Bloom (YC X25), ETH ESOP Scholar & Robotics. Meet David.

Currently: Co-founder @ Bloom (YC 25) Studies: MSc Robotics (ESOP) @ ETH Zurich ‘23, BSc Aerospace @ TU Delft ‘19 (top 1% graduate w/ honors) Previous start-ups: Fireview, Giftit Experiences: Software Engineer @ Verity, Co-founder & Chief Engineer @ Talaria, Autonomous Software Lead @ MIT Driverless Origin: Spain, NetherlandsLinks: LinkedIn, Twitter

Bloom is hiring!

They’re an incredible team based in Zurich, building the future of the creator economy.

Supported by Founderful Campus! They run the VC Academy program, selecting top students from Swiss universities that invest in the hottest university startups.

Growing up and early education

Where did you grow up?

I grew up in Barcelona, Spain until the start of high school. Then I moved to a tiny village along the coast of Barcelona that had about 2,000 people. That's where I did high school. During that time, I also did an exchange year in Canada in Kamloops and participated in three back-to-back summer programs in STEM, including a summer at MIT as part of the RSI program.

Why did you decide to study Aerospace engineering?

I liked so many different things in STEM – math, physics, chemistry, engineering, computer science – basically everything. So I decided to do aerospace for my undergrad because I could do all of those things in one degree. I did that in Delft in the Netherlands.

Then I fell in love with robotics through a minor, so I came to ETH Zurich to do my master's. At the end of my master's, I fell in love with app development and software engineering, and that's how I ended up doing startups.

Studying at Delft University and side projects

What projects did you work on at Delft?

I followed the standard curriculum there, but the main highlight was that we set up Talaria, a "dream team." We were trying to build an electric vertical takeoff and landing jet. We basically went from nothing to a 40-person team, all students. We actually built a working prototype over a couple of years, and it was really cool.

Why did you decide to do a minor at ETH Zurich?

There were a couple universities you could pick from for your minor. My top two choices were Princeton and ETH Zurich. At that point, I was mostly just looking at rankings and how good these universities were.

I wasn't someone that looked at their minor as a way to take a six-month vacation and go to Asia and chill. I was looking at what minor would allow me to grind the most and learn the most from.

ETH was cool because I was in my undergrad and could basically take entirely master's courses in robotics. I could get a really good taste of what a master's would look like and what robotics would be like.

I loved the challenge. I had never done machine learning, robotics, or much software engineering before. I remember going to the advanced machine learning lecture, and the professor asked, "How many of you are doing your PhD?" 30% of the room raised their hands, and I was like, "Holy s**t, okay."

But I loved that challenge. I had to learn the entire Introduction to Machine Learning course in about two weeks so I could catch up with the content in the advanced lectures. It was super theoretical and mathematical.

I also took a course called Duckie Town, which was very applied robotics. I met really cool people there and got super inspired by the technical excellence and rigor of these lectures and professors. I fell in love with that and knew I wanted to potentially come back for my master's.

What did you do during your gap year after undergrad?

After my bachelor's, I did a gap year. The main reason was that I knew about this student team - I knew Delft was working with MIT Driverless, and there was an opportunity for me to be the autonomous software lead at this team and work with really cool people on both sides, in Delft and at MIT.

I really liked the applied part of robotics where you would work on software, deploy it on hardware, and see it run - but on a larger scale, on an actual full-scale electric driverless car.

That year I was going to be going to MIT to lead the team halfway through, but then COVID hit. So we completely shifted gears and started working on a simulator to do the competitions online instead of in real life.

It was also an experience of learning C++ on the go. I hadn’t written a single line of C++ before I started with Driverless, and everything was in C++.

Studying at ETH Zurich and starting Bloom

What made you decide to do so many side projects (like Before They Change the World and YouTube) at ETH that most students don't normally do?

Honestly, it's hard to explain where this comes from, but I think I take a lot of inspiration from others. When I see someone being excellent at a certain craft—and it can be literally anything like Carlos Alcaraz destroying Djokovic at Roland Garros, or Magnus Carlsen playing blindfold chess against 10 people simultaneously, or Charlie Puth coming up with a melody with his perfect pitch—I can get very easily inspired.

I realize these people are human, and I'm human too. There's no law of physics limiting me from doing something like that or being able to do something like that. So I just love to take on challenges and see how plastic my brain and body are—how quickly can I learn something and change my abilities.

How was your experience at Verity?

I interned at Verity during my master's, which was a cool experience. Although it was during COVID, so I didn't get a lot of face time with the people I was working with, I did work on an important part of the pipeline. I deployed code to actual drones and watched these drones fly around at night in an IKEA warehouse, which gave me a kind of "Night at the Museum" feeling.

What made you transition from robotics to app development?

The transition to app development happened later in my master's, especially when I was doing my thesis. I was starting to work with my current co-founder, and I basically listened to my gut. I saw how excited I was to work on my thesis versus how excited I was to work on this app we were building, and there was a big difference.

I think the main factor was probably iteration speed. I was doing reinforcement learning in robotics, which sounds really cool at a high level. But the day-to-day reality is different—it's a lot of tweaking parameters, waiting two days for a training run, looking at graphs, and hypothesizing what needs to be changed.

With app development, it's more immediate: I want to build this feature, I crank out some code, and I instantly see if it works and how it looks. I can quickly add new features and talk to customers. I love that aspect of it, plus the freedom of controlling your own destiny.

I also found that the output of a startup—a product that people can actually use and get value from—was a lot more fulfilling for me than a research paper. So that's how I can rationalize it, but in the end, just listening to my gut would have been enough. I was way more excited about app development, so I knew I would put in more hours and get much better at that craft because of how motivated I was.

Building Giftit, Fireview & Bloom

How did you know when to move on from an idea and what lessons did you learn about validating startup ideas?

This is a very tricky balance and nobody has really figured it out. I do think you get better at it over time.

There's what YC calls the "good kind of crazy" versus the "bad kind of crazy" when thinking about ideas. Some people will call your idea stupid, and sometimes they can be right, sometimes they can be really wrong. The Airbnb founders are an example of the "good kind of crazy"—people thought they were insane, but they had seen a few data points and experienced something magical themselves, so they had conviction.

My main advice is: don't care what others say except if they're your users. Anyone can say "this is a stupid idea," but ignore them unless they're your target users. Ground yourself in user feedback and don't be afraid to uncover the truth by forcing someone to tell you "No, I don't want to pay for this."

First-time founders often postpone talking to users and asking the hard, awkward questions like "Is this a problem for you right now?" or "Would you pay for this?" because they're afraid of someone saying no and killing their fantasy. At some point, you stop marrying yourself to your ideas and actively try to invalidate them until someone proves you wrong.

More experienced founders actually validate an idea before they even start building, which is the opposite of what first-time founders do.

With our first startup Giftit, I had so much energy and excitement that I could just keep going even when things weren't really working. But my co-founder was losing conviction, so it became clear we needed to think about something else.

We had a few ideas in the back of our minds, one of them was a vague idea of what became Fireview, which would solve a problem we had with Giftit and that we had seen other startups and agencies have.

This time, we didn't just build the product. We started with a landing page and waitlist, saying "we're not going to build this product only to realize nobody wants it." We got some decent traction on the waitlist, built the product in a few months, launched it, and started making revenue pretty quickly.

We even got an acquisition offer from a company that wanted to integrate our product into their tool, though we didn't take it because it wasn't a "hell yes" offer.

Around that time, the idea for Bloom had been floating around in our heads, but it was very abstract—like, what about a meta-app where you could draw something on screen and it would come to life, or where you could talk to your phone and it would build an app?

But we weren't thinking far enough and weren't anticipating how good AI coding models would get. We were limiting our vision to super simple apps and wondering who would want this—hobbyists? Designers for prototyping?

During the holidays in 2024, I went to Barcelona for a couple weeks and spent that time experimenting to see if this was a pipe dream or if there was a feasible way to make it happen. I came back with something promising, and in mid-January convinced my co-founder that this could be extremely big and change the world.

The YC deadline was coming up in three weeks, so we built a prototype, submitted the application, and got an interview. I got a bunch of help from people in the SPH mafia who had gone to YC before, which really helped with our prep.

Building Bloom and lessons from Y Combinator

What is Bloom and what is the long-term vision?

Bloom is the fastest way to build and share native mobile apps. What Bloom allows you to do is literally talk into your phone, describe an app—it could be anything: an AI-powered app, a personal app that persists your data, a social app with authentication or payments—and Bloom builds both the front end and the backend for this app, deploys everything for you.

You don't have to do anything else than just describe what you want, and then you instantly can use that app natively on your phone with haptics, gyroscope, camera API, contact sync, calendar, even HealthKit. It's cross-platform for Android and iOS, and you can share this app with a single link.

We're basically taking the convenient web sharing experience of links and bringing it to native mobile apps, which is notoriously difficult even for developers. Normally you'd have to build in Xcode, have an Apple developer account costing $100/year, submit for review, use TestFlight for sharing, etc.

With Bloom, you build the app in seconds, get a link, drop it in any chat, and people can just tap that link to open the app on their phone. It's the most frictionless way to build and share native apps.

Short term, we're empowering people who already think of software as a creative medium—designers, entrepreneurs, developers. They already come up with app ideas but building them takes a lot of time or money, especially for non-technical entrepreneurs who have to hire someone.

But what's really exciting is expanding the range of people who think about software as a creative outlet. The best analogy is what YouTube and phone cameras did to media, where initially making videos was limited to videographers or journalists or filmmakers, and now literally everyone makes videos.

I think the same thing will happen with software. We're unlocking the creator economy, but for software. Right now the creator economy is mostly photos, videos, and music, but software is a superset of all those. The creator economy can become so much larger when everyone can build software for themselves, their communities, and their fan bases.

Getting into Y Combinator

What was the Y Combinator experience like?

We spent most of the batch building, which isn't the best use of time. Ideally, you start the batch, launch immediately, and then use the batch to grow, sell to your batchmates, get growth advice from partners, use the YC brand as much as possible.

For us, we weren't ready to launch. We launched on Bookface (the internal YC platform) about halfway through the batch, saw issues we needed to fix, then launched properly a few days before fundraising week, which is one week before demo day.

The batches are shorter than they used to be, and demo day comes at you so fast. It's crazy.

SF was great, and YC especially was great because they organized events that brought people together—alumni reunions, talks, alumni demo day, demo day itself. These events brought together really awesome people. But most of the batch, we were literally just in our apartment coding.

The main value was that the three of us were in the same apartment, living and working together, which created a very good atmosphere that I'm going to miss.

What were the biggest lessons and value-adds from Y Combinator?

I haven't actually thought about ranking the main lessons, but there are tons of them. At the start of the batch, there's an event where all the partners in your group tell their founder stories in great detail, and you learn so much from their experiences and mistakes.

All the talks about fundraising, how to approach it, and the famous Dalton talk about how not to die as a company are incredibly valuable. He shares statistics from thousands of YC companies and what happens to them after the batch.

But the main value-add from YC was the brand and how that helps you, especially in fundraising. That's where you have the biggest unfair advantage.

It creates an artificial deadline that usually doesn't exist in fundraising, and it gives you confidence because you have high leverage—you have $500K in the bank so you don't need to raise money. You're in a position of very high leverage where you can say, "I want to raise at a 20 to 30 million valuation or higher," and investors will say, "Yeah, okay, YC standard practice." If you tried to do that outside of YC given the stage most companies are at, people would laugh at you.

One of the things I loved about the experience was seeing people I previously thought of as "gods" in person. I had listened to so many podcasts of Brian Chesky, Sam Altman, Jessica Livingston, and Paul Graham. Then I saw them in real life, and the world felt so small in that moment.

You realize these people are just humans like you—they make stupid mistakes, they have dumb questions. When they tell their stories from the beginning of their journey, they're very human and relatable. So there's pretty much no excuse for you not to go for something yourself.

How did Bloom raise $3.4M and what advice do you have about fundraising decks?

That round was raised in five days, more or less, before demo day. I have to say, it was raised without a single pitch deck. I want to tell this to people who are first-time founders because I remember how much I obsessed about pitch decks when I was getting started with Giftit, spending so much time on them.

For this round, I just jumped on calls the week before demo day, told the story of me and my co-founder working together on different things and landing on Bloom. Then I shared my screen, gave them a demo of the product, and told them the vision and how big it could become.

The people who understood the product, the technical problem it was solving, or the pain point it addressed—and had a bit of imagination—understood its potential easily. They didn't need a pitch deck with a TAM slide or anything.

We had launched a few days before, so we had limited traction. YC definitely helped because investors knew demo day was coming and other investors would be looking at our startup. Investors needed to move fast.

What's funny is that you still get some people, mostly European investors or those who haven't dealt with YC companies before, who book a meeting, have a call, and then tell you, "We have our investment committee next Monday and can follow up after that." They don't realize the round might close before then.

It was a super exhausting week with so many back-to-back meetings. I was basically talking the whole day, telling the story, and doing demos. But the product really sold itself in the fundraise.

What advice do you have for people considering starting a company or taking risks?

Especially at this age, I think the biggest risk is not taking risk. You should be making bets when you're in your 20s.

Honestly, at this time when AI is taking off and industries are being completely shaken by new technology, there's no advantage to being in this field for seven years versus one year. Nobody knows how things are supposed to be done—there are no rules.

Young people have a crazy advantage now more than ever. You're taking a huge risk if you're not taking a risk right now.

Bloom is hiring!

They’re an incredible team based in Zurich, building the future of the creator economy.

Hey Arnie here - hope you enjoyed this blog!

Always happy to hear feedback, who should I bring next & what questions should I ask?

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.ambitiousxdriven.com