At Gruve AI, I've built production MCP servers connecting LLM agents to HappyFox, Jira, and Confluence, while developing GPT-4-supervised evaluation and knowledge-distillation systems that reduced firewall validation from over four hours to 2–5 seconds.
I've delivered enterprise AI across security, healthcare, financial services, and supply-chain workflows—from zero-PII-leakage redaction for 50K+ financial documents to multi-agent patient interviews that raised diagnostic accuracy from 67% to 91%. My work spans GraphRAG, agent harnesses, RLHF, DPO, LLM-as-judge evaluation, hybrid retrieval, and multi-cloud deployment across AWS, Azure, and GCP.
I lead cross-functional teams, mentor junior engineers, and build reliable, measurable AI systems that reduce costs, improve safety, and move work from prototype to production.
