AI-powered financial decision prioritization and historical validation: turns a household's financial profile into a ranked, explainable list of actions (pay down high-APR debt, build an emergency buffer, and so on), then tests that ranking against historical data.
- Ranking: an XGBoost learning-to-rank model combined with a deterministic multi-factor index
- Validation: backtests against historical economic data and stress scenarios
- Audit trail: every approved, modified or rejected recommendation is logged
- Stack: Python, FastAPI backend, Docker
The full project, including setup instructions, benchmarks and the API, lives in Documents/jen_projects/PrioriFi — start with its README.
Built for the GIBC V2 hackathon, Track 02 (Applied AI: Finance).