Adrian Wedd
I build things that work for people who need them — and I break things on purpose to make sure they actually do.
Greenpeace Actions to government cybersecurity to AI safety. Same methodology every time: enumerate failure modes before you move.
Current Work
All projects →SPARK
Non-coercive AI robot companion for neurodivergent children.
ActiveFailure First
Adversarial evaluation framework. 120+ models, 18,000+ prompts.
ActiveThis Wasn't in the Brochure
Neurodivergent co-parenting guide — the brain you actually have.
ActiveAfterwords
Local voice output for Claude Code — 17 cloned voices, zero cloud.
Have a problem worth solving?
Tell me what's broken or what you're trying to build. I'll tell you whether I'm useful.