• Minnesota Timberwolves Data Pipeline

  Built a containerized Airflow ELT pipeline in Python/Docker to pull NBA API data and load Timberwolves games and player stats into Snowflake.

  Implemented config, staging, and quality controls (CSV staging, date partitioning, retries, logging) to support daily automated runs.

  Connected Snowflake to PowerBI for dashboarding and analysis.

• Ryder Rank

  Created a custom match play index by aggregating and normalizing metrics to quantify player suitability for match play.

  Optimized rosters with ridge regression to predict player readiness and linear programming to select the team with the highest predicted performance.

  Validated model effectiveness through sensitivity analysis and historical backtesting against 2023 Ryder Cup results, demonstrating a higher theoretical win probability for the algorithmically selected team.