to you, and yeah, I would love to sort of get feedback wherever I can be helpful.
43 00:06:13.970 ⇒ 00:06:19.890 Miranda Wen: Okay, awesome. Yeah, let me… Let’s see the preview…
44 00:06:40.620 ⇒ 00:06:45.259 Miranda Wen: Oh, weird? Where do I have this, like, preview… in preview setting?
45 00:06:45.260 ⇒ 00:06:50.690 Uttam Kumaran: Yeah, so you can actually click on, this… Right here.
46 00:06:52.540 ⇒ 00:06:56.240 Uttam Kumaran: This, like, paper thing, yeah.
47 00:06:56.600 ⇒ 00:06:58.619 Miranda Wen: Oh, okay, okay, cool, cool, cool.
48 00:06:58.870 ⇒ 00:07:10.460 Miranda Wen: Yeah, so I created, like, a more detailed doc, and, related to, like, from your stand-up plan, what we discussed with Sam the other day, as well as, like, my own thinking. So,
49 00:07:10.980 ⇒ 00:07:23.899 Miranda Wen: So, I think, like, our goal here is, like, as you said, like, it’s a… this… so, first off, what do you think of, like, Forge Lab? Like, do you think that’s the name we should consistently using, or you prefer another name I can brainstorm further, too?
50 00:07:33.920 ⇒ 00:07:37.300 Uttam Kumaran: Just switched my speaker. Say it one more time, though, just the last part.
51 00:07:37.300 ⇒ 00:07:45.349 Miranda Wen: Yeah, do you… what do you think of the name for the lab? Should I just change all the docs in this name, or I can brainstorm more names about it?
52 00:07:45.350 ⇒ 00:07:50.010 Uttam Kumaran: Yeah, I guess I don’t have a strong opinion on…
53 00:07:50.140 ⇒ 00:07:58.210 Uttam Kumaran: the name, I guess I’m okay either way. Maybe it’s a good conversation for… honestly, it’s probably a good conversation with Bryle.
54 00:08:00.230 ⇒ 00:08:07.439 Uttam Kumaran: Because I’m not as creative. We basically had, like, work, like, Brainforge work. Forge Lab is also fine. Kind of like…
55 00:08:07.640 ⇒ 00:08:12.470 Uttam Kumaran: That, that… Worth… worth discussing with the team, or doing a poll, or something.
56 00:08:12.720 ⇒ 00:08:15.549 Miranda Wen: Okay, okay, cool, yeah.
57 00:08:15.550 ⇒ 00:08:21.500 Uttam Kumaran: It could be fine if people want to vote on it, but either way. We could also run with something for now and change it later.
58 00:08:21.720 ⇒ 00:08:29.360 Miranda Wen: Okay, cool, sounds good, yeah. I’ll talk to V later, too. So this, Forge Labs thing, and so, like.
59 00:08:29.580 ⇒ 00:08:44.530 Miranda Wen: right now, like, the unified service is to make the AI skills more usable product, easy to run, easy to share, easy to govern, and reliable enough for everyday worker across teams, and so it means to abstract away underlying complexity so people can just, like.
60 00:08:44.630 ⇒ 00:08:59.009 Miranda Wen: directly go on to execute the repeatable workflows faster, consistent at scale. That’s where… how I frame the goal. And for the scope right now, like, so we have this, like, entry surface…
61 00:08:59.200 ⇒ 00:09:18.290 Miranda Wen: to have, like, the skill creation, some workflows, and have, like, an early rollout to select an internal user group, as well as, like, develop features, as you… and Sam mentioned, like, hoping… hopefully, when we have, like, a internal user group being rolling out, I can also develop more features that pillar to them.
62 00:09:18.560 ⇒ 00:09:26.109 Uttam Kumaran: Yeah, I guess maybe a question, just to start, like, you now kind of hopefully had a week of, like, trying out Cursor, like.
63 00:09:26.980 ⇒ 00:09:33.170 Uttam Kumaran: I guess… in, like, I want to hear Miranda’s… I wanna hear your opinion, like.
64 00:09:33.270 ⇒ 00:09:35.170 Uttam Kumaran: What do you think about…
65 00:09:35.470 ⇒ 00:09:48.690 Uttam Kumaran: like, let’s say my hypothesis, right, and I want to hear your feedback. My hypothesis is that cursor, right, if we think back to our conversation last week, my hypothesis is that cursor is overkill.
66 00:09:49.190 ⇒ 00:09:56.809 Uttam Kumaran: for a majority of the work that people are doing right now, and in fact, I think the way the future is looking.
67 00:09:57.250 ⇒ 00:10:03.919 Uttam Kumaran: Most people are… there’s gonna be a group of people that are, like, developing the skills, the integrations, but most people are just gonna be using them.
68 00:10:04.160 ⇒ 00:10:09.669 Uttam Kumaran: And those are gonna be two separate audiences, right? And that’s my hypothesis. I’m kind of interested in, like.
69 00:10:10.590 ⇒ 00:10:18.249 Uttam Kumaran: Reflecting back on what we discussed on last week, like, how do you think your view has changed?
70 00:10:18.420 ⇒ 00:10:26.670 Uttam Kumaran: like, yeah, I’m just interested in, like, your hypothesis now on, like, just, products, surfaces in general.
71 00:10:27.110 ⇒ 00:10:36.179 Miranda Wen: Yes, I think, like, for me right now, outside… let’s kind of, like, relate to, like, how I was talking about the target users. So I’m thinking, like, for, like.
72 00:10:36.290 ⇒ 00:10:48.219 Miranda Wen: workflow operators, or admin team owners, including most of the teams, basically, other than the builder teams. I think, like, I agree with your hypothesis overall, where, like.
73 00:10:48.380 ⇒ 00:11:01.999 Miranda Wen: what we want is just, like, skills that are, like, readily available, and maybe just a little bit tweaked, like, maybe… or maybe even honest, honest, just, like, really fitting to their workflow, and…
74 00:11:02.320 ⇒ 00:11:06.850 Miranda Wen: that aren’t really available, I can just, like, pull them out and directly use that, instead of, like.
75 00:11:06.970 ⇒ 00:11:15.639 Miranda Wen: Because I do think, like, just thinking of, like, a good skill and make sure it works consistently does take some time. It’s…
76 00:11:15.640 ⇒ 00:11:18.049 Uttam Kumaran: That’s the product, almost, these days, right? It’s like…
77 00:11:18.050 ⇒ 00:11:19.689 Miranda Wen: Yeah, yeah, I mean, although.
78 00:11:19.690 ⇒ 00:11:20.260 Uttam Kumaran: Yeah.
79 00:11:20.260 ⇒ 00:11:23.899 Miranda Wen: But it’s, like, still kind of similar to, like, you developing a feature.
80 00:11:24.170 ⇒ 00:11:25.090 Uttam Kumaran: Yeah.
81 00:11:25.270 ⇒ 00:11:34.510 Miranda Wen: Yeah, I think, like, for example, the document console, if I need to, like, create a skill from scratch, I think it’s still… you need to have that idea, thinking, oh, this…
82 00:11:34.510 ⇒ 00:11:35.270 Uttam Kumaran: Yes.
83 00:11:35.270 ⇒ 00:11:39.159 Miranda Wen: help, and really test on different documents, it can help with that.
84 00:11:39.160 ⇒ 00:11:43.049 Uttam Kumaran: Which is what I did, which is… I… I used it a hundred times.
85 00:11:43.050 ⇒ 00:11:43.580 Miranda Wen: Yeah.
86 00:11:43.580 ⇒ 00:11:53.900 Uttam Kumaran: I was like, this sucks, I need this to be better, and then I, like, sort of… but then I… for all… most of the skills there, it’s sort of like a one shot, I tried it, it worked, we pushed it. But this is also where, like.
87 00:11:54.150 ⇒ 00:12:11.339 Uttam Kumaran: you know, it’s similar to… it’s similar to social media. A lot of people are gonna be creators, and a lot of people are gonna be consumers. And the ratio is gonna be, like, way off, right? Yeah. And not everybody’s good at making skills, and frankly, they shouldn’t… like, not everyone should make skills, or like…
88 00:12:11.860 ⇒ 00:12:17.130 Uttam Kumaran: I’m trying to… because our team is very small, right? And, like, there’s a lot of people opinionated, I’m like.
89 00:12:17.310 ⇒ 00:12:20.459 Uttam Kumaran: I’m trying to teach them to make skills that use skills.
90 00:12:20.610 ⇒ 00:12:26.840 Uttam Kumaran: But at some point, maybe that you have to, like, you can create one, and you can submit it, or you can submit a change, but ultimately.
91 00:12:26.990 ⇒ 00:12:31.190 Uttam Kumaran: I think there’s already a lot there, you know? Yes. Yeah.
92 00:12:31.690 ⇒ 00:12:37.860 Miranda Wen: Yes, I totally agree with that. And I think, like, for example, a person who’s on Southside, like.
93 00:12:38.000 ⇒ 00:12:50.389 Miranda Wen: I don’t think their… their focus should be, like, oh, even spend, like, 20% time in beauty and skill. I don’t think that that should be their job skill. They really should, like, pour their 100% attention in talking with clients.
94 00:12:51.160 ⇒ 00:12:52.170 Miranda Wen: And.
95 00:12:53.060 ⇒ 00:12:56.729 Uttam Kumaran: Preparing to talk to clients, or preparing strategy. Yeah. Yeah.
96 00:12:57.430 ⇒ 00:13:13.330 Miranda Wen: Yeah, I think that’s what… a reason that the first skill is just, like, since laid out, for example, before they need to, like, manually put into everything to solve for, but now, like, they don’t need to do that, and it’s, like, more clear, more intuitive, more easy to share. I feel like that’s the direction we’re going for.
97 00:13:13.770 ⇒ 00:13:14.390 Uttam Kumaran: Yeah.
98 00:13:14.640 ⇒ 00:13:15.390 Uttam Kumaran: Yeah.
99 00:13:15.520 ⇒ 00:13:16.270 Uttam Kumaran: Fair.
100 00:13:17.090 ⇒ 00:13:22.959 Miranda Wen: Yes, and for… So I’m saying all of Scope is definitely, like,
101 00:13:23.220 ⇒ 00:13:34.910 Miranda Wen: as you said, like, infrastructure integration, and later on, rolling on to, like, teams that are more computer-oriented, like engineers, as well as, like, self-serve onboarding that would be, like.
102 00:13:35.260 ⇒ 00:13:50.919 Miranda Wen: down the road. And some interest strategies right now, like, I feel like, I listed these top 3 as the, top internal users. And today, I actually just talked to Robert, and we were talking more on the business side.
103 00:13:50.920 ⇒ 00:13:51.470 Uttam Kumaran: Yeah.
104 00:13:51.470 ⇒ 00:13:56.159 Miranda Wen: Basically, I was talking about, I will also add this information more, because I was still.
105 00:13:56.160 ⇒ 00:14:02.049 Uttam Kumaran: Yeah, even, like, the quotes or stuff is gonna be good. You know, like, any quotes from those meetings.
106 00:14:02.280 ⇒ 00:14:07.930 Uttam Kumaran: I think is helpful. Like, maybe I can even give you one side of the coin.
107 00:14:07.990 ⇒ 00:14:17.710 Uttam Kumaran: like, today, I’m having… I’m having our client teams present project reviews. And let me show you an example of one, because it’ll kind of give you a sense of, like.
108 00:14:17.760 ⇒ 00:14:29.949 Uttam Kumaran: okay, what is the bar for, like, a good project plan at Brainforge? And maybe I’ll… I’ll share this with you. And then, very nicely, I also created, like, basically, like, a project reviewer skill.
109 00:14:30.040 ⇒ 00:14:34.880 Uttam Kumaran: And so, kind of like, if I was to show you the ABC…
110 00:14:35.600 ⇒ 00:14:41.200 Uttam Kumaran: project plan. So take a look at this. Let me know if you can… if you can see this.
111 00:14:51.720 ⇒ 00:14:52.350 Miranda Wen: Premium.
112 00:14:58.280 ⇒ 00:15:01.210 Miranda Wen: Okay, yeah, wait, let me share the screen.
113 00:15:03.400 ⇒ 00:15:11.800 Uttam Kumaran: So basically, we have a client called ABC Home and Commercial. Pranav is leading this as CSO, and this is his project plan.
114 00:15:11.860 ⇒ 00:15:24.519 Uttam Kumaran: And so, project plans, are an artifact that the CSO has to create, and has to go get approved by him, the service lead, and me.
115 00:15:24.790 ⇒ 00:15:43.089 Uttam Kumaran: and then ultimately the client to begin working. This is not something that we’ve done historically, but it is our current process, which just allows us to, like, raise the bar on that. Everybody’s on the same page on, like, what a project is. So, ultimately, the CSO has to go through what we’re calling project reviews. It’s almost like a defense.
116 00:15:43.200 ⇒ 00:15:56.040 Uttam Kumaran: It’s, like, almost like a thesis defense, where they… they come… they… we basically schedule a meeting, and they show up, and I ask a wide range of questions. It could be, hey, how does ABC make money?
117 00:15:56.570 ⇒ 00:16:02.490 Uttam Kumaran: It could be, why is step 4… why are we doing step 4 of, like, milestone 2?
118 00:16:02.730 ⇒ 00:16:12.460 Uttam Kumaran: Right? It could be, why did we choose this technology? Right? And so, ultimately, the client success owner has to either be able to answer that, or be able to, like.
119 00:16:12.770 ⇒ 00:16:21.449 Uttam Kumaran: basically be, like, in a few minutes, be able to find the answer from their notes, right? And I think this is a great example of, like.
120 00:16:21.900 ⇒ 00:16:22.800 Uttam Kumaran: Okay.
121 00:16:23.720 ⇒ 00:16:38.619 Uttam Kumaran: what you’re ultimately trying to do, and even… and this is where, like, I’m sort of playing two roles, right? One is for you to have confidence in looking at your project plan for… for Forge Lab on, like, okay, I can speak from everything to, like.
122 00:16:38.820 ⇒ 00:16:42.429 Uttam Kumaran: why this is needed, what is the ROI,
123 00:16:42.600 ⇒ 00:16:45.510 Uttam Kumaran: for Brainforge and for, like, sales.
124 00:16:45.690 ⇒ 00:17:03.920 Uttam Kumaran: And then all the way down to, like, okay, what is the technical milestones, right? So this is, like, probably a good outline. In the… in cursor, you can go in and say, like, how do I write a project plan? But there’s also a skill today that will mimic how I will, basically review your plan.
125 00:17:04.000 ⇒ 00:17:22.630 Uttam Kumaran: It, it, like, I don’t have exact questions, but I will basically… it’ll give you a sense of, like, okay, you’re missing this, or this is confusing, and it’ll allow you to, like, almost mimic, like, a project review, but I think this is just helpful, I wanted you to take a stab at, like, a format first.
126 00:17:22.839 ⇒ 00:17:34.030 Uttam Kumaran: And then show you that, like, there is a little bit of, like, a… there’s a good structure for you to mimic as well. Ultimately, we’re an internal team, right? So, there’s… the client is us, there’s, like.
127 00:17:34.190 ⇒ 00:17:42.909 Uttam Kumaran: We… I would like us to have also, like, how is this affecting revenue, or time, or some metric, but just wanted to share you with this.
128 00:17:42.980 ⇒ 00:18:00.430 Uttam Kumaran: Because I think it’s… I think it would actually be awesome, you know, and it could be later this week, for you to do, like, a project review, like, with a couple people from the team, right? Where me, B, Robert can come, and… and you sort of are like, hey, I want… I’m here to pitch this project.
129 00:18:00.600 ⇒ 00:18:06.790 Uttam Kumaran: And, like, here’s the scope, and here’s… here’s all the interviews and all the details on why this is the number one thing.
130 00:18:06.960 ⇒ 00:18:08.810 Uttam Kumaran: That we should be working at, you know?
131 00:18:09.270 ⇒ 00:18:12.279 Uttam Kumaran: And I think it’s helpful as, like, the,
132 00:18:12.470 ⇒ 00:18:18.079 Uttam Kumaran: As the product manager, you’re… you’re almost like the client success person, right?
133 00:18:18.200 ⇒ 00:18:25.240 Uttam Kumaran: For this. Luckily, it’s an internal project, but, like, that doesn’t mean we should hold it to any lower of a bar, you know?
134 00:18:26.760 ⇒ 00:18:28.030 Miranda Wen: Totally, totally.
135 00:18:30.320 ⇒ 00:18:32.349 Uttam Kumaran: Sorry, but I kind of derailed, but…
136 00:18:32.350 ⇒ 00:18:34.830 Miranda Wen: Man, no, it’s, it’s good.
137 00:18:35.580 ⇒ 00:18:36.490 Miranda Wen: Yeah.
138 00:18:37.510 ⇒ 00:18:39.969 Miranda Wen: Wait, let me… sure.
139 00:18:40.750 ⇒ 00:18:44.672 Miranda Wen: Back to my… Hold on the
140 00:18:47.350 ⇒ 00:19:03.330 Miranda Wen: Yeah, yeah, I was, like, talking with, I think, so far, I still think, like, going for the go-to-market team is a good strategy, and I talked with, like, Robar a little bit more. We were thinking, like, maybe some first repeatable workflow ones, like, for their…
141 00:19:03.460 ⇒ 00:19:05.739 Miranda Wen: Partnership management, because.
142 00:19:05.740 ⇒ 00:19:06.090 Uttam Kumaran: Yeah.
143 00:19:06.680 ⇒ 00:19:16.309 Miranda Wen: robot is having those things in the sales partnership channel, but it’s, like, not very so clear. I think, like, given that the Forge Lab
144 00:19:16.480 ⇒ 00:19:27.800 Miranda Wen: advantages, like, to have this, like, UI. We can have, like, some, like, stakeholder visuals or something. I still need to think about that, but maybe, like, partnership.
145 00:19:28.310 ⇒ 00:19:46.869 Miranda Wen: tracking which stage, what are stakeholders involved, could be one thing, and the other thing, probably would be the part about, like, their demos, how we can help with their… having, like, quick demos, customized demos that they can be readily presentable. I feel these two probably, like.
146 00:19:47.000 ⇒ 00:19:52.709 Miranda Wen: Some repeatable workflows we can test, we can add on skills and to let them try.
147 00:19:52.880 ⇒ 00:19:53.770 Uttam Kumaran: Yeah. Okay.
148 00:19:53.770 ⇒ 00:20:04.179 Miranda Wen: And then from there, we will see probably the marketing side, and then, strategy, CSO side, to gradually roll out. Yeah.
149 00:20:04.180 ⇒ 00:20:04.750 Uttam Kumaran: Okay.
150 00:20:04.900 ⇒ 00:20:07.830 Uttam Kumaran: Yeah. Yeah, I would almost, I would almost,
151 00:20:08.880 ⇒ 00:20:10.710 Uttam Kumaran: I would do 1 and 3.
152 00:20:12.030 ⇒ 00:20:12.810 Miranda Wen: Yeah.
153 00:20:16.090 ⇒ 00:20:23.900 Uttam Kumaran: Yeah, because you’re gonna… I’m trying to put in… I’m trying to put in the first, the thing, the people that will give you the most feedback.
154 00:20:24.210 ⇒ 00:20:24.980 Miranda Wen: Mmm.
155 00:20:24.980 ⇒ 00:20:29.209 Uttam Kumaran: Like, marketing, they’ll just use whatever we give them. Same with operations.
156 00:20:29.400 ⇒ 00:20:34.590 Uttam Kumaran: And recruiting, like, they’re not gonna be good, design partners.
157 00:20:34.770 ⇒ 00:20:35.200 Miranda Wen: Mmm.
158 00:20:35.200 ⇒ 00:20:38.920 Uttam Kumaran: So, business is good. I think you pick the business, and then for number 3,
159 00:20:39.080 ⇒ 00:20:41.659 Uttam Kumaran: I think you either pick… pick, like.
160 00:20:42.470 ⇒ 00:20:48.410 Uttam Kumaran: you either pick one person from each, or pick, like, just the CSOs.
161 00:20:49.080 ⇒ 00:20:53.699 Uttam Kumaran: But, yeah, I would… I… I would… I would flip, I would swap them.
162 00:20:54.370 ⇒ 00:20:54.980 Miranda Wen: Okay.
163 00:20:54.980 ⇒ 00:21:01.820 Uttam Kumaran: By the time it gets to marketing, you want it to sort of be ironed out, because they’re not gonna… they’re not gonna be able to give good feedback, you know.
164 00:21:02.780 ⇒ 00:21:17.189 Miranda Wen: That makes sense, yeah, I will… I will edit that and change that. And for the third, I think, like, yeah, I need to talk to their people more in this week, that’ll be my goal, like, to really understand, like, what are… maybe one for each word, just focus on CSO. Yeah.
165 00:21:17.410 ⇒ 00:21:31.300 Miranda Wen: And I actually want, like, to see, like, this use case hypothesis I have for the business users, and I want to see if, like, if you have any take on that, because I feel this is a really key part of the product, like, how we frame the problem.
166 00:21:31.300 ⇒ 00:21:36.600 Uttam Kumaran: Yeah, so Forge Lab packages repeatable, it workflows to guided, non-ID-friendly
167 00:21:38.080 ⇒ 00:21:41.749 Uttam Kumaran: The business users with recurring operational work will adopt it.
168 00:21:42.650 ⇒ 00:21:45.690 Uttam Kumaran: Users return… Yeah…
169 00:21:50.970 ⇒ 00:21:52.780 Uttam Kumaran: So I think I would probably…
170 00:21:53.280 ⇒ 00:22:01.449 Uttam Kumaran: make, like, a couple of wording changes, because I think this is almost… this could be similar to, like, what is the mission statement for this project, right?
171 00:22:01.450 ⇒ 00:22:01.880 Miranda Wen: Yes.
172 00:22:03.960 ⇒ 00:22:06.279 Uttam Kumaran: I think the first thing
173 00:22:06.840 ⇒ 00:22:13.940 Uttam Kumaran: for this whole thing is you should say, hey, we’re strictly doing this for internal use. That way, you can basically
174 00:22:14.430 ⇒ 00:22:19.920 Uttam Kumaran: start to narrow down, like, okay, what are we? We’re a consultancy, we have these roles, right? So…
175 00:22:20.310 ⇒ 00:22:21.200 Uttam Kumaran: you can…
176 00:22:21.420 ⇒ 00:22:27.520 Uttam Kumaran: That way, let’s say at a project where someone’s like, hey, is this for clients, or is this for us? You can say, it’s currently just for clients.
177 00:22:27.720 ⇒ 00:22:36.270 Uttam Kumaran: we’re currently just for us, and the clients are gonna come next, right? The second piece is non-IDE friendly. I would probably just say, like.
178 00:22:41.760 ⇒ 00:22:48.019 Uttam Kumaran: I don’t know, I don’t think… I think non-IDE friendly is a weird thing. I think you might want to just say, like.
179 00:22:48.600 ⇒ 00:22:56.289 Uttam Kumaran: It’s… it’s almost like… you want to have the same functionality as an IDE,
180 00:22:56.550 ⇒ 00:23:00.759 Uttam Kumaran: Minus all of the, like, training required, right?
181 00:23:00.960 ⇒ 00:23:01.610 Miranda Wen: Yeah.
182 00:23:01.610 ⇒ 00:23:10.639 Uttam Kumaran: One way is, like, you could always say, oh, ID is complicated, but okay, it’s complicated, but you can teach someone, right? But ultimately, you want it to be intuitive where you don’t need to teach.
183 00:23:10.900 ⇒ 00:23:13.600 Uttam Kumaran: Everybody’s familiar with ChatGPT already.
184 00:23:13.790 ⇒ 00:23:19.749 Uttam Kumaran: Right? ChatGPT already taught everybody how to use that product. We should just build on that.
185 00:23:20.090 ⇒ 00:23:21.030 Miranda Wen: Right?
186 00:23:21.030 ⇒ 00:23:21.870 Uttam Kumaran: Versus…
187 00:23:22.150 ⇒ 00:23:31.509 Uttam Kumaran: teaching a cursor-style IDE is always difficult, no matter what. So I think, like, something about that could be helpful. Like, my question here is, like.
188 00:23:31.760 ⇒ 00:23:35.249 Uttam Kumaran: why… why is non, like, why are IDEs
189 00:23:35.580 ⇒ 00:23:40.739 Uttam Kumaran: like, why not IDE, right? And so, a good answer there is, like, well.
190 00:23:40.990 ⇒ 00:23:50.259 Uttam Kumaran: like, we already noticed that it’s already a struggle. You, yourself, took a while to onboard, and ChatGPT already taught people, so, like, we might as well use that, right?
191 00:23:50.380 ⇒ 00:23:52.900 Uttam Kumaran: So, I think that, that could be helpful.
192 00:23:53.260 ⇒ 00:24:03.260 Uttam Kumaran: I think, I like the second thing, like, generic chat or specialist AI tooling. I may,
193 00:24:04.780 ⇒ 00:24:11.079 Uttam Kumaran: I may… I may honestly expand that to, like, list out some of those tools. Like, generic chat, what is that like?
194 00:24:11.260 ⇒ 00:24:18.600 Uttam Kumaran: Claude, chatGPT… Perplexity
195 00:24:18.710 ⇒ 00:24:26.410 Uttam Kumaran: and specialist AI tooling, I would list out a couple of those, too. Like, whatever. So that way you can… you pinpoint the exact
196 00:24:26.750 ⇒ 00:24:35.140 Uttam Kumaran: ones that you’re referencing. And the core bet is that users will return, I love this, like, easy-to-discover.
197 00:24:37.080 ⇒ 00:24:42.739 Uttam Kumaran: I think there’s two things that I really love in this. One is, like, you write easy to discover, which is great.
198 00:24:42.850 ⇒ 00:24:49.380 Uttam Kumaran: I also think that, you have Guided at the top.
199 00:24:49.720 ⇒ 00:25:01.820 Uttam Kumaran: Like, that… those two are things that I think you should just, like, highlight with a marker as, like, why this is different. Because not only is this just, like, an… it’s… yes, it looks… it’s easier, it’s simpler, fair.
200 00:25:02.030 ⇒ 00:25:03.840 Uttam Kumaran: but guided…
201 00:25:04.060 ⇒ 00:25:12.629 Uttam Kumaran: right? It’s something that’s really, really tough in cursor. Like, it doesn’t tell you, like, where to go to do anything, right? So that’s one thing. Imagine, like.
202 00:25:12.770 ⇒ 00:25:29.340 Uttam Kumaran: you log into… you log into Forge Lab, and it’s like, first thing, you should click here. Write your first skill, execute your first skill, right? Discover is also great. How can someone at Brainforge literally, like, discover skills? And… and, like, how can they, like, basically say.
203 00:25:29.650 ⇒ 00:25:35.370 Uttam Kumaran: like, take a look at all the stuff I did today. What are some skills, or what are some automations I should be using, right?
204 00:25:35.370 ⇒ 00:25:36.900 Miranda Wen: Exactly. That’s what I’m thinking, too, yeah.
205 00:25:36.900 ⇒ 00:25:40.429 Uttam Kumaran: Yeah, so those are huge, and, like, that’s what’s gonna really…
206 00:25:40.640 ⇒ 00:25:51.290 Uttam Kumaran: that’s how you go from, like, okay, this is just a replacement for Cursor, versus, like, this is better, because Cursor does not do that, right? Like, ChatGPT does not do that.
207 00:25:51.530 ⇒ 00:25:55.950 Uttam Kumaran: And the fact that we are developing this internally at Brainforge.
208 00:25:56.290 ⇒ 00:26:01.059 Uttam Kumaran: on a platform that is customizable is what allows us to do that. So I think both of those are, like.
209 00:26:01.630 ⇒ 00:26:06.819 Uttam Kumaran: I think the rest is great, some feedback, but those two things are how this is, like, much, much better, you know?
210 00:26:07.780 ⇒ 00:26:11.880 Miranda Wen: Awesome, awesome. Cool, cool, cool, yeah, I will make those edits, yeah.
211 00:26:12.300 ⇒ 00:26:13.140 Uttam Kumaran: Okay.
212 00:26:13.380 ⇒ 00:26:25.039 Miranda Wen: Yeah, so I have, like, an experimentation plan, and I feel like, for example, this one will be, like, something I will be doing this week and the following week with, Robert’s team.
213 00:26:25.290 ⇒ 00:26:26.180 Uttam Kumaran: Yeah. Okay.
214 00:26:26.310 ⇒ 00:26:32.309 Miranda Wen: And, and gradually… And also, like, this part,
215 00:26:33.110 ⇒ 00:26:39.900 Miranda Wen: to… I think probably this part… two part… oh, okay, actually, this three part… I think I should highlight that this three-part will be, like, the top.
216 00:26:40.090 ⇒ 00:26:46.700 Miranda Wen: focus… For this, like, first rollout and developing design partner.
217 00:26:46.700 ⇒ 00:26:47.400 Uttam Kumaran: Yeah.
218 00:26:47.910 ⇒ 00:26:49.330 Uttam Kumaran: Yeah, I think,
219 00:26:53.720 ⇒ 00:26:57.149 Uttam Kumaran: So you said the top two, and then the knowledge re… which ones?
220 00:26:57.360 ⇒ 00:27:01.660 Miranda Wen: Oh, the workflow execution, knowledge reuse, and business user adoption.
221 00:27:02.340 ⇒ 00:27:03.010 Uttam Kumaran: Okay.
222 00:27:04.260 ⇒ 00:27:10.920 Miranda Wen: Actually, we’re gonna use our foods later. Ashley just, like, yeah, the first, the first two and knowledge we use.
223 00:27:11.220 ⇒ 00:27:12.010 Uttam Kumaran: I agree.
224 00:27:12.310 ⇒ 00:27:12.860 Miranda Wen: Yeah.
225 00:27:13.740 ⇒ 00:27:20.509 Uttam Kumaran: Yeah, I think one of the big things that I’d love to see as part of this plan is, like, the measurement, you know? One thing in cursor.
226 00:27:20.740 ⇒ 00:27:25.180 Uttam Kumaran: Unless you have a cursor Enterprise plan, you can’t measure anything. Same with ChatGPT.
227 00:27:25.310 ⇒ 00:27:30.320 Uttam Kumaran: So, like, there’s no way for us to log who’s using it, what they’re using it for, what are their prompts.
228 00:27:30.540 ⇒ 00:27:35.940 Uttam Kumaran: You know, and I think that’s a big piece of this that also is, like, How does, like.
229 00:27:36.370 ⇒ 00:27:43.700 Uttam Kumaran: You know, regardless of what our hypothesis is now, the fact that we can now see the data allows us to, like, get to the next version
230 00:27:43.940 ⇒ 00:27:50.770 Uttam Kumaran: or next hypothesis faster, right? All of my hypothesis now is built on asking people and using it.
231 00:27:50.960 ⇒ 00:27:54.599 Uttam Kumaran: But it’s not built on any data. I know how much we’re spending.
232 00:27:54.730 ⇒ 00:27:59.819 Uttam Kumaran: We’re spending a lot of money. I know how many tokens are getting used, but, like, I don’t really know, like.
233 00:28:00.410 ⇒ 00:28:04.719 Uttam Kumaran: The nitty-gritty, like, what skills are being used, what are the prompts?
234 00:28:04.960 ⇒ 00:28:08.559 Uttam Kumaran: Like, how often… Things like that.
235 00:28:09.430 ⇒ 00:28:12.180 Uttam Kumaran: How long does something take? Things like that, you know?
236 00:28:12.540 ⇒ 00:28:15.119 Miranda Wen: Ethan, how can we get that, like…
237 00:28:15.580 ⇒ 00:28:18.399 Uttam Kumaran: Yeah, so, I mean, this is where I… I think…
238 00:28:18.760 ⇒ 00:28:22.209 Uttam Kumaran: You, as… with your product manager hat, don’t have to be like.
239 00:28:22.530 ⇒ 00:28:25.240 Uttam Kumaran: I need to do that, so don’t worry about the how.
240 00:28:25.420 ⇒ 00:28:38.189 Uttam Kumaran: Right? So when we switch to talking about technical, we will talk about that. I think me, you, and Sam will meet on, like, how do we get there? But I think you should make sure that it’s a requirement that we measure
241 00:28:38.330 ⇒ 00:28:39.029 Uttam Kumaran: you know.
242 00:28:39.030 ⇒ 00:28:39.620 Miranda Wen: Right.
243 00:28:39.620 ⇒ 00:28:41.900 Uttam Kumaran: Those things, and that you, yourself.
244 00:28:42.060 ⇒ 00:28:53.029 Uttam Kumaran: you can log in, and within, like, there’s an… you can just see it at, like, near real time, like, what people are doing, you know? And I think that’s a fair requirement for you to put in there.
245 00:28:53.890 ⇒ 00:29:02.060 Miranda Wen: Okay, okay, cool, cool, cool. Yeah. Yes, yeah, that’s definitely something, like, I care a lot about, like, because I do have some, like, success criteria.
246 00:29:02.280 ⇒ 00:29:08.600 Miranda Wen: That I think that are very important as, like, guidelines for us to really measure if it’s worth it, yeah.
247 00:29:09.050 ⇒ 00:29:10.460 Miranda Wen: Yeah,
248 00:29:11.120 ⇒ 00:29:16.270 Miranda Wen: Yeah, and I have, like, some open questions, but I guess, like, we run out of time for this meeting, so…
249 00:29:16.270 ⇒ 00:29:18.190 Uttam Kumaran: No, no, I can keep going, let’s keep going.
250 00:29:18.190 ⇒ 00:29:20.329 Miranda Wen: Oh, okay, okay, that would be great, yeah.
251 00:29:20.550 ⇒ 00:29:25.930 Miranda Wen: Yeah, I think I have some open questions I want to… like,
252 00:29:26.300 ⇒ 00:29:29.730 Miranda Wen: Go over, especially the first 5?
253 00:29:29.850 ⇒ 00:29:36.229 Uttam Kumaran: Okay, let’s go through it. I guess, like, my request here is I would love for you to answer first.
254 00:29:36.450 ⇒ 00:29:36.900 Miranda Wen: Yeah.
255 00:29:37.130 ⇒ 00:29:38.409 Uttam Kumaran: Yeah.
256 00:29:38.410 ⇒ 00:29:43.380 Miranda Wen: Yeah, I think for now, like, for the first questions, I think my idea is, like.
257 00:29:43.970 ⇒ 00:29:57.520 Miranda Wen: And that’s the hypothesis… I was agreeing with Xiang, I think we kind of covered that. I think it’s more so for the operator side execution. So have, like, all the skills packet ready for them to use to avoid them from, like, the orchestration
258 00:29:57.920 ⇒ 00:30:01.250 Miranda Wen: orchestration burden, I would say. That would be, like, our…
259 00:30:01.430 ⇒ 00:30:04.629 Miranda Wen: Have all this great time, and what we’re batting on.
260 00:30:04.840 ⇒ 00:30:15.770 Miranda Wen: And, and the first workflows that are high frequency, high value, and simple enough to package well for the initial rollout is, like, what I talked with Robart. The partnership thing, as well as, like.
261 00:30:15.890 ⇒ 00:30:21.060 Miranda Wen: The demo thing. What’s the demo thing pop-up gonna look like? I’m still not sure, but he…
262 00:30:21.060 ⇒ 00:30:22.350 Uttam Kumaran: Yeah, I agree, yeah.
263 00:30:22.380 ⇒ 00:30:23.060 Miranda Wen: Yeah, I was thinking.
264 00:30:23.060 ⇒ 00:30:26.150 Uttam Kumaran: I also think that, like, you should,
265 00:30:26.470 ⇒ 00:30:32.429 Uttam Kumaran: message B in the channel, Bryle, he, has some data on, like.
266 00:30:32.680 ⇒ 00:30:35.740 Uttam Kumaran: I think he has some data on, like, what skills are being used.
267 00:30:36.200 ⇒ 00:30:40.190 Uttam Kumaran: Or he may be also able to tell you, like.
268 00:30:40.620 ⇒ 00:30:43.690 Uttam Kumaran: These are the first sets of skills we’d like to move over there.
269 00:30:44.100 ⇒ 00:30:45.280 Uttam Kumaran: Oh.
270 00:30:45.980 ⇒ 00:31:02.820 Uttam Kumaran: Because, for example, there’s some skills that everybody’s using on every team, and then there’s also going to be features that we’re just working on just for, like, sales, right? Yeah. So, I think it’s helpful, like, on the… if I was to… on the engineering side, it would just be helpful for me to have a list of, like, here’s the functionality we need, you know, and then I can start
271 00:31:03.200 ⇒ 00:31:04.570 Uttam Kumaran: Muscling through that.
272 00:31:06.290 ⇒ 00:31:07.380 Miranda Wen: Okay, yes.
273 00:31:08.210 ⇒ 00:31:12.550 Miranda Wen: Yeah, so for that, it’s, like, in the AI channel, right?
274 00:31:13.290 ⇒ 00:31:18.870 Uttam Kumaran: I would… you could put it in the platform channel. I think we should just centralize there, basically.
275 00:31:18.870 ⇒ 00:31:20.380 Miranda Wen: Cool, cool, cool, okay.
276 00:31:22.070 ⇒ 00:31:34.279 Miranda Wen: Yeah, I was also thinking, like, this, like, timeline is realistic for key milestones, such as, like, pilot setup, first user testing, early validation. Like, do you have, like, a… like, a timeline, like, in your mind?
277 00:31:34.680 ⇒ 00:31:38.409 Uttam Kumaran: I would say as soon… as soon as humanly possible. So…
278 00:31:38.680 ⇒ 00:31:44.270 Uttam Kumaran: I think if you tell me one month, then I will ask, why not two weeks?
279 00:31:44.920 ⇒ 00:31:50.410 Uttam Kumaran: I think this is where I almost want you to, like, build it up,
280 00:31:51.810 ⇒ 00:31:57.659 Uttam Kumaran: I mean, honestly, what’s helpful is, like, if you can build out the linear, like, milestones.
281 00:31:58.740 ⇒ 00:32:01.149 Uttam Kumaran: Then we can shift into tasks.
282 00:32:01.480 ⇒ 00:32:03.419 Uttam Kumaran: Okay. And then it’s basically, like.
283 00:32:04.260 ⇒ 00:32:06.259 Uttam Kumaran: However fast we can do them.
284 00:32:06.610 ⇒ 00:32:10.859 Uttam Kumaran: For example, we set a month deadline, then it’ll get done in a month.
285 00:32:10.960 ⇒ 00:32:19.430 Uttam Kumaran: But, like, I’m gonna push on some of the ways that we do the engineering for this. So instead, what I want you to think about is, like.
286 00:32:20.000 ⇒ 00:32:23.870 Uttam Kumaran: and this is something that’s new, I feel like, in product management, is don’t…
287 00:32:24.190 ⇒ 00:32:27.189 Uttam Kumaran: Assume the timeline is however fast we can do it, right?
288 00:32:28.120 ⇒ 00:32:32.629 Uttam Kumaran: So instead, I think it’s helpful is to have, like, clear milestones.
289 00:32:33.320 ⇒ 00:32:40.299 Uttam Kumaran: of, like, okay, this is the first milestone, we will know that we reached the milestone when XYZ is true.
290 00:32:40.820 ⇒ 00:32:45.010 Uttam Kumaran: When you can go in, you can do this, you can run this, you can see this.
291 00:32:45.210 ⇒ 00:32:46.989 Uttam Kumaran: That’s gonna be helpful.
292 00:32:47.550 ⇒ 00:32:52.779 Uttam Kumaran: And then… If we can just see all those milestones.
293 00:32:53.270 ⇒ 00:32:57.919 Uttam Kumaran: then I think that’s perfect, because then we’ll ticket it all out, and then we’ll just run as fast as we can.
294 00:32:58.300 ⇒ 00:32:59.100 Miranda Wen: Okay.
295 00:32:59.210 ⇒ 00:33:02.729 Miranda Wen: Okay, cool, yeah, I would note down on… no doubt on that.
296 00:33:02.730 ⇒ 00:33:07.179 Uttam Kumaran: Ultimately, there’s some friction, right? You have to meet with people, onboard, things like that, but, like.
297 00:33:07.340 ⇒ 00:33:13.330 Uttam Kumaran: on the engineering side, I don’t know, like, if you give me all the tickets, I may just try to one-shot as much as possible.
298 00:33:13.510 ⇒ 00:33:23.540 Uttam Kumaran: overnight. So, like, I think for me, more… that’s… more that’s important is, like, the requirements need to be really, really good, because I’m gonna have AI do as much of it as possible.
299 00:33:23.710 ⇒ 00:33:25.839 Uttam Kumaran: And I don’t want it to guess, like.
300 00:33:26.300 ⇒ 00:33:29.860 Uttam Kumaran: When are we done, right? For each… for each phase?
301 00:33:30.280 ⇒ 00:33:30.950 Miranda Wen: Yeah.
302 00:33:31.330 ⇒ 00:33:40.600 Uttam Kumaran: And there’s actually a lot in the platform also about, like, ticket requirements and how to think about, like, milestones versus initiatives versus projects and linear.
303 00:33:40.920 ⇒ 00:33:49.570 Uttam Kumaran: But like, again, I think start with this broader plan, and then move on to that. So, yeah, that’s sort of the answer on the timeline piece.
304 00:33:49.980 ⇒ 00:33:50.760 Miranda Wen: Hmm.
305 00:33:50.900 ⇒ 00:33:53.180 Miranda Wen: Okay, okay, sounds good, sounds good.
306 00:33:53.550 ⇒ 00:33:57.129 Miranda Wen: Yeah. Yeah, I think, like, my…
307 00:33:58.000 ⇒ 00:34:07.419 Miranda Wen: fifth question is kind of related to that. I also have to jump to the fifth question first. So for the, like, what do you think, like, are some, like, prerequisites that…
308 00:34:07.680 ⇒ 00:34:13.959 Miranda Wen: are good for, like, when we’re shipping new features, for example, for the first user group, and…
309 00:34:14.320 ⇒ 00:34:31.869 Miranda Wen: And should, feature releases beyond the initial pilot be managed through separate product planning documents? So does features consider us, like, part of the product plan, or it’s, like, more so, like, more… more individualistic, separately speaking? I don’t know, like…
310 00:34:32.159 ⇒ 00:34:37.159 Uttam Kumaran: Yeah, I mean, this is where I think, like, in a traditional world, like, for…
311 00:34:37.449 ⇒ 00:34:39.509 Uttam Kumaran: Like, product requirements docs.
312 00:34:39.619 ⇒ 00:34:44.029 Uttam Kumaran: you would basically be like, these are all the things I’m confident in, like, let’s not deviate.
313 00:34:44.289 ⇒ 00:34:46.379 Uttam Kumaran: because of, I think, AI,
314 00:34:46.509 ⇒ 00:34:59.359 Uttam Kumaran: I actually don’t mind if, like, you give me this plan, we rip through it, and then next week you’re like, I got some feedback, we need to tweak it, because you’re gonna use AI to, like, update the plan, right? Yeah. So…
315 00:35:00.389 ⇒ 00:35:10.489 Uttam Kumaran: I think, again, what I’ll say is, like, as soon as we can sort of approve the plan, and you can run a project review, I’m good to start building at that point.
316 00:35:10.659 ⇒ 00:35:13.589 Uttam Kumaran: I think…
317 00:35:14.419 ⇒ 00:35:19.809 Uttam Kumaran: I will let you think about more of, like, how do we communicate? Because one thing is, like.
318 00:35:20.079 ⇒ 00:35:30.619 Uttam Kumaran: there’s gonna be change management here. This thing is now available. This, like, I wanna show… I wanna… we need to run, like, a webinar to teach people. I’m gonna leave that to you.
319 00:35:31.779 ⇒ 00:35:42.569 Uttam Kumaran: So I think if there is, like, a way in terms of, like, literally the marketing to the team that you want to organize around, like, every Friday we do a release, or every Monday we do a release.
320 00:35:43.120 ⇒ 00:35:54.419 Uttam Kumaran: totally game for however you want to do it, but we’re gonna try to ship every day, you know? But then, if you’re like, hey, I don’t want to, like, change things every day, let’s just consider once a week.
321 00:35:54.570 ⇒ 00:35:58.060 Uttam Kumaran: then I’m… I’m also open to discussing that.
322 00:35:58.580 ⇒ 00:36:07.649 Uttam Kumaran: So, I think the answer to this is, like, as soon as the project sort of plan, we feel confident, Sam signs off, and, like, a couple other people sign off.
323 00:36:07.840 ⇒ 00:36:10.799 Uttam Kumaran: then we’ll… we’ll just run, and then I’ll…
324 00:36:11.130 ⇒ 00:36:19.139 Uttam Kumaran: I think this is where, like, I’m gonna start to be the primary developer, and then I think I’ll start to loop you in so that you can also start developing features.
325 00:36:19.520 ⇒ 00:36:19.900 Miranda Wen: Alright.
326 00:36:19.900 ⇒ 00:36:22.620 Uttam Kumaran: Because then you’ll be kind of like a one-man army on this, so…
327 00:36:23.290 ⇒ 00:36:26.249 Miranda Wen: Yes, okay, okay, sounds good, sounds good.
328 00:36:26.480 ⇒ 00:36:38.630 Miranda Wen: Yeah, and also, I think I have, like, very PM questions, yeah. So I think it’s, like, this kind of user… although I had a pretty great conversation with Shilbert, but he… he…
329 00:36:38.850 ⇒ 00:36:45.890 Miranda Wen: we talk about, like, different things they’re… they’re doing, like, we… so that’s how we exclude, like, the Alban part of…
330 00:36:45.890 ⇒ 00:36:46.540 Uttam Kumaran: Yeah.
331 00:36:46.540 ⇒ 00:36:52.629 Miranda Wen: things and stuff. But also, I think, like, that I know you were doing product, right? I was thinking, like, what are some, like.
332 00:36:52.790 ⇒ 00:36:55.179 Miranda Wen: When we were doing user interviews, or.
333 00:36:55.180 ⇒ 00:36:55.870 Uttam Kumaran: Yes.
334 00:36:55.870 ⇒ 00:37:01.620 Miranda Wen: What are some key questions you think for this case we should ask them to really make sure, like, this, like.
335 00:37:02.190 ⇒ 00:37:06.540 Miranda Wen: It’s a product they would actually use, and and,
336 00:37:07.510 ⇒ 00:37:16.080 Miranda Wen: I’m basically just, like, what are… doing the user interviews, what are some questions, or how do I really collect the things that are actually important for the product?
337 00:37:16.140 ⇒ 00:37:19.070 Uttam Kumaran: Yeah, it’s a great question. I mean, I honestly…
338 00:37:19.200 ⇒ 00:37:25.259 Uttam Kumaran: will… I think the densest way for you to learn these, there’s, like.
339 00:37:25.630 ⇒ 00:37:30.189 Uttam Kumaran: There’s basically, like, two books that will answer, like, all of these questions. I don’t know if you’re…
340 00:37:30.370 ⇒ 00:37:32.050 Uttam Kumaran: If you’re a big reader.
341 00:37:32.530 ⇒ 00:37:33.489 Miranda Wen: I am, I am.
342 00:37:33.490 ⇒ 00:37:37.620 Uttam Kumaran: Okay, then… I… I read this book…
343 00:37:37.620 ⇒ 00:37:38.690 Miranda Wen: bookcase.
344 00:37:38.690 ⇒ 00:37:41.499 Uttam Kumaran: Really? What are you reading right now? Are you reading anything interesting?
345 00:37:41.770 ⇒ 00:37:49.399 Miranda Wen: I’m now reading, like, a book, basically talking about, like, how does, Netflix view their culture.
346 00:37:49.400 ⇒ 00:37:50.809 Uttam Kumaran: Okay, yeah.
347 00:37:50.810 ⇒ 00:37:53.520 Miranda Wen: Yeah, yeah. But anyway, yeah, continue, yeah.
348 00:37:53.870 ⇒ 00:37:59.530 Uttam Kumaran: Yeah, so I… let me send you this, let me send you this, buck.
349 00:38:00.020 ⇒ 00:38:07.330 Uttam Kumaran: I think… this is the book I read when I first started as a product manager, and it, like.
350 00:38:07.850 ⇒ 00:38:10.940 Uttam Kumaran: answered, like, 98% of the questions I had.
351 00:38:11.260 ⇒ 00:38:16.159 Uttam Kumaran: I just shared in the Zoom… in the Zoom channel. I’d be happy to buy for you, by the way.
352 00:38:16.350 ⇒ 00:38:17.110 Uttam Kumaran: Okay.
353 00:38:17.110 ⇒ 00:38:18.880 Miranda Wen: Thank you.
354 00:38:19.710 ⇒ 00:38:23.690 Uttam Kumaran: So that’s a really great one. And then,
355 00:38:24.280 ⇒ 00:38:29.550 Uttam Kumaran: this is another one, Lean Analytics, but I would read, the…
356 00:38:33.030 ⇒ 00:38:35.819 Uttam Kumaran: I would read that first one.
357 00:38:38.820 ⇒ 00:38:43.429 Miranda Wen: Okay, awesome. Wait, I don’t think I’ve ever heard of all these two books before.
358 00:38:43.430 ⇒ 00:38:47.510 Uttam Kumaran: So you should… so you should read this. I’ll read it with you again, if you want.
359 00:38:47.510 ⇒ 00:38:48.639 Miranda Wen: Oh, yeah.
360 00:38:48.640 ⇒ 00:38:58.190 Uttam Kumaran: So, if you buy it and get it shipped, or whatever, just submit a receipt to the operations team, and they’ll reimburse you.
361 00:38:58.500 ⇒ 00:39:06.269 Uttam Kumaran: But I’m telling you, you read this, you’ll be like, holy shit, I, like, I get it. This is, like, the best book ever on product
362 00:39:07.240 ⇒ 00:39:07.770 Uttam Kumaran: Yeah, it’s really cool.
363 00:39:07.770 ⇒ 00:39:09.180 Miranda Wen: Very excited for that.
364 00:39:09.180 ⇒ 00:39:22.550 Uttam Kumaran: Really good. Okay, so if you’re a reader, then what you’re gonna find is that I’m sort of a fraud. Like, I read every… I just read books on how to do things, and I learn how to do them, so this is a really good one. I think read this, and then
365 00:39:22.890 ⇒ 00:39:36.629 Uttam Kumaran: like, you may have to read it, I mean, see if you can crush it, but, this Lean Analytics book is also good, but the analytics piece, it’s… I don’t think it’s gonna be that hard once you have an intuition about, like.
366 00:39:37.180 ⇒ 00:39:43.369 Uttam Kumaran: how the product is working, because then you’re gonna be like, should I measure skills? Should I run a survey? Should I look at…
367 00:39:43.810 ⇒ 00:39:47.749 Uttam Kumaran: How long people are in the product, like, we can discuss all that.
368 00:39:47.910 ⇒ 00:39:51.650 Uttam Kumaran: But I think I want you to get an intuition for, like.
369 00:39:53.060 ⇒ 00:40:04.210 Uttam Kumaran: okay, how do I… how do I get someone at Brainforce to log in and be like, I need this product? Like, I cannot do anything anymore without this tool. It’s like, think about ChatGPT, like…
370 00:40:04.480 ⇒ 00:40:05.090 Miranda Wen: Yeah.
371 00:40:05.090 ⇒ 00:40:15.490 Uttam Kumaran: you know, there’s something deeper about building great products beyond, like, oh, someone is spending 2 hours in the tool, like… it’s like, why YouTube over Vimeo? Like…
372 00:40:15.880 ⇒ 00:40:23.400 Uttam Kumaran: You know, it’s like… Yeah, I think there’s just… it’ll expand your mind to thinking about, like,
373 00:40:23.820 ⇒ 00:40:25.569 Uttam Kumaran: Product building a little bit better.
374 00:40:26.190 ⇒ 00:40:29.650 Miranda Wen: Okay, okay, okay, wait, this makes me so hype about this, too.
375 00:40:31.180 ⇒ 00:40:32.200 Miranda Wen: for recommending.
376 00:40:32.200 ⇒ 00:40:39.639 Uttam Kumaran: Yeah, I have a lot of other wrecks, so I feel like give this one a go, and then I have some more about,
377 00:40:40.190 ⇒ 00:40:44.169 Uttam Kumaran: digital products and some AI stuff that we can read next, so…
378 00:40:44.360 ⇒ 00:40:46.870 Miranda Wen: Okay, okay, awesome, yeah, I hear you.
379 00:40:46.870 ⇒ 00:40:55.249 Uttam Kumaran: I didn’t know you were a big reader, so in that case, you’ll love… you’ll love this one. It’s not, like, that boring, it’s… he’s, like, he, like, he’s a really good writer, he’s kind of funny, and he’s, like.
380 00:40:55.500 ⇒ 00:40:56.149 Uttam Kumaran: talks about.
381 00:40:56.150 ⇒ 00:40:58.270 Miranda Wen: Oh, really? Not those, like, rigid…
382 00:40:58.270 ⇒ 00:41:05.050 Uttam Kumaran: No, this isn’t, like, that bi… this isn’t… it’s like… it’s a biz… it’s a business book, but it’s kind of about the art of, like.
383 00:41:05.870 ⇒ 00:41:11.639 Uttam Kumaran: like, product management, you know, a lot of people are like, oh, product management’s like, I’m making tickets, and this is that, like…
384 00:41:11.770 ⇒ 00:41:14.200 Uttam Kumaran: I don’t care at all. That’s like… that’s like…
385 00:41:14.610 ⇒ 00:41:14.960 Miranda Wen: That’s a lot.
386 00:41:14.960 ⇒ 00:41:16.020 Uttam Kumaran: It doesn’t matter.
387 00:41:16.020 ⇒ 00:41:16.620 Miranda Wen: Word, yeah.
388 00:41:16.620 ⇒ 00:41:19.140 Uttam Kumaran: It’s like, when I ask you, like, hey.
389 00:41:19.280 ⇒ 00:41:23.719 Uttam Kumaran: Why is this, like, why do we feel like we have the…
390 00:41:23.980 ⇒ 00:41:29.740 Uttam Kumaran: We have the, like, confidence to tell people to use… Forage Labs.
391 00:41:29.800 ⇒ 00:41:30.670 Miranda Wen: Like…
392 00:41:31.240 ⇒ 00:41:40.880 Uttam Kumaran: You know, like, how do we have the, like… who are we to tell people that this is a better product, right? Like, how are you gonna sit… like, let’s say you were to talk to someone at Brainforge you haven’t met.
393 00:41:40.990 ⇒ 00:41:44.620 Uttam Kumaran: And they’re like, I don’t like this tool, like, I don’t believe in it.
394 00:41:45.030 ⇒ 00:42:03.959 Uttam Kumaran: Right? How do you develop the intuition where, like, no, just give it a shot? Like, try this thing, or tell me what you don’t like about it. So, he’ll walk through, like, user research, user interviews, like, kind of getting an understanding for your customer, and then also, like, things like product requirements docs, working with tech.
395 00:42:04.050 ⇒ 00:42:06.720 Uttam Kumaran: But I think it’s a… it’s a good book.
396 00:42:07.440 ⇒ 00:42:10.380 Miranda Wen: Awesome, awesome. Thank you so much. Yeah.
397 00:42:10.520 ⇒ 00:42:14.149 Miranda Wen: Yeah, I’ll keep you updated on my book delivery and my book reading.
398 00:42:14.150 ⇒ 00:42:17.909 Uttam Kumaran: Okay, okay. Well, tell me, I’ll read it too, I’ll read it with you.
399 00:42:17.910 ⇒ 00:42:19.350 Miranda Wen: Okay, okay, yeah.
400 00:42:19.350 ⇒ 00:42:21.350 Uttam Kumaran: I haven’t read in a long time, so…
401 00:42:22.770 ⇒ 00:42:27.650 Miranda Wen: Yeah, I think this heard some, like, open questions that I need to…
402 00:42:27.780 ⇒ 00:42:29.629 Miranda Wen: discuss that right now. I feel like…
403 00:42:29.750 ⇒ 00:42:44.069 Miranda Wen: Do you think, like, we should go over, like, success metrics? Do you think, like, for success metric setting, now I’m setting about some things, but I just thought if I can get the data on that. Do you think, like, it’s more so I should know what are some data that I can collect, and then…
404 00:42:44.470 ⇒ 00:42:51.349 Miranda Wen: make my success criteria really realistic, or ideal, let’s say this is the success criteria, and I’ll figure out how to do that.
405 00:42:52.100 ⇒ 00:42:53.799 Uttam Kumaran: Yeah, I,
406 00:42:57.360 ⇒ 00:43:00.479 Uttam Kumaran: It’s tough, like, success criteria…
407 00:43:00.680 ⇒ 00:43:04.890 Uttam Kumaran: it’s like, if you put yourself in my shoes, what I want to see is that, like.
408 00:43:05.970 ⇒ 00:43:10.230 Uttam Kumaran: If you think about, like, ultimately in our business, you want to see impact to…
409 00:43:10.620 ⇒ 00:43:19.199 Uttam Kumaran: our profit and loss report, right? So, is open work helping us run clients a better margin? Are we able to increase the revenue per employee?
410 00:43:19.710 ⇒ 00:43:21.889 Uttam Kumaran: The problem with that is, like, that may be, like.
411 00:43:22.380 ⇒ 00:43:25.180 Uttam Kumaran: that may be, like, a downstream effect, right? So…
412 00:43:25.180 ⇒ 00:43:25.590 Miranda Wen: Oh, God.
413 00:43:25.590 ⇒ 00:43:26.780 Uttam Kumaran: What are you measuring?
414 00:43:26.960 ⇒ 00:43:28.890 Uttam Kumaran: Here, let me give you a bad example.
415 00:43:29.410 ⇒ 00:43:47.070 Uttam Kumaran: I just want to measure the number of tokens used, right? That doesn’t mean anything, right? And you’re seeing that a lot of companies are doing that right now. They’re morons. Like, they don’t know what… what are they talking about? Like, people are just spamming to try to get their tokens up. So that’s not good, right? Also, you can say skill use.
416 00:43:47.700 ⇒ 00:43:50.069 Uttam Kumaran: Okay, also, maybe not that good.
417 00:43:50.300 ⇒ 00:43:54.429 Uttam Kumaran: But easy to measure, right? Token use is super easy to measure.
418 00:43:54.430 ⇒ 00:43:55.140 Miranda Wen: Yeah.
419 00:43:55.390 ⇒ 00:43:56.660 Miranda Wen: Charge.
420 00:43:56.660 ⇒ 00:43:59.450 Uttam Kumaran: Easy to… yeah, but then also, it’s like, what is…
421 00:44:00.620 ⇒ 00:44:07.509 Uttam Kumaran: So then there’s gonna be a difference between what the right metric is, and what is easy or hard to measure.
422 00:44:07.510 ⇒ 00:44:08.429 Miranda Wen: Yeah, yeah, yeah.
423 00:44:08.430 ⇒ 00:44:12.520 Uttam Kumaran: And so you want to start with the… with this side. Think about, like.
424 00:44:12.650 ⇒ 00:44:19.760 Uttam Kumaran: And I think Greg is a good person. You should actually talk to Greg a little bit, once you have… I don’t know if you already met… you’re meeting with him tomorrow, right?
425 00:44:20.600 ⇒ 00:44:23.889 Miranda Wen: I don’t think so. Am I? I don’t think so.
426 00:44:24.350 ⇒ 00:44:25.750 Uttam Kumaran: Have you met with Greg yet?
427 00:44:26.030 ⇒ 00:44:28.129 Miranda Wen: I only saw her in the meeting last…
428 00:44:28.360 ⇒ 00:44:29.040 Uttam Kumaran: him.
429 00:44:29.450 ⇒ 00:44:30.519 Miranda Wen: Yeah, him. I only…
430 00:44:30.520 ⇒ 00:44:36.200 Uttam Kumaran: I think… He used to work in product analytics at Stack Overflow.
431 00:44:36.500 ⇒ 00:44:37.020 Miranda Wen: Okay.
432 00:44:37.020 ⇒ 00:44:39.389 Uttam Kumaran: Has a lot of good ideas about this.
433 00:44:39.760 ⇒ 00:44:44.750 Uttam Kumaran: He may be someone that’s good to brainstorm with on, like, what is,
434 00:44:44.960 ⇒ 00:44:48.660 Uttam Kumaran: what is a good criteria?
435 00:44:49.750 ⇒ 00:44:53.590 Uttam Kumaran: So that… I think that would be… like…
436 00:44:55.190 ⇒ 00:45:00.629 Uttam Kumaran: I think that would be effective. But I also… I also kind of want you to read this book.
437 00:45:01.130 ⇒ 00:45:05.180 Uttam Kumaran: Or at least skip to the place where he talks about this, because it’ll… it’s like…
438 00:45:05.770 ⇒ 00:45:10.759 Uttam Kumaran: You’ll realize that, like, you don’t want to set yourself up for failure, but you also don’t want to, like.
439 00:45:11.880 ⇒ 00:45:17.110 Uttam Kumaran: look at, oh, are people spending 4 hours in the product? Like, you don’t want to pick a bad metric.
440 00:45:17.390 ⇒ 00:45:22.070 Uttam Kumaran: You want to pick a metric that actually is closer to, like, are people getting value?
441 00:45:22.480 ⇒ 00:45:28.610 Uttam Kumaran: But you may find that you need to measure it elsewhere. Like, a good example, maybe if people use the tool.
442 00:45:28.850 ⇒ 00:45:34.230 Uttam Kumaran: they… like, my job gets easier. Like, the number of escalations that come to me go down.
443 00:45:34.410 ⇒ 00:45:36.920 Uttam Kumaran: Maybe you see, like, the number of bugs
444 00:45:37.320 ⇒ 00:45:39.769 Uttam Kumaran: Right? You see, like, a downstream effect.
445 00:45:40.440 ⇒ 00:45:45.920 Uttam Kumaran: So this is where you kind of have to get creative on, like, what are some things that are easy to measure, but are, like.
446 00:45:46.240 ⇒ 00:45:49.539 Uttam Kumaran: Maybe, like, good indications.
447 00:45:49.650 ⇒ 00:45:52.129 Uttam Kumaran: And then what are some things that are maybe harder to measure?
448 00:45:52.450 ⇒ 00:45:54.369 Uttam Kumaran: Which means maybe we have to use survey?
449 00:45:54.520 ⇒ 00:45:56.809 Uttam Kumaran: Maybe you literally have to go talk and be like.
450 00:45:56.990 ⇒ 00:45:59.580 Uttam Kumaran: Did you… did… did you see this? Did you see this?
451 00:45:59.580 ⇒ 00:45:59.960 Miranda Wen: Huh.
452 00:45:59.960 ⇒ 00:46:07.649 Uttam Kumaran: But they’re, like, That’s the value, right? Metric creation is also, like, a little bit of an art.
453 00:46:07.780 ⇒ 00:46:12.819 Uttam Kumaran: I would ultimately say, Pick 1 to 3 max.
454 00:46:13.430 ⇒ 00:46:16.269 Uttam Kumaran: And make them things that, like.
455 00:46:17.050 ⇒ 00:46:23.490 Uttam Kumaran: Know that if those go up and to the right, or go down, you’re succe- you’re successful. Like…
456 00:46:23.640 ⇒ 00:46:26.619 Uttam Kumaran: It’s sort of my broad… my broad feedback.
457 00:46:27.390 ⇒ 00:46:32.580 Miranda Wen: Gotcha, got true. Okay, okay, I think that’s some good direction to go. Yeah, I will talk to Greg.
458 00:46:32.700 ⇒ 00:46:35.620 Miranda Wen: And put more salt into it. I think, like.
459 00:46:35.760 ⇒ 00:46:47.639 Miranda Wen: in that case, I think, like, at least for the two repeatable workflows, I’m thinking, right now, like, their success criteria could be, like, very different. Like, one for them, one for the partnership, I feel like…
460 00:46:47.850 ⇒ 00:46:52.219 Miranda Wen: Because the outcome they deliver will be different, and…
461 00:46:52.220 ⇒ 00:46:52.820 Uttam Kumaran: Yeah.
462 00:46:52.980 ⇒ 00:47:05.329 Miranda Wen: yeah, it probably needs to be, like, more… At this stage, I feel it’s, like, probably more need to be tailored to, like, each… each skill, and when they use that, and then conclude with something, like, broader.
463 00:47:05.590 ⇒ 00:47:12.890 Uttam Kumaran: Yeah, and this is not an easy question to answer, but I’m gonna force you to have an answer, so, but it’s… but it’s…
464 00:47:13.880 ⇒ 00:47:22.179 Uttam Kumaran: When in doubt, pick… pick… pick one to three, pick… choose less. Choose ones that are easy… that are reasonable to measure.
465 00:47:22.370 ⇒ 00:47:23.190 Miranda Wen: Yeah.
466 00:47:23.190 ⇒ 00:47:24.720 Uttam Kumaran: Like, for example, if, like.
467 00:47:25.090 ⇒ 00:47:30.049 Uttam Kumaran: what I don’t want to happen is, like, we release it, and then you’re like, it’s going really well, and there’s, like.
468 00:47:30.310 ⇒ 00:47:39.219 Uttam Kumaran: Okay, but what does that mean? What does going well mean? Like, oh, well, people are using it. Okay, but, like, is the company winning because of that? So that’s what I think…
469 00:47:40.100 ⇒ 00:47:43.429 Uttam Kumaran: This is an art in itself, so worth reviewing.
470 00:47:44.170 ⇒ 00:47:45.980 Miranda Wen: Okay, totally, okay.
471 00:47:48.410 ⇒ 00:48:00.889 Miranda Wen: Yeah, wait, could you, like, just elaborate a little bit more about, like, the project plan you mentioned that… that, you think, like, down the line I should… I should be doing? Like, for example, like, what kind of, like, project plan?
472 00:48:02.060 ⇒ 00:48:04.960 Uttam Kumaran: Yeah, so, I mean, I think, like.
473 00:48:05.080 ⇒ 00:48:08.310 Uttam Kumaran: I think mimicking that notion that I sent you.
474 00:48:09.070 ⇒ 00:48:12.670 Uttam Kumaran: gonna be effective. I think all you’re gonna see there is that, like.
475 00:48:12.970 ⇒ 00:48:24.949 Uttam Kumaran: I think this is a net new product, and so compared to what they put in there, there’s not a lot of specifications. I think yours requires specifications, but what I do want to see is, like.
476 00:48:25.840 ⇒ 00:48:28.579 Uttam Kumaran: Also, like, what is the build plan, right?
477 00:48:28.710 ⇒ 00:48:29.420 Miranda Wen: Hmm.
478 00:48:29.420 ⇒ 00:48:39.629 Uttam Kumaran: So this is something that, like, I think can come in a later phase, but when you talk about milestones and timelines and features, I want you to… the first part is to have the core thesis.
479 00:48:39.900 ⇒ 00:48:42.240 Uttam Kumaran: And, like, the core fundamentals, and then…
480 00:48:42.630 ⇒ 00:48:44.739 Uttam Kumaran: me, you, and Sam can work on
481 00:48:45.580 ⇒ 00:48:47.830 Uttam Kumaran: What are the pieces that we need to build?
482 00:48:48.050 ⇒ 00:48:51.979 Uttam Kumaran: So you can accomplish your goal, right? Not the other way around.
483 00:48:53.340 ⇒ 00:49:05.239 Uttam Kumaran: Right? We think this platform is great, we have a hypothesis, now it’s like, what are our goals? Like, what does success look like? And then what features allow us to do that, right?
484 00:49:05.320 ⇒ 00:49:19.139 Uttam Kumaran: People need to be able to discover skills. People need to be able to use skills. People need to be able to, like, set up alerts or something, right? Those ladder to this KPI. Otherwise, what’s gonna happen in product is you’ll just build random stuff.
485 00:49:19.230 ⇒ 00:49:22.859 Uttam Kumaran: And guess what? Look at all the digital tools you use.
486 00:49:23.000 ⇒ 00:49:35.699 Uttam Kumaran: you know, just like everybody, half the stuff nobody uses. Some product manager at somewhere was like, I just want to build this today. They built it, right? They didn’t measure, and you’ll see that great products, like, think about the iPhone, right?
487 00:49:35.900 ⇒ 00:49:44.499 Uttam Kumaran: you know how difficult it is to make a change, like, inside Apple? Literally, they will not let you do that unless there’s, like.
488 00:49:44.860 ⇒ 00:49:46.390 Uttam Kumaran: It’s perfect.
489 00:49:46.880 ⇒ 00:49:47.280 Miranda Wen: Hmm.
490 00:49:47.280 ⇒ 00:49:49.620 Uttam Kumaran: You can’t even change, like, a setting.
491 00:49:49.860 ⇒ 00:49:58.410 Uttam Kumaran: Right? And think about it, there’ll be 100 product managers working on, like, the Wi-Fi settings. It’s, like, insane mode, right? And so…
492 00:49:58.550 ⇒ 00:50:04.500 Uttam Kumaran: But then also think about, like, Something like, like Zoom, for example.
493 00:50:04.810 ⇒ 00:50:23.609 Uttam Kumaran: there’s all types of shit now. There’s, like… look, there’s so many settings, I don’t even know half these things. So think about both of those kind of companies, right? And where do we want to be? Internally, we can be a little bit like Zoom. We can ship things, we can try it. For clients, we have to be a lot more, like, everything is, like.
494 00:50:23.780 ⇒ 00:50:25.309 Uttam Kumaran: There for a reason.
495 00:50:25.610 ⇒ 00:50:26.090 Miranda Wen: Mmm.
496 00:50:26.090 ⇒ 00:50:28.440 Uttam Kumaran: The problem is, like…
497 00:50:28.550 ⇒ 00:50:33.979 Uttam Kumaran: Apple has been in business for 20 years, right? Took them 20 years to get here. We…
498 00:50:34.190 ⇒ 00:50:42.429 Uttam Kumaran: How in a few months’ time can we build something sophisticated, that we’re proud of, that works, that we can go sell for a lot of money?
499 00:50:42.760 ⇒ 00:50:50.610 Uttam Kumaran: And then internally, how can we use internally as a testbed, maybe for beta stuff, and, like, for things that are slightly broken? So that’s kind of, like.
500 00:50:51.330 ⇒ 00:50:55.500 Uttam Kumaran: how I kind of would like you to think a little bit about .
501 00:50:55.780 ⇒ 00:50:56.330 Miranda Wen: Fair.
502 00:50:56.330 ⇒ 00:51:05.640 Uttam Kumaran: like, the sophistication of the product and what to prioritize. But ultimately, I think right now, there’s… the risk internally is low.
503 00:51:05.900 ⇒ 00:51:15.209 Uttam Kumaran: But I think the risk internally is, like, if you’re, like, if I don’t… if it’s not clear the features to build and why we’re building it, we may build too much stuff.
504 00:51:15.350 ⇒ 00:51:18.449 Uttam Kumaran: Because it’s a platform, right? Yeah. For example.
505 00:51:19.060 ⇒ 00:51:25.379 Uttam Kumaran: I… looking at your list of things, yeah, maybe there’s some governance, like, what can you do at a user level versus your team?
506 00:51:25.630 ⇒ 00:51:28.599 Uttam Kumaran: Is that priority zero? No, I don’t think so.
507 00:51:28.830 ⇒ 00:51:32.780 Uttam Kumaran: Priority zero is, like, can you use the skills? Yeah. You know, today?
508 00:51:33.020 ⇒ 00:51:35.810 Uttam Kumaran: And so that’s also for you to have an understanding of, like.
509 00:51:36.630 ⇒ 00:51:40.940 Uttam Kumaran: Your customer may ask you for something, but then for you to be like, are they right?
510 00:51:41.380 ⇒ 00:51:50.030 Uttam Kumaran: like, is this actually what they want? And you’ll find that Apple doesn’t listen to anybody. It’s like, famously, like, Steve Jobs always said, like, the customer, like, doesn’t know
511 00:51:50.190 ⇒ 00:51:52.440 Uttam Kumaran: really what they want. It’s up to us.
512 00:51:52.660 ⇒ 00:51:56.040 Uttam Kumaran: To determine that. To give them something, right?
513 00:51:56.520 ⇒ 00:52:02.720 Uttam Kumaran: So, again, as you start to get into the product world, I think you’ll… you’ll find from this book, and you’ll start to see, like.
514 00:52:02.900 ⇒ 00:52:18.129 Uttam Kumaran: the world around us is somebody made a decision on a product, right? And you can tell whether they were lazy, or whether they actually thought about it. If you’re using a product every day, and you’ve never thought about how it’s used, and you’ve just been happy, that’s a great design.
515 00:52:18.330 ⇒ 00:52:25.430 Uttam Kumaran: Right? But if there’s something in your house where you open the microwave and it’s annoying, that’s a bad product design.
516 00:52:25.680 ⇒ 00:52:31.169 Uttam Kumaran: You know? And so, that’s some things I think you’re gonna have to develop the kind of, like, intuition, you know?
517 00:52:31.970 ⇒ 00:52:41.140 Miranda Wen: Gotcha, got you. Yeah, makes sense. So I guess, like, for this product planning, it’s, like, more so when we are at the point of, like, we are rolling out features, like, it passed, like, the…
518 00:52:41.260 ⇒ 00:52:49.270 Miranda Wen: the skills and the repeatable workflow, having, like, a few repeatable workflows work for the go-to-market team, like.
519 00:52:49.270 ⇒ 00:52:53.270 Uttam Kumaran: Yeah, but I think, again, I think you’re being on guided.
520 00:52:53.850 ⇒ 00:52:58.280 Uttam Kumaran: And then discoverable is also, like, really, really important. That’s gonna be, like, wow.
521 00:52:58.570 ⇒ 00:53:02.410 Uttam Kumaran: Right? Like, think about it, our users are, like, previously in cursor.
522 00:53:02.560 ⇒ 00:53:07.739 Uttam Kumaran: They’re gonna switch to this. They’re gonna be like, this is… how do you turn them to, like, oh my god, this is, like…
523 00:53:09.050 ⇒ 00:53:10.519 Uttam Kumaran: Crazy, right?
524 00:53:10.520 ⇒ 00:53:11.730 Miranda Wen: Huh.
525 00:53:11.730 ⇒ 00:53:17.219 Uttam Kumaran: So, that’s what I want you to think about, like, I’m talking to a lot of people at the company about what is a floor?
526 00:53:17.360 ⇒ 00:53:20.450 Uttam Kumaran: The floor is, like, cool, people move to this, they’re able to do this stuff.
527 00:53:21.000 ⇒ 00:53:29.950 Uttam Kumaran: Excellent is, like, oh my god, like, this is insane, I’ve never seen anything like this. I, like, am doing the job of 10 people, like…
528 00:53:30.700 ⇒ 00:53:31.970 Uttam Kumaran: That’s so…
529 00:53:32.180 ⇒ 00:53:46.620 Uttam Kumaran: how do we drive towards, like, the delight? And you’ll talk… you’ll see, like, in product, they talk about, like, how do we delight the user? So they’re like, this is amazing, I’ve never seen. And the edge that we have in AI is that me and you can just build it. We don’t need
530 00:53:46.900 ⇒ 00:53:50.850 Uttam Kumaran: Like, a couple years ago, we would need, like, 100 people to build this thing, you know?
531 00:53:50.850 ⇒ 00:53:51.250 Miranda Wen: Damn.
532 00:53:51.250 ⇒ 00:53:53.060 Uttam Kumaran: Me and you are gonna do it alone.
533 00:53:53.370 ⇒ 00:53:55.399 Uttam Kumaran: I think you’re gonna do it, I’m not even gonna work on it.
534 00:53:55.870 ⇒ 00:53:59.839 Uttam Kumaran: So, like, that’s, like, what we’re here to do, you know?
535 00:54:00.470 ⇒ 00:54:13.780 Miranda Wen: Gotcha, gotcha. Okay, awesome. Yeah, I think, like, definitely clear of a lot of air here, and I also have Noel who I’m gonna reach out to and, better this, and also, cool. Started really, like, getting to…
536 00:54:13.950 ⇒ 00:54:19.319 Miranda Wen: having, like, some prototypes, making the skill work. I think, yeah, I have a lot of things on my mind now, yeah.
537 00:54:19.320 ⇒ 00:54:26.099 Uttam Kumaran: Okay, so I think one thing that could be helpful this week is maybe we set an internal deadline to just try to do a little, like, project review.
538 00:54:26.530 ⇒ 00:54:37.619 Uttam Kumaran: Like, I think we can just have me, you, and, like, I can invite one or two people, and we’ll just… we’ll have you just pitch the plan to them, and ask questions.
539 00:54:37.810 ⇒ 00:54:41.940 Uttam Kumaran: Like, I think if you want to do that, Thursday?
540 00:54:42.510 ⇒ 00:54:52.579 Uttam Kumaran: that gives you another day or so, to just, like… I know it’s tight, but, like, I want you to also be able to… I want us to start building a couple things, and for you to see that process.
541 00:54:52.920 ⇒ 00:54:56.510 Uttam Kumaran: So… Incorporate as much feedback as you can.
542 00:54:56.670 ⇒ 00:55:04.229 Uttam Kumaran: And then I think we just try to, like, get a small group together. You’ll see the stakes are really low, so I want us to, like, I want us to really…
543 00:55:04.930 ⇒ 00:55:12.450 Uttam Kumaran: I want you to hear directly from the users, and hear them talk about, like, what’s annoying about Cursor, what’s annoying about their job.
544 00:55:12.610 ⇒ 00:55:21.500 Uttam Kumaran: how does this solve that, right? Yeah. And I think that’s… it’s gonna… you’re… you’re gonna… it’s gonna answer all your questions, you know, hearing directly from folks.
545 00:55:21.970 ⇒ 00:55:26.599 Miranda Wen: Gotcha, okay, then probably we will still use, like, the stock as a reference. I will make some edits today.
546 00:55:26.600 ⇒ 00:55:28.079 Uttam Kumaran: Yeah, yeah, yeah, totally.
547 00:55:28.080 ⇒ 00:55:32.469 Miranda Wen: Okay, cool. Sounds good, yeah. I will send you an invite shortly.
548 00:55:32.470 ⇒ 00:55:32.870 Uttam Kumaran: Okay.
549 00:55:32.870 ⇒ 00:55:33.989 Miranda Wen: And you can.
550 00:55:33.990 ⇒ 00:55:43.330 Uttam Kumaran: And then, yeah, I think next couple days, send some stuff in platform, like, whatever you’re reading, or whatever you’re up to, like, I think everybody’s excited to see and be helpful, so…
551 00:55:43.610 ⇒ 00:55:47.280 Uttam Kumaran: You’ll see that, like, we operate on, like.
552 00:55:47.390 ⇒ 00:55:51.370 Uttam Kumaran: Kind of, like, rapid speed here, so the faster we can give feedback.
553 00:55:51.510 ⇒ 00:55:57.370 Uttam Kumaran: The faster you’ll feel part of the team, and you’ll be able to kind of… the answers will come to you as you start to see everything, you know?
554 00:55:57.370 ⇒ 00:55:59.939 Miranda Wen: Mmm, okay, cool, ho, ho, cool.
555 00:56:00.050 ⇒ 00:56:02.420 Miranda Wen: Noted. Okay, thank you so much for your feedback.
556 00:56:02.420 ⇒ 00:56:06.920 Uttam Kumaran: Thank you. Yeah, let me know when you get the book, and like, yeah, enjoy the rest of your day.
557 00:56:07.080 ⇒ 00:56:08.400 Miranda Wen: Yeah, thank you, thank you.
558 00:56:08.400 ⇒ 00:56:09.480 Uttam Kumaran: Okay, alright.
559 00:56:09.810 ⇒ 00:56:10.830 Miranda Wen: Have a great day.
560 00:56:10.830 ⇒ 00:56:11.489 Uttam Kumaran: You too, bye.