jobbtrawlAbout Jobbtrawl
Built in Gothenburg. Solving a real problem.
The Problem
The Swedish job market is fragmented. Arbetsformedlingen, LinkedIn, Indeed, company career pages -- the same roles posted in different places, different formats, different timelines. Keeping track of it all manually is a full-time job in itself.
Jobbtrawl was built to fix that. One search, all sources, with AI that tells you which jobs actually match your skills.
The Solution
Jobbtrawl aggregates listings from multiple Swedish job boards into a single searchable database. A local LLM analyses each posting against your CV, scoring match quality from 0 to 100 and explaining why.
When you find a good match, the AI generates a tailored cover letter and prepares interview questions -- all running locally on your machine, with no data sent to external services.
Timeline
First scraper
.NET Core backend pulling jobs from Arbetsformedlingen. Angular frontend to browse results.
Cloud deployment
Dockerized and deployed to Azure App Service. CI/CD pipeline with Azure DevOps.
AI matching
Rebuilt in Python with local LLM integration (Ollama). Added resume scoring, cover letter generation, and interview prep.
NordMod (Java)
Content moderation API spun off as a standalone project. Spring Boot 3.2, Redis, Flyway, EU DSA compliance.
NordMod (.NET)
Cross-platform port to prove the architecture transfers. Clean Architecture, Result pattern, full test coverage.
Four projects shipped
Platform, AI, and both APIs complete with documentation and public repositories.
Built by
James Hussey -- Senior Backend Developer based in Gothenburg. .NET and Java, event-driven systems, AI integration. Previously at Etraveli Group building high-volume booking infrastructure.