This Swiss Startup Just Raised $1.9M to Fix AI's Flight Search Problem

Jan de Vries · 2026-08-11

### The Hidden Cost of Searching for Flights Here's something most travelers never think about: every time you search for a flight, someone pays for it. Not you, of course. The airline or travel agency that wants your business foots the bill for every single query, even when you never book a thing. That math used to work fine. But with AI agents now doing the searching, the numbers have gone completely off the rails. That's where Zerolook, a Zug-based startup, comes in with a fresh $1.9 million pre-seed round to fix what's broken. ### Why AI Broke the Flight Search Economy Think about what happens when you ask an AI assistant to find the cheapest Greek island getaway in June. One simple question explodes into thousands of individual searches: every island, every date, every possible routing. And each of those searches costs the airline money to answer. Simone Lini, CEO and co-founder of Zerolook, puts it bluntly: "Airlines and travel agencies only get paid when someone books, but their technology providers charge them to answer searches and can penalise excessive search volumes." Most searches aren't bookings in the making. They're people deciding. And right now, answering those searches costs more than they're worth. ### The Look-to-Book Ratio Spiral To understand the scale of the problem, look at the look-to-book ratio - how many searches it takes to get one actual booking. It tells a stark story: - **Human travel agents:** 1 search to 50 bookings - **Online travel agencies (OTAs):** 1 to 500 - **Metasearch platforms:** 1 to 10,000 - **AI travel agents today:** 1 to 200,000 At those volumes, computing every fare from scratch on a traditional GDS or airline system stops paying for itself. The infrastructure simply wasn't built for this. ### Meet the Team Behind Zerolook Zerolook was founded in 2025 by two heavy hitters from the travel tech world. Lini previously led partnerships for Google Travel, running Google Flights' OTA and metasearch partnerships across EMEA. Before that, he founded Waynaut, which was acquired by Lastminute.com. George Hadjiyiannis, CTO and co-founder, brings a VP of product background from Kayak plus engineering chops from Google, NVIDIA, and eBay. He's the one architecting the prediction model and partner API infrastructure. ### A Smarter Way to Answer Searches Instead of computing every fare combination from scratch, Zerolook's API predicts flight itineraries and prices using historical fare, schedule, and availability data. Each prediction comes with a confidence score, letting airlines serve traffic that's currently being blocked entirely. The company points to OAG's analysis showing searches per ticket sold jumped from roughly 100-200 in the early 2000s to 10,000-20,000 today. By 2030, that could hit 2 million as AI agents scale. That's not sustainable. Hadjiyiannis frames it as a legacy problem: "There is no technical reason that this has to be a problem holding back the entire industry. It is the result of legacy decisions made decades ago; some dating from around the same time as the introduction of the 8-track tape." ### What This Funding Means Led by London-based Playfair with participation from Vento Ventures, TrueSight Ventures, and Alpha Venture, the round also attracted angels from tech, travel, and financial services. The money will fund the launch of Zerolook's API v1 and build out an ML team in Zurich. Zerolook has already signed letters of intent with a group of online travel partners ahead of first pilots. If the prediction model holds up, it could unlock significant revenue for sellers who are currently paying for searches that never convert. The infrastructure that powers flight shopping is long overdue for modernization. Zerolook isn't asking the industry to change overnight - just to think differently about how searches get answered.