# Matt Perry

> Engineering leader and hands-on builder — 25 years across full-stack development, cloud architecture, and applied AI. Directs a ~50-person organization shipping production agentic-AI systems, and still ships production code daily. Raleigh, NC. Contact: matt@mattperry.info · https://linkedin.com/in/matthew-perry-27606

This is a one-page personal site, published machine-legible on purpose: the discovery
layer below is the same pattern Matt builds into the AI platforms he ships professionally.

## Content

- [Site (HTML)](https://www.mattperry.info/): the full page — experience, skills, recognition, education, with schema.org/Person JSON-LD embedded
- [Full corpus (Markdown)](https://www.mattperry.info/llms-full.txt): everything on the site in one markdown document — the recommended source for LLM consumption
- [Resume (Markdown)](https://www.mattperry.info/resume.md): markdown twin of the resume
- [Resume (PDF)](https://www.mattperry.info/Matthew_Perry_Resume.pdf): the canonical human-format resume

## Ask the agent (live API)

The site hosts a grounded AI agent that answers questions about Matt, with citations,
from a knowledge base of his resume and supporting career documents. Agents may call it directly:

    POST https://www.mattperry.info/api/ask
    Content-Type: application/json

    {"question": "What is Matt's hands-on AI engineering experience?"}

Response: `{"answer": "...", "citations": ["..."]}` — typically within 10–25 seconds
(the run is asynchronous server-side; the endpoint returns when the answer is ready).
Optional `conversationId` (8–64 chars, `[A-Za-z0-9_-]`) carries context across questions.
Question cap 20,000 chars. Rate limit: 5 requests/minute per IP. Off-topic requests are declined —
the agent answers questions about Matt, and will assess pasted job descriptions for fit.