Kevin Phan

I’m a senior AI full-stack engineer with 8+ years building AI-powered applications for healthcare and SaaS. I specialize in scalable, HIPAA-compliant systems that combine LLMs, agentic workflows, and RAG—delivered with production reliability in mind.

Most recently at Abridge AI, I developed an end-to-end healthcare platform with Python (FastAPI)/Node.js backend services and a React/TypeScript clinician-facing UI. I built secure REST APIs for PHI-compliant data exchange, designed multi-agent clinical workflows for automation, and created real-time voice-driven clinical agents using LiveKit, Deepgram, and ElevenLabs—along with evaluation frameworks and safeguards that reduced failures and improved reliability.

At Ambience Health, I built full-stack AI features end-to-end, including a voice AI receptionist, LLM-powered document summarization, and a context-aware chatbot grounded in retrieved patient context. I improved output reliability through prompt refinement, validation, and structured response formatting, reducing malformed responses and improving factual accuracy.

Previously at Zing Health and Google, I shipped enterprise and internal tools across the stack—building member-facing portals used by 20,000+ daily users, care coordination services for clinicians, and population health analytics pipelines. I bring a product-focused engineering mindset rooted in testing, debugging, observability, and cross-team collaboration.