Noah S.

I specialize in QA and evaluation for AI-powered products: conversational AI, voice agents, and LLM systems. My core focus is making sure AI has the instructions, tools, and knowledge base to stay grounded, accurate, and helpful in real-world deployments across hospitality, insurance, and healthcare. How I Can Help Your Project: ✅ AI QA & Evaluation - Testing LLM output for grounding accuracy and hallucination - Evaluating conversational flow and guardrails - Reviewing production deployment logs for performance - Document findings as structured, reproducible, developer-ready reports ✅ RAG Pipeline Work - Building knowledge corpora from raw source documents (menus, websites, healthcare plan documents) into structured JSON and Markdown - Testing retrieval quality and grounding faithfulness end-to-end - Iterating on corpus format, semantics, and phrasing to improve reliability ✅ End-to-End Deployment Support On recent projects I've owned the full cycle: - Translating client requirements into a campaign spec - Creating initial structure and prompt - Coordinating with developers on tooling and retrieval infrastructure - Running initial testing and iteration - Managing post-launch performance review My background is in traditional QA - Structured methodology, detailed documentation, and cross-platform coverage. This is the foundation that makes my AI work rigorous rather than guesswork. I identify why a model fails, how often it fails, and how we can improve, not just "model provided incorrect response" or "model hallucinated."