PerennialA11y - Digital Accessibility & Responsible AI Engineering

Developer in Digital Accessibility & Responsible AI

I help humans build responsible & inclusive software and AI systems.

I'm a developer, not an auditor. I've shipped production software for over ten years across startups, Fortune 500 companies, and non-profits, and I practice two crafts that share one root: making software work for every human, and making AI trustworthy enough to deserve a place in your business. I write the code, run the tests, and ship the fixes. The deliverable is working software, not a PDF you need to interpret.

I'm currently consulting for C4AT and building Mir'at, a user-owned, local-first personal AI with an epistemic integrity layer. I use everything I sell.

Digital Accessibility Engineering

WCAG audits delivered as pull requests, embedded accessibility leadership, and remediation that holds up over time. I find the barriers automated tools miss, because as an autistic engineer I don't simulate the user experience. I have it.

Explore accessibility engineering services

Responsible AI Engineering

Private AI systems on hardware you own, company knowledge bases that cite their sources, and honest model evaluations before you trust AI with real work. Your data stays yours, your costs stay flat, and I run this exact stack myself every day.

Explore responsible AI engineering services

I'm a developer, not an auditor.

I sit with the code and the models. I run tests, write findings, ship fixes, and drive them to resolution across teams if that's what it takes. My job is only done once your engineering culture carries the work without me.

I have lived neurodivergent insight.

I'm neurodivergent and autistic. I experience the barriers that automated tools, neurotypical testers, and even dedicated accessibility teams miss.

I treat AI as a craft, not a marketing layer.

I evaluate models before I trust them, publish my methods, and tell you when AI is the wrong answer. I built a first-principles evaluation framework across 9 models from 6 frontier families and observed self-identity hallucinations in 4 of the 6 families.

Interested?

If you're serious about building truly accessible, responsible software and AI systems, and a better world, I can help. Book a free 30-minute consultation. I'll listen first. If I'm the right person for the work, we'll talk through how it would actually run. If I'm not, I'll tell you and point you toward someone who is.

Email me