Principal AI Engineer | Agent Platforms, RAG, Evals, Voice AI

Principal AI engineer specializing in production agentic systems, with 9+ years of ML and LLM infrastructure experience reaching back to pre-LLM production NLP.

I co-founded Steno.ai and led AI engineering for the platform behind the Tony Robbins AI mobile app: realtime LiveKit voice/text agents, Azure RAG over 10TB+ of source data, agent evaluation infrastructure, multi-session memory, fine-tuning pipelines, and tooling for building, testing, and publishing assistants.

• Agent systems: OpenAI Agents SDK, LiveKit Agents, ElevenLabs voice, tool calling, structured outputs, context engineering, multi-agent search, memory, prompt/runtime orchestration, and realtime voice constraints.
• RAG: Azure AI Search, Azure Blob Storage, Postgres, pgvector, embeddings, hybrid and semantic search, custom chunking and chapterization, facet routing, LlamaParse, and LangChain.
• Evaluations: Simulated user testing, agent version comparison, regression detection, failure triage, parallel reruns, source-grounding checks, eval dashboards, and trace/debug artifacts.
• Model training: Synthetic data generation, SFT/DPO datasets, OpenAI fine-tuning, base-vs-tuned evaluation, and model comparison.
• Infrastructure: Python, FastAPI, Pydantic, async systems, TypeScript, React, Next.js, Docker, Kubernetes, Azure Functions, Azure App Service, Modal, GCP, Azure, CI/CD, monitoring, and debugging.

Most of my recent production work lives in private company repositories.