FPV Ventures' Nikunj Kothari on startup lessons, AI app wave, and why data ownership determines winners

in sales Amazon's marketleading cloud business generated last year. Let's hear it for Amazon generating 107 billion dollars in sales for AWS. That's his gong. Get the gong. Well, we have our next guest in the studio. We have our next guest with a major trade deal. Major trade deal. Let's hear it. TVPN to discuss.

It's great to finally have you finally have printed your posts. We've reacted to your posts and now you're finally here. Uh, would you mind kicking us off with a little introduction on yourself, though? Um, thank you so much for having me.

It's been my claim to fame is being your first under 100 followers and following your journey. Uh yeah and so it's it's been an honor honor to come on here.

Um and uh yeah I I'm Nikon I spent 10 years building startups everything from logistics networking hardware real estate and um I just joined a new fund um joined this fund called FPV Ventures um with two GPS West with West Chan and Pega who are the two GPS here.

Um and yeah very excited to be here uh after you know dipping into my investing for at Coastal Ventures for a year. So very cool. What does FPV stand for and what is your focus at the fund? Yeah, so FPV stands for founders point of view. I think that's what kind of like really drew me into joining this fund.

Um we mostly just cared about like what the founders are doing. Um we're a generalist fund. We just raised our fund to um $525 million. uh just closed a few months ago and then uh we have a Thank you. we have a billion management. It's great. Yeah.

Uh and it's a generalist fund which means like you know it's and they they put the money we put our money where where our mouths are. We've done everything from biotech to all the way to consumer fintech uh to rag infrastructure to anything you can kind of think of.

And I think as I was kind of figuring out what I want to do next, the world is very exciting and kind of like both atoms and bits, AI is affecting kind of every small detail and you know excited to be here and be a small part of the team. Yeah, I love a typical generalist fund says yeah we're generalists.

We invest in enterprise software and AI usually at the intersection. Uh I mean speaking of that, do you think the uh the the ship has sailed for uh the investments in the foundation modeled companies? Is the game over and now we're in the vertical AI, application layer AI, consumer AI.

Is that where the most interesting companies are getting started and scaled today? I think as the large models are kind of stabilizing like a lot of the value is in the application layer because now you can kind of build on top of that.

a lot of the infrastructure was has been built and so now every company every enterprise is kind of being like okay we can automate large percentages of work that can't be done do I think like a large model is kind of the ship has sailed unless you have like new completely new architectures I do think so like if everybody if you're just building on the transformer side it's hard for me to be excited given the data and from a money perspective too you're just kind of out bid but I'm seeing like a lot of small models kind of come up and more application focused models come up.

I used to work a company called meter. I know you had Anil on the on the on the podcast a couple of days ago. They just raised their lounge around part of one is to build the networking model which you know they control the infrastructure for and they're partnering very closely with Microsoft.

That kind of data is not available anywhere else right you can't get packet data and so I get more and more excited about that. Is there like unique data that you're capturing and then can you translate that to efficiency? Um I think that's where bets can still be placed and you can see outsized value.

Uh otherwise like I think on the large model side um it seems a little bit hard unless there's a brand new architecture and I'm really excited. I've been reading technical papers trying to figure out is there a new architecture? Is there a new transformer like model that will come up? So totally.

How do you think about the convergence of every company into being an app builder? You highlighted I think in a post sometime in the last 24 hours. Now you have Replet, Bolt, uh, Figma, u, Lovable, there's a bunch of different players.

Uh, basically if you, if a company, if a big company is doing a launch and around AI, you can guess that maybe there's a 50% chance that it'll be some type of, you know, promptbased uh, software builder. Is that uh, do you think a lot of that, you know, it makes sense in a lot of different contexts?

Um but at the same time it feels like there's some FOMO element companies seeing you know this tremendous growth from some new players and saying oh maybe we should have uh a horse in that race but but how are you looking at that market? Yeah I mean air tableable announced theirs as well.

If you look at retool they announced theirs as well a week ago. Um our portfolio company canva has one.

I think there is there are some companies who are doing it for defense because they're seeing the replets and the boards kind of coming at it and taking a lot of their business away especially if you have like you know you can build internal tools then like why would you use and so I think some some of it is coming from defense but some of them are coming from offense because they see that they can go deeper into the workflows.

Uh from my perspective, a lot of it is for context also, right? Like if you're able to pull in context from all these different sources, it increases your retention. You can have people do more things on there. It makes it easier for you to highlight them in one place.

And I think we we I think we all agree we're in the era of bundling. More and more companies like more and more enterprise want like one solution. They don't want to pay for like five different solutions. And so you're able to see all of that kind of combined there.

Uh it is it is wildly interesting to see on the sidelines.

I have no predictions on kind of how it goes that the the rise of like lovable replet and bold all doing such insane revenue numbers is probably like what's causing some people to play defense but it just means software is just going to keep growing and growing and spend on software is just going to go parabolic in the next 5 years.

Totally.

How do you how do you think about uh how are you looking at uh application layer opportunities in the context of companies like open AAI just want you know having this tremendous appetite for data right you know we talked about on the show uh clearly has this sort of interesting experience of this sort of p being able to get the power of LMS passively right just sort of in an interview or you know putting putting uh our intern on clue so he can answer random questions that we have um that feels like something that OpenAI will will uh eventually you know if if it works they'll want to roll it out.

How do you think of that the kind of competitive dynamic between kind of upstart application layer companies and and some of these bigger uh labs? I think it's the question like kind of like every decade you can pick up dominant incumbent and ask what if X does this.

So like you know everybody open the the difference this time though is the pace is just staggering. Like in the last few months Open AI has launched like a meeting with recording notes and pulling in context and so I think that's what's scaring people.

Uh but I do think like building a horizontal and vertical at the same time with deep workflows it's just hard.

It's like how many how many things can you do and then the interface becomes like a jumbled mess and then how do you kind of do that and so there is a lot of lot of depth in each application each job each vertical there's so much to do I don't think like open AAI or a single company can capture all the value openai is uniquely positioned where they have a dominant consumer company in like chat GPT and also marketleading APIs but there's claude that has incredible APIs for coregen as All um but yeah I think there's like plenty of depth.

I do think like companies that are on the edge of productivity or just building small wedges or like features that they don't have the founders who don't have the depth to like think through like how does this go in the next few years. I think those as in the past will continue to get like decimated by incumbents.

Uh but I still think there's so much value to be had. Uh I mean it's just it's this this the size of the pie is too large for one company to own it. Um and I'm opening a big span but I still think there's plenty to do.

So how are you thinking about go to market for new AI products either in consumer or uh or B2B uh there it feels like there's multiple strategies right now. You know you see that cursor chart and it's just it's a line directly straight up because they just got to hundreds of million.

Are they at 500 million now or something? 500 million AR. Yeah, we had we had Amjad from Reploid on yesterday going from 10 to 100 in a matter of months, right? So fast. Um and then and then you see the the Kulies of the world and AI at friend. com did another strategy which was uh very viralitydriven.

And so there's there's kind of two two strategies that I'm seeing pop up with the new products. One is kind of like build silently, let the product kind of just grow into this behemoth and then everyone starts talking about it because of the growth.

The other is is take over the timeline, become popular ahead of growth and then use that to convert into subscriptions or revenue growth. Um are are are you looking at both types of deals? Do you see one as kind of more of a positive signal than the other?

How are you thinking about how the next generation is growing their businesses after they build something? It depends on who they're targeting. I think uh it is being a media arm of a company is important now. I think it's hard to like get away from the noise unless you are showcasing your product in some way.

That's why we're seeing our timeline flooded with launch videos. But that's also getting saturated now. So I think more and more what I've been probing for these bottoms up companies that are trying to go after like the individual is understanding like hey where are you spending your time? Are there any channels?

If you're just doing yet another Twitter launch video, then maybe that's like, you know, oversaturated. Like what are other channels? Do you go multiprong? Are you like, you know, are you doing local events or like local communities where you can kind of own it and just build such a good product that people will do?

So virality and PLG has been there for 10 years. It's not new. I think like virality on Twitter is just the newest element of it. And then and then on the enterprise side though and this is one of the reasons I joined FPV like Pega on our team.

She's incredible at like really working with the team and understanding like you know who's the ICP? What's the org chart looks like? How do you break in? Is this like a bottoms up play in the enterprise side or are you top down? Do you need the decision maker? How do you get them in a room? Is this a burning problem?

Is this not a burning problem? Where are they like actually evolving? Is this in budget? Is this not in budget? I think those are the questions uh people need to ask.

Very few companies like cursors or like you know of the world are able to break through like you know cursor like kind of grew from the bottoms up to to the point where enterprises had to come knocking to them and said like hey can you build us an enterprise plan I don't think they even had one.

Uh but that's more of a rarity and sometimes I think uh people take hype on Twitter or just like everything needs to be PLG as the gospel which I I don't buy that that's the only way to do business. So how are you seeing the PM role evolve?

Do you have any predictions or or you know one way to make a prediction would just be to kind of identify like what's happening today. So I'm curious what what you're seeing. Yeah. Um so one I like I never treated my PM role as a PM role. I always took a GM responsibilities wherever I joined.

So even at meter I had P&L responsibilities at open door like you know growth was my last job there and I took try to take as much P&L responsibility and I think if you were just like being a product manager and just want to project manage and kind of just like you know not do sales or not do marketing or not just kind of like you know I'm a product manager and a product owner.

I think that job is dead. I think you are going to be out of a job very quickly. I think it's all become a yes and you kind of have to own multiple things and kind of treat your job as like hey I'm going to sell this product. I'm going to figure out what the marketing copy is. I'm going to work with everyone.

I do think the role of an orchestrator is what all of us will do. Like I think more all of us are going to have we already have like you know I was like playing with claude clude yesterday. I had five instances running and I was watching TVPN.

I was just like building a call tool and I was just yeah nun said like tvpn clips are perfectly timed with like you know running basically. So bullish for us. Yeah. I I I wrote a long piece on this.

I actually think sorry slight tangent, but I actually think entertainment is going to have a massive ups thing because what do you do for 15 minutes? That's why X's numbers are actually through the roof.

I can't like I I I can't they don't have any like stats from them, but it's like well you're writing you're having an agent do the work for you. What do you do? You need a dopamine rush. You go to X or you watch TDPN. And so I actually think that's the future.

But going back to the PM side, like in the orchestration side, PMs are excellent orchestrators. like maybe you kind of have to keep all of these things handing.

So I think the PMs who are like upleveling themselves, learning what sales is, trying to figure out what marketing is, doing some of the wide coding, prototyping themselves, and then being the core orchestrator, I think they'll do phenomenally well because I think like most engineers you talk to, most marketers you talk to, they don't want to deal with like people problems.

They want like an orchestrator. But if you're like a PM who just wants to hold on to like I have a PRD and I'm a project manager, the work behind the work is dead. like that is just going to die a slow death in the next like 3 to 5 years. It'll take a while to trickle down to the larger companies.

And I'm seeing it in startups. No startup is coming to me and saying like, "Hey, I want a PM. " Everyone's like, "Cracked engineer, cracked marketer, cracked salesperson. " That's all they want. Nobody's asking me for like a PM today. Uh at least they're like 50 people. Uh what about uh pricing?

are are non-AI companies getting uh valuation lifts just by nature of raising at the same time as this AI boom or you know are you seeing that when you're when you're kind of looking at opportunities in other categories?

I I was the question like are AI native companies like just seeing a massively high no so it's very clear that like AI native companies are getting this massive valuation just lift do you think that's benefiting sort of non AI businesses by just nature of investors being like well I just did a seed round at 60 posts like grocery stores making 1.

5% profit but with AI could 10x that who knows but yeah yeah I'm just curious like you you let's say you look at some company that's doing, you know, farming focused biotech, are they just naturally getting like a slight lift or or maybe maybe it's too anecdotal? It really depends.

Sorry, I know that's a [ __ ] answer, but like I think it like some of it is kind of like how much data do you have like you know if you've been working on a good example is case text, right? like Case spent 8 years figuring out the perfect case law and GPT 3.

5 came in and they just like skyrocketed and so I think a lot of investors are are trying to look outside of purely AI native mostly to understand are there things are there like built-in brand reputation data reputation things you've done in the past that allow you to then kind of layer AI on top as an accelerant and I think those could be good to go invest in and sometimes help bring in the talent that can help do them um or it's like you know the atoms world physical world still has a lot of alpha is still out there I think like you know I think software as that gets democratized there's a lot of things happening that's why in bio and robotics in like manufacturing supply chain like robots in the field you're going to see a lot of people a lot of excitement there because they see that's the next world that's kind of incoming um but the problem is venture is momentum backed and so as a venture investor you want to see like that inflection point and sometimes if you're not AI native And one of the hard discussions is like, well, they're a great company.

We're seeing some inflection, but we're not sure yet. And if you're not going to see the inflection, then it's hard to be a venture back company today. That's showing up in the series A that or in, you know, the I don't like the word, but the zombie corns that we're seeing from the last few years. So, yeah, totally.

Well, thank you so much for coming on. Y and thanks for helping get Alex on who's coming on next.