Norwegian online auction platform that sells a used car in a single day
Instead of listings and haggling, the car goes to auction: dealerships bid for it and the seller has the best offer the same day. The car is inspected before the bidding starts, so the dealership knows exactly what it is buying. Six years with the same client, from startup to enterprise, with the auctions kept at high uptime throughout.
Screens
The challenge
Nettbil sells cars in live auctions, where seconds decide the price. All of it ran on the architecture from the startup days, which had to keep earning while the company grew — the platform, the infrastructure and the team needed to scale at the same time, and the auctions could never stop.
Our approach
We started with the product: the flow private sellers use to sell their car, the dealership experience in the sales process, and the internal tools that run it all — then the auction platform itself, with bidding as real-time events. As the marketplace grew, we re-architected the platform into microservices along domain boundaries, moved production onto a managed container platform, and kept the entire cloud estate in code. Every deploy passed through a dedicated QA environment, monitoring told us about problems before customers did, and we mentored new engineers until they owned services of their own.
The results
Sales grew from roughly 3,000 cars a year to more than 18,000, and market share grew with them, until Nettbil was a household name for selling a car in Norway. The auctions never paused — not during the migration, and not when Schibsted, the largest media group in the Nordics, acquired the majority stake. The platform came through the technical due diligence the deal brought with it, and we stayed on as the engineering partner afterwards.
What we did
- Built the digital product private sellers use to sell their car in one day
- Developed the dealership side of the sales process, and the internal tools that run it
- Built the auction platform, with bidding delivered as real-time events
- Split the front-end apps and backend services into microservices along domain boundaries
- Scaled the engineering team through a year of expansion and launched the platform in a second country
- Moved production onto a managed container platform, with the whole cloud estate as code
- Gated every production deploy through a dedicated QA environment with hundreds of automated tests, and monitoring on every service
- Integrated the ERP and a long list of third-party APIs, and machine-learning services as the platform matured
- Mentored new hires until they owned services of their own, and kept backups and the recovery procedure in shape