Fraud Detection Analytics Case

End-to-end fraud analysis project simulating a real fintech environment, combining data engineering, SQL analytics, risk scoring and BI dashboards.

Tech Stack

• Python (Pandas, NumPy)

• SQL (PostgreSQL)

• Power BI

• Tableau

• Git & GitHub

• Key Insights

• Identified high-risk patterns in cross-border transactions

• Detected fraud concentration in specific merchant categories

• Highlighted increased fraud rate in newly created accounts

• Built a risk scoring model to prioritize investigation

• Project Structure

• python/ → data generation, fraud rules and scoring

• sql/ → data modeling and advanced queries

• data/ → raw and processed datasets

• dashboards/ → Power BI and Tableau files

• docs/ → business context and findings