Sports betting data engineer

Sports Betting Data Engineer / Python + TypeScript Developer Project is ready for handover to the right person as we are coming up to offseason. I’m looking for a technically strong developer to work part-time on an existing sports betting intelligence system covering Aus markets NRL, AFL and Horse Racing and possibly expanding into other markets. The system is already built and operational. It includes Python/TypeScript betting engines, data pipelines, Supabase, historical odds/results data, model evaluation, settlement, Telegram outputs and a React/Vite dashboard. This is not a greenfield build and I’m not looking for someone to redesign everything. I want someone who can understand the current system, improve it carefully and help make the data/model pipeline more reliable, measurable and efficient. The main work will include: * reviewing and improving Python/TypeScript betting and data-processing code * validating data quality and historical data * improving model/backtest methodology and preventing data leakage * reviewing probabilities, edge calculations, calibration and performance metrics * improving dashboard statistics and reporting * maintaining Supabase/database workflows * improving runtime reliability, settlement and monitoring * simplifying duplicated or unnecessary processes * evaluating model performance using ROI, CLV, calibration, drawdown and sample size * documenting changes clearly in GitHub Experience with the following is preferred: * Python * TypeScript / JavaScript * SQL / PostgreSQL / Supabase * data engineering * statistics or machine learning * sports analytics or quantitative systems * APIs and automated data pipelines * Git/GitHub Sports betting experience is highly desirable, particularly knowledge of bookmaker probability, vig/overround, closing line value, expected value, calibration and walk-forward testing. The system has strict data-integrity requirements. No fake odds, proxy statistics or fabricated probabilities should be introduced. Model outputs need to be based on real data and reproducible calculations. The existing system and technical handover documentation are available in GitHub, so the first task will be to review the current architecture, understand the existing workflow and identify the highest-value improvements rather than starting again. Hours: approximately 5–10 hours per week Potential: ongoing work for the right person When applying, please briefly explain: 1. your experience with Python and data pipelines 2. any sports analytics, betting or quantitative modelling experience 3. your experience with SQL/Supabase/Postgres 4. whether you are comfortable working on an existing codebase rather than rebuilding it 5. one example where you improved data, model or system reliability Open to all offers and suggestions