Proof of work
What one veteran and an AI workforce actually build and run. Everything on this board is real and in use. We show the concept and the value here — the blueprints stay home.
The builds
A wall-mounted board that pulls one household's world onto a single screen: the weather, the day's plan, training progress, reminders, and the status lights of the systems it runs. It refreshes itself, writes its own morning brief, and does it all on hardware that lives in the house.
It is the working answer to a simple question: what does useful AI look like when it serves a family instead of a feed? Not a gadget demo — a board the household actually reads every morning.
A capable AI that answers at home, on machines the family owns. Questions, drafts, and daily work stay in the house — private by architecture, not by a policy page's promise.
This is the flagship principle behind everything VetTech teaches: you can get real value from AI without handing your life to someone else's server. The rig is the proof that it works, running every day.
A demonstration we run in person. A visitor plants a fake secret — a made-up password, a pretend account number — into a question. Before any AI model sees the text, a deterministic guard catches and strips it. Deterministic means a checkable rule, not another AI making promises. The answer comes back with a provenance stamp: 0 bytes left the box.
Most AI safety talk asks for trust. This one hands you the receipt.
A text war game fought from the radio net: you give the orders, the intel lies, and the enemy has a plan. The opponent plays to win by real doctrine, resolved by a deterministic engine — not a script you can memorize.
Designed, built, and shipped end-to-end in-house, and it's public: you can play the demo right now.
No page builder, no template. Hand-built pages, self-hosted fonts, an animated circuit eagle, a site guide named Idun in the corner, and it keeps working when your connection drops. The page you are reading is the web-design portfolio piece.
A patent-pending architecture that splits AI work between machines you own and the cloud, so the private part never leaves hardware you control. The design stays confidential while the patent process runs. The principle doesn't: use the cloud's strength without handing it the keys.
The method
Everything on this page was built by one person running AI as a workforce — Claude and local models doing the drafting and heavy lifting, the human owning every decision and every check. The loop never changes:
That loop is also the product: it is exactly what the workshops teach schools, small businesses, and veterans to run for themselves.
AI is a strong drafter and a confident liar. So nothing here — no claim, no number, no line of code — ships on the AI's word alone. It gets checked against a primary source or a live test first.
It's the standard behind the motto: Grounded in truth. The thinking is public — read the blog for where this method comes from.
Call or text (513) 278-7137, or email ron@vettechhomefront.com
Get in touch