PlatAtlas
PlatAtlas gives agentic and physical AI work a verifiable accountability layer. Per-org consoles map what agents and machines do, sign every action, and keep a replayable record. OpenCastor is its open-source robotics edition.
Systems engineer · prototyper & builder
By day I keep enterprise systems running for member utilities. The rest of the time I ship: in one 48-hour run this July, four of my iPhone apps went to App Review, a bird journal that paints itself, a rear-view radar for cyclists, the military heat-flag system as a pocket coach, and an Auracast remote for hearing aids, every one built end to end with Claude Code. I’m a Prototyper who spans into Builder: I explore a lot of ideas, then turn the ones that hold up into real, shipping systems. Accessibility isn’t a feature I add; it’s the reason I build.
Sacramento, California · hard-of-hearing engineer · accessibility-first by default

What I’m building now
PlatAtlas is the platform: per-org consoles that map agentic and physical AI work, sign every action, and keep a replayable record. OpenCastor is its open-source robotics edition, and it grew up: a 1.0 iOS app on the App Store, a flashable Raspberry Pi image that goes from blank SD card to pairing QR, and a bench where you say “drive forward one meter” and the car does it, inside a signed approval, with a receipt you can verify offline. ROBOT.md is still the spine: one declarative file that lets any MCP-aware agent safely understand and operate a physical robot. I prototype fast, then build it into a coherent system, pairing with Claude Code.
The OpenCastor edition, to start today
pip install opencastor
castor up Proof it’s real
One month of prototyping and building, paired with Claude Code. Measured, not estimated:
Hardware in the loop: a SO-ARM101 arm named bob, registered as RRN-000000000001. Every feature is exercised against real servos before it merges.
The system
The open-source robotics edition of PlatAtlas is itself four layers, each its own repo with its own release cadence and one clear concern. A change at the bottom ripples up cleanly, which is what keeps a stack this size tractable to prototype, build, and hold together.
Selected work
Computer vision, accessibility, edge AI, open standards. Every one a system built to keep someone safer or more included. The robotics work didn’t come from nowhere.
PlatAtlas gives agentic and physical AI work a verifiable accountability layer. Per-org consoles map what agents and machines do, sign every action, and keep a replayable record. OpenCastor is its open-source robotics edition.
OpenCastor is the open-source robotics edition of PlatAtlas. A single ROBOT.md manifest, the RCAN wire protocol with Python and TypeScript SDKs, an MCP server so any LLM can drive a robot, and a public registry with the first five EU AI Act-aligned endpoints.
An open-source bench where a capable model designs a quantum circuit or chip and a deterministic, numpy-only judge re-simulates and ACCEPTs or REJECTs it. The wedge for a verifiable-efficiency referee: an honest map of where machine intelligence actually gets more efficient.
An interactive, browser-based lab that walks you down a temperature gradient and shows, with real PDB/AlphaFold structures and rigorous prose, how proteins like HSF1 sense and survive heat. The thing I built after I built the wrong thing.
A response to Anthropic's Frontier Red Team work on multiagent systems: reordering identical inputs can invert a model's decision. Eleven pre-registered studies, two zero-dependency Python tools, and a failure record published at the same standard as the findings.
Public identity and compliance for AI-driven robots: permanent RRNs, Ed25519 ownership proof, and the first five EU AI Act-aligned endpoints, live.
How I learned to ship at scale
Before the robots, I spent fifteen years making enterprise platforms ship safely: migrations, automation, governance, CI/CD. That’s where I learned that the hard part of any system isn’t the code; it’s keeping every moving piece coherent.
Why I build
“I build the infrastructure I wish existed, for robots and for the people who usually get left out of the brief.”
Being hard of hearing shaped how I work: I read systems for a living, I document everything, and I design for the person who isn’t in the room. The accessibility tools and the robot registry come from the same place: a belief that the things that decide for us should be legible, signed, and accountable.