Led a cross-functional team of 9 in the design, development, and deployment of a multi-agent generative AI chatbot to handle patient inquiries, prior authorization requests, and medication reminders.

Architected the system using a microservices framework on AWS ECS, with different agents (Specialist Agent, Triage Agent, Data Retrieval Agent) powered by LangChain and AWS Bedrock.

Implemented RAG (Retrieval-Augmented Generation) with a vectorized knowledge base, reducing chatbot hallucinations by over 40% and ensuring responses were grounded in up-to-date clinical guidelines.

Integrated with EHR systems via FHIR APIs to provide personalized context, resulting in a 30% reduction in routine inquiries to human agents and improving patient satisfaction scores.

Ensured strict HIPAA compliance across all platforms by implementing robust security controls, encryption-at-rest and comprehensive audit logging.

Built and maintained scalable data pipelines using Java, C++, Python, and SQL, processing diverse data sources.

Conducted thorough code reviews, providing detailed rationales to improve design and code quality.

Mentored junior engineers on best practices in debugging and architectural patterns, fostering a culture of excellence.

Designed microservices deployed on AWS with Kubernetes.

Streamlined releases through CI/CD pipelines, reducing deployment times by 35%.

Evaluated and ranked code diffs for correctness, efficiency, and maintainability.

Partnered with compliance and product teams to ensure security and regulatory compliance.