AAKSHI SHARMA

· Chargeback Model: Designed and deployed a real-time chargeback model with 90% precision and 85% recall, resulting in savings of $9.6M.

· Fraud management tool: Enhanced fraud management tool by developing rapid evaluation functionality and rule anomaly alert system, saving $13M annually.

· Driver Arrear Reduction: Controlled and reduced driver arrears significantly using risk rules, adhoc analysis and AB experiments

· IN Portfolio Lead: Managed frauds and losses for PayPal India by creating strategies, risk rules and performing dynamic statistical analysis along with global stakeholders management

· Incentive Gaming: Implemented controls on Incentive Abuse, tagging users based on their behavioral and transaction patterns using velocity counters and machine learning classifiers. Instrumental for launching campaigns and received recognition for the same.

· Product Classification (E-commerce, Cloud): Developed machine learning classification model to classify 3M SKUs into pre-defined product hierarchy with ~95% accuracy.

· Theme Detection and Sentiment Analysis (Retail, NLP): Built NLP tool to detect trending themes in customer survey for an e-commerce and helped the clients to identify underlying sentiments in survey open-ends.

· Customer Segmentation (Retail, Big Data): Built unified customer profiles using demographic segmentation and behavioral clustering and subsequently derived customer segments.

· Dashboarding: Designed and developed multiple dashboards for a leading retailer tracking customer experience metrics.

· Responsible for understanding business challenges and formulate efficient analytic solutions and mitigating the risk.