Full-Stack Software Engineer & AI/ML Integrator. MCP servers, agent architecture, LLM evaluation.

San Diego, CA · tounsils.github.io · LinkedIn · tounsils@gmail.com

Twenty years shipping production systems end to end: architecture, backend, cloud, frontend, data pipelines. The last three of those layering LLM applications, agent orchestration and evaluation pipelines on top.

What I care about is the part most AI products skip: making the thing provable. Typed tool contracts instead of prompts, answers grounded in real data instead of generated, and an evaluation corpus wired as a release gate so a failure blocks the ship rather than starting a discussion.

A public reference implementation of that pattern, and the fastest way to see how I build.

github.com/tounsils/ask-me-mcp · MIT · listed in the official MCP Registry as io.github.tounsils/ask-me-mcp

Six typed tools, JSON-schema strict. External grounding, so the model never invents. Confidence and sources on every response. A 15-case evaluation corpus that gates releases. OAuth 2.1 with anonymous dynamic client registration and PKCE. Two transports: stdio for local, Streamable HTTP for the deployed endpoint.

It answers questions about my work, patterns and availability. Install it and interrogate it rather than taking my word for anything:

https://ask-me-mcp-xi.vercel.app/api/mcp

• Senior engineering at NXT Robotics. Operator console and agent layer turning device telemetry into decisions for physical-security and field-equipment fleets.
• VP of Engineering (advisory) at AIMIA. Multi-agent assessment platform: specialist agents plus a supervisor synthesizer, real-time voice, multilingual output, encrypted PII with blind-index lookup.
• Founder of digitalqrcard.com. Node, Express, MongoDB, Vite + React.
• Taking a small number of fractional CTO and fixed-scope engagements alongside those → scope and terms

Languages. TypeScript, Python, JavaScript, PHP, Bash AI / LLM. MCP servers, multi-agent orchestration, RAG, evaluation corpora and certification gates, real-time voice, AWS Bedrock, Vertex AI Frontend. React, Next.js, Vite, TanStack, Tailwind Backend. Node.js, Express, Laravel, FastAPI, REST Cloud / DevOps. AWS (Lambda, ECS, Aurora, Cognito, SAM), GCP Cloud Run, Vercel, Docker, CI/CD Data. MongoDB, PostgreSQL, MySQL Testing. pytest, Jest, deterministic harnesses, eval runners

tounsils@gmail.com · LinkedIn · résumé