Increased the F-1 score of the entire anti fraud system by ~ 41%. Reduced response time of the entire anti-fraud service to 10% of the original time. Developed an end-to-end automated blocking and flagging system of suspicious users, anomaly detection and cluster fraud detection systems by using Boosting, ARIMA models. Developed and enhanced a community detection system which identifies fraud clusters in e-commerce consumers using a Bi-Partite graph. Organized Internal Hackathon across Engineering department.

India Developed a real time fraud flagging system to alert on risky transactions. Building multiple reports for internal dashboards. Working on ad-hoc tasks to solve fraud cases.