Every company I've built grew out of the one before it. Not by design — by necessity. When one thing stopped working, I looked at the edges for what came next. That pattern took me from a student accommodation marketplace to a Bitcoin arbitrage bot to a short-term rental portfolio generating over $1M in annual revenue.
ErasmusInn: scaling supply the hard way
After three years in finance, I launched ErasmusInn in 2015 — a proptech student accommodation marketplace backed by 500 Global and angel investors. The idea was simple: over 4 million students move across Europe every year and have nowhere reliable to find housing. We'd be the platform they turned to.
I personally validated and scaled the supply side to 10,000 rooms by onboarding hosts across 12 European cities. I stayed in a different Airbnb each night, meeting landlords, pitching the platform, and signing up inventory city by city — Berlin, Lisbon, Madrid, Istanbul, and everywhere in between. We led a 6-person sales team and were finalists at Startup Turkey 2016.
But underneath the traction was a structural problem. Without channel-manager tools or exclusive listing agreements, hosts distributed their inventory everywhere — on our platform, on Airbnb, on local classifieds, on Facebook groups. We had no supply lock-in. Bookings would fall through because a room had already been let somewhere else. The operation became unsustainable, and we wound it down.
Finding spread where others saw noise
After ErasmusInn, I was looking for what came next. I noticed that Bitcoin traded at meaningfully different prices on USD exchanges versus Turkish-lira platforms like Paribu and BTCTurk. The spread was sometimes as wide as 20% per trade.
I built a simple arbitrage bot — buy BTC on USD exchanges, sell on Turkish-lira platforms, pocket the spread. It worked well for a while. But like all arbitrage, the margin compressed as more participants entered. The window closed.
What that experience reinforced was a principle I've carried into every business since: if you can see a pricing inefficiency before the market corrects it, move fast.
One listing that changed everything
While running the arbitrage bot, I spotted another spread — this time in real estate. I listed my Istanbul flat on Airbnb as an experiment. Short-term letting generated roughly $4,000 per month versus around $1,000 on a long-term lease. A 4x spread, hiding in plain sight.
I took that insight and ran with it. I started leasing more apartments, furnishing them, and listing them on Airbnb — a model known as rental arbitrage. The economics were compelling: lease at long-term rates, operate at short-term yields, and keep the spread.
Scaling to 10% of Istanbul
Over the next few years, I bootstrapped that first listing into a portfolio that captured roughly 10% of Istanbul's professionally managed short-term rental listings and generated over $1M in ARR with a 45% EBITDA margin. We called the company Oval.
The business worked because Istanbul had massive tourist demand, a weak lira that made the city cheap for foreign visitors, and a fragmented supply of professionally managed rentals. We moved quickly, locked in favourable leases, and built out the operational muscle to manage dozens of properties simultaneously.
The seeds of what came next
But scaling Oval also exposed a fundamental truth about hospitality operations: the more properties you add, the more chaos you create. Every new unit meant more guest messages, more cleaner coordination, more vendor management, more exceptions — all stitched together with WhatsApp threads and spreadsheets.
That pain — the feeling that the operational engine was always one step from breaking — is exactly what led me to build Cendra. But that's a story for another post.