EiTanol

Many drivers always fill up at the same station out of habit, without knowing whether one a little farther away would be worth it. Price information is scattered, outdated or depends on a single source.

Aggregate multiple price sources and turn them into a simple decision: find stations, check prices by fuel type and compare options, considering price and distance.

• Modular monolith in Spring Boot, organized by domain (station, price, location, user, vehicle…), with no microservices before there is a need for them.
• PostgreSQL with PostGIS for geographic queries such as nearby stations and distance to the user.
• Multiple price sources supported from the data model up: each observation stores its source and confidence level.
• Monorepo with backend, web, docs and infra; local environment with Docker Compose.

• Foundation complete: Spring Boot backend with a health check, Next.js frontend and PostgreSQL + PostGIS via Docker Compose.
• Domain model with Station, FuelType, PriceSource, PriceConfidence and PriceObservation as immutable classes covered by unit tests.
• Price history preserved: an observation is never overwritten.
• Public, staged roadmap, from ANP data integration to a mobile app.

• Modeling price as a historical observation with source and confidence, rather than a field on the station.
• Deferring tools like Redis, queues and observability until there is a real need, recording each decision in ADRs.