• Developed machine learning fraud detection models using Python (Scikit-learn, XGBoost, TensorFlow) analyzing 2.5M+

daily transactions, reducing fraudulent losses by 38% through improved anomaly detection

• Architected AWS credit risk pipeline (S3, Redshift, Glue, Lambda) integrating multiple data sources, improving loan

approval accuracy by 34% and reducing processing time by 92%

• Built automated Tableau/Power BI dashboards for AML/KYC compliance tracking 50+ risk indicators, reducing manual

review time by 42% while ensuring regulatory adherence

• Implemented customer segmentation models on 8M+ records using Python and SQL, driving targeted campaigns that

increased cross-sell conversion rates by 23%

Skills: Python (Programming Language) · SQL · Machine Learning · Tableau

• Analyzed RPA performance metrics across 500+ automated workflows using Python and SQL, identifying bottlenecks that improved process efficiency by 41%

• Developed Power BI dashboards monitoring intelligent automation KPIs (robot utilization, error rates, processing volumes), enabling decisions that increased ROI by 35%

• Conducted OCR accuracy analysis on 2M+ invoices using Python (Pandas, NumPy, Matplotlib), implementing ML validation rules that improved accuracy from 82% to 94%

• Built ETL pipelines with Apache Airflow integrating SAP, Salesforce, and Oracle data into Azure Synapse Analytics, enabling unified reporting for 200+ business users

Skills: Python (Programming Language) · SQL · Tableau · Amazon Web Services (AWS) · Data Visualization

• Analyzed logistics data from 5,000+ vehicles using SQL and Python, identifying route optimization opportunities that reduced miles-per-load by 12%

• Developed predictive maintenance models using Random Forest and Gradient Boosting on vehicle telemetry data, forecasting equipment failures with 84% accuracy

• Built Power BI and Tableau dashboards tracking fleet utilization and delivery efficiency, enabling dispatching decisions that improved on-time delivery from 91% to 96%

• Executed 70+ QA test cases for C++-based TMS validating shipment tracking accuracy, ensuring 99.6% data integrity for 12,000+ daily shipments