I built Tresta to solve a problem I had personally: stuck with a non-transferable SJ train ticket in Sweden
Last month I was in Sweden, holding a non-transferable SJ train ticket I couldn't use. No refund. No name change. Just money down the drain.
I thought: there has to be someone else out there who needs this exact ticket.
That frustration became Tresta — a free, peer-to-peer marketplace to lend, borrow, and swap unused tickets. No platform fees. No payments processed on site. Just a match-making layer between people.
The problem is bigger than I realized:
- SJ (Sweden), IRCTC (India), RedBus — none allow name transfers
- Millions of tickets go unused every year
- Secondary markets exist but are sketchy and unstructured
- There's no trust layer for P2P ticket exchange
How Tresta works:
- Post your unused ticket (route, date, price paid)
- Or post what you're looking for
- We match you with someone on the other side
- You exchange directly — outside the platform
Stack: Next.js + Supabase + Vercel. Shipped solo in ~3 weeks while doing an MBA exchange in Gothenburg, Sweden.
What's live now:
- Sweden: SJ, Västtrafik
- India: IRCTC, RedBus, IntrCity
What I'm still figuring out:
- How to build trust between strangers exchanging tickets
- Whether to add a reputation/review layer
- How to expand to more countries and rail networks
Would love your honest feedback — especially from anyone who's been burned by non-refundable tickets.
Live at tresta.xyz
Trending on Indie Hackers
I know a few individuals burned by non-refundable train or bus tickets, so I could ask them if they'd be willing to answer some questions for you for free.
The "exchange outside the platform" decision is the most interesting part to me. It removes a tonne of compliance and Stripe/SCA pain, but it also means you can't enforce trust through the money rails (chargebacks, escrow, dispute resolution). That's a real tradeoff for a marketplace. How are you planning to handle the inevitable "I sent the ticket and they ghosted me" case without payments going through you? I'm curious whether a lightweight escrow or even just a dual-confirmation step would change the calculus enough to be worth taking on the payments scope.