At Stripe, I architected a merchant pricing optimization system and developed XGBoost and LightGBM models for transaction volume forecasting. Feature engineering and hyperparameter optimization reduced WAPE from 22.4% to 17.8%, while an offline pricing simulation improved contribution margins by 11.4% compared with rule-based pricing.
At Databricks, I built distributed demand forecasting and inventory planning systems using Apache Spark and Databricks. My independent Enterprise RAG & LLM Evaluation Framework project combines document retrieval, RAGAS evaluation, and a Dockerized FastAPI service; I also hold AWS Certified Machine Learning Engineer – Associate and Databricks Certified Machine Learning Professional certifications.

