At Netsmartz, I develop and ship AI features for client projects, from LLM integrations and prompt logic through production. I build LangGraph multi-agent systems with MCP tool calling and Redis-backed conversation memory, and implement RAG pipelines using Pinecone and Chroma.
On Steadywell OS, I developed an AI voice agent platform for appointment booking and scheduling, handling 2,000+ calls per month and reducing manual booking effort by 70%. I also built production machine-learning models for Dyre ID that predict likely health conditions for dogs and cats, achieving 92% top-10 accuracy.
