Kevin Huang

At Databricks, I build full-stack AI applications and GenAI observability platforms that track LLM performance, latency, token usage, and production errors. I developed RAG pipelines with Pinecone and FAISS, plus LLM evaluation and prompt optimization workflows that reduced token costs by 20%.

Previously at Anthropic, I built AI-powered applications integrating Claude APIs with React, TypeScript, Python, and FastAPI. I also developed scalable real-time APIs and data pipelines for LLM training, evaluation, and monitoring.

At ProductiveEdge, I designed Go-based service-oriented backend systems that reduced API latency by approximately 40%. I delivered enterprise RBAC workflows, responsive React interfaces, secure JWT/OAuth authentication, and containerized services on GCP.

I enjoy improving performance across the stack—from caching, SQL optimization, and asynchronous processing to reliable Kubernetes deployments. I also mentor junior engineers, review code, and strengthen testing and documentation practices.