One person. Local AI. Unlimited scale.

I'm Matt Gladding, and Glad Labs is my proof that the rules of business just changed. One developer with the right AI systems can do what used to require an entire team — and I'm building that reality in the open.

My path here didn't go in a straight line. I grew up around heavy equipment at the family excavation company in Jamestown, Rhode Island — operating machinery, doing septic installs and new-home foundation work. NEIT gave me a B.A.Sc. in Computer Games & Programming Skills (2015) — the degree that taught me to think in systems, state machines, and update loops before I knew that's also how AI pipelines work.

After NEIT I spent over three years as a Production Implementation Specialist in Providence — ETL data streams, SQL procedures against Oracle warehouses, root cause analysis on the kind of dashboards where moving a column the wrong way breaks somebody's morning. After that came a stretch managing residential construction projects, and since 2021 I've worked as an application engineer in geotechnical instrumentation — measurement systems and sensor arrays for civil and infrastructure projects.

Glad Labs is the nights-and-weekends project. Built outside of working hours, on my own hardware, with the idea that one operator plus the right AI systems can run a real content business without giving up the day job until it earns its keep.

That mix — heavy-equipment hands-on, ETL data plumbing, project management, sensor systems, and a game-programming degree — is exactly what an AI content pipeline turns out to need. Reliable infrastructure, idempotent migrations, real observability, no silent failures.

The system running this site is Poindexter — an open-source autonomous content pipeline I built and run on my own hardware (Ryzen 9 9950X3D + RTX 5090, no cloud LLM dependency). AI agents research, draft, critique, refine, and publish — with quality scoring at every stage. Apache 2.0 licensed, every line of code is on GitHub, every config knob lives in a settings table you can read.

The big idea? A Personal Media Operating System — autonomous content generation, multi-platform distribution, social scheduling, stream clipping, photo-to-multi-format — everything a creator needs to run a professional media presence without the operational overhead. I'm building it for myself first, then opening it up.

This isn't corporate AI theater. No pitch deck, no buzzwords, no pretending GPT wrappers are products. I write about what I actually build, what actually works, and what doesn't. If something breaks, you'll read about that too — usually with a commit hash.