Building a Unicorn Startup in the AI Era | Pareen Lathia | ETHMumbai 2026

Hello Mumbai, how are you? Great to be here. Uh great job Chandesh part and everyone in the team uh pulling this off. It's getting bigger and bigger and I think I was just telling someone that ETH Mumbai has a spirit of Mumbai because it always happens in the bare market probably peak of bare market where even Bangalore guys are very afraid to do a conference. uh these guys just have the grit and the resilience to pull it off which is what is needed.

Uh the topic that I've chosen to talk about today is actually building in AI era. It sounds pretty generic but essentially I think a lot has changed. Uh oh. Okay. Yeah.

So, I think the uh everyone keeps talking about AI changing everything. Uh AI has changed everything for sure, but how we don't know. We'll talk about that. Uh one thing it hasn't changed is fundamentals. Uh you still require a founder, you still require the grit, you still require a large market.

uh maybe less money than it was needed earlier. Uh how does it change? Uh today is the time to look at that. I Yeah. Okay.

So this is actually all there is. uh we've seen in the history web 2 before that web 3 uh there will be a small team two people three people they will come up with an idea they will build an MVP prototype uh they will try to talk to users they will get to a stage where they are like yeah this is this is something that really works let's spend a few years of our life building this see if it becomes a company or a big startup uh once they do that then the first first thought that everyone has in mind is let's raise capital right this is usually called angel round preede round whatever you want to call it uh ranges anywhere between $50,000 to $2 million uh then they go out and actually try and achieve what people call PMF by hiring a team which is pretty crazy like this used to take from idea to PMF generally ally at least few months maybe 6 months to couple of years and that's how we've taken it for granted like this is what it means uh even today I think uh we see a lot of decks where people are like I've built this cool product that's all is the product live on mainet we don't know is the product really going to work we don't know uh this completely changes with AI today uh the best of founders in the future. I think with 4.6 six opus we crossed some kind of threshold uh probably I've built three products in last one month just like that and now I think all three of them have more than 100 users I have I haven't tweet tweeted about any of them haven't done any marketing as such just been sharing it with people who I think might find it interesting so to reach 100 users for three products took me one month and zero money uh $200 cloud plan that's All that was needed. What is the stage?

Yeah. What is the stage that was needed which is to ship extremely extremely fast probably over a weekend is where what we are expecting right now. No money is needed to build a product. Zero money on idea stage. Zero money to hire a team.

You can see the biggest thing was oh I need three devs because my product is so complicated. I have to do this. I have to do that. The only money probably that is legit today is for audits. If you're building something that's going to handle a ton of money, that's also needed.

But uh there are private launches these days without audits. People are launching smart contracts on mainnet only for 10 users. By informing them that this is an unodudited product, it's up to you. If it is in innovative enough, if it is interesting enough, try it out with $50. And there are enough people who are willing to try out unodudited products for $50.

So every reason to raise money is basically just going away. Give those $200 to anthropic and we are sorted. Uh what becomes very important now is actually PMF. People used to call it PMF and it had a lot of vague aura around it. Uh let's boil it down and call it retention.

Right? So the 100 users that I have for my products, do they come back on a weekly basis? Do they come back on a daily basis? If they do, then that's insane. That means there's a small group of engaged users who actually care so much for the product that I know there is some value to this.

This takes less than weeks. Frankly speaking, this is where uh I would say Paul Graham's mantra of doing things that don't scale come into picture. You go manually and you sign up 100 people. I would say ambitious founders today in this hackathon uh could whip up a product in one afternoon have an MVP deploy it on Versel on some chains mainet uh and then spend next two days going to each one of these people and sign up 100 users by Sunday you know whether your product is working or not. If it is not working just leave it.

If it is something that has some signs of retention, go ahead and build it. Then you go and raise. Imagine in one weekend you've got 100 users. Maybe you listen to them. You fill them in a telegram group and start talking to them.

In 2 weeks, 3 weeks, 4 weeks you'll know whether these 100 people are coming back or not. If they're coming back and then you go to the investors or angel investors or accelerators or wherever they are way more likely to fund you and your chances of success just go up by a lot. So this entire thing has changed completely. Uh I think this is going to be the focus. The money is only for distribution.

I don't think this is very early for me to say. uh probably most VCs haven't said this but in 6 months I don't think any VC is going to invest on a product and a team uh they are going to give you money only when you know exactly what your customer acquisition cost is and how much are customers paying for it if customers are not paying for your product then it's going to be very very hard to raise money so uh if you see everything remains the same here you still need a veg you still need a loop that customers need to go through or users need to go through and then you need to go conquer that market. So fundamentals absolutely remain the same. The sequence has completely changed and that's actually good because gritty founders earlier could not do everything by themselves. Today if you are a really gritty founder uh you can just do most of the things by yourself and that's a beautiful thing for a lot of people.

That's very scary for those who were shilling white papers and demos that really did not matter and were absolutely scared to go and talk to users. Talking to users has become primary skill at this point frankly. So that being said, what I would say is if there was one line uh that has changed with AI, that's cost of PMF has gone down by a lot. There's frankly it's $200 uh plus maybe two months of your time, 3 months of your time till you get 100 paying users. I have to say paying users because in web 3 founders are not used to talking about revenue but that's going to be the new trend thanks to no one buying each other's tokens anymore.

Uh that's a good thing. That's a very good thing. Uh funding is not a signal anymore. Frankly, your number of paying users have become signal. So, if I see someone funded $5 million uh but has 1,000 users, that's actually a negative signal.

Uh although you managed to fool some VCs, but this is actually not the era to do this. So yeah, that and then I would say if you want to showcase what what replaces that $5 million funded by let's say Sequoia, A16Z, Paradigm, Coinbase, whatever top tier VCs, the biggest signal could be 1,000 users repeating on a daily basis. Are they using your product on a daily basis? If you say this out loud on Twitter saying, "I have a product that 1,000 users use on a daily basis, uh 25 VCs will be in your inbox in in a couple of hours. That's what you should be aiming for, frankly speaking.

And I think my time is done, but that's all that was talk that was the talk. This all of these things remain the same. acquire a very very short narrow market before you go and try to even scale like don't go to 100,000 users or even 10,000 users don't shout out on Twitter frankly this has become very old don't tell chains to retweet your product very old tactic nothing happens with that go and talk to each one of the users uh yeah this is a 90-day plan if you want to click a picture pretty simple exactly what I told you guys right now and I think this should be the last slide. Yeah. Uh overfunded startups are done.

Uh you need to make sure that your startup looks like inevitable. If thousand people are so obsessed with your product, then it's inevitable that you will raise money and you will build a unicorn. So thank you so much