I built a Java-based interview preparation platform that delivers personalized insights, secure authentication through Gmail/LinkedIn SSO with OIDC, and JWT-based session management. I developed a GraphQL Gateway that integrates multiple backend services and supports flexible queries across topics, difficulty, and ratings. To ensure performance and reliability, I integrated Resilience4j and Redis for intelligent request handling and caching, reducing database load by 30% during peak usage.

I designed an end-to-end ML pipeline, including data preprocessing, model training, validation, deployment, and continuous monitoring. I also led a project where I developed a DDPG-based reinforcement learning algorithm using PyTorch, tailored to the dynamics of bipedal robots for continuous control in high-dimensional action spaces.

Developed Java-based APIs to facilitate the interaction of airline data between front-end and back-end systems.