As a freelance Data Scientist, I’ve worked on fraud detection, risk scoring, and machine learning across independent client engagements. For Takhassos Karmania Online Exchange, I developed a fraud-detection solution using imbalanced classification and cost-sensitive thresholding; for Maalem Insurance, I built and validated risk-scoring models against baseline heuristics.
I also designed and built an end-to-end FastAPI RAG backend that answers from a curated news corpus and withholds answers when evidence is insufficient. My data work spans feature engineering, model evaluation, API development, and reporting, building on an earlier engineering career that included supervising construction teams and using schedules, cost baselines, and quality metrics to support decisions.
