Software Engineer II – Full Stack & Generative AI at G.E. Healthcare

Architected a multi-agent agentic AI platform using LangGraph/LangChain on AWS Bedrock (Claude 3.5 Sonnet), orchestrating autonomous diagnostic workflows that reduced manual triage effort by 40%. Built and deployed RAG-based knowledge retrieval using OpenAI embeddings, Pinecone, and FastAPI, enabling real-time medical documentation Q&A with 92% retrieval accuracy at p99 latency under 300ms. Shipped React + Spring Boot microservices for an internal AI-assisted reporting dashboard used by 200+ clinical ops users, enforcing distributed tracing with OpenTelemetry and Datadog across 12 services. Led event-driven pipelines using Apache Kafka and AWS Lambda processing 3M+ healthcare device events/day with 99.95% uptime and sub-50ms P95 throughput. Implemented prompt engineering, guardrails, and LLM evaluation harnesses (Ragas + custom scoring) reducing hallucinations from 8% to 2.1% in production. Modernized CI/CD with GitHub Actions, Docker, and Kubernetes (EKS), reducing deployment cycle ti