I’m part of Safe Intelligence, an Imperial College London spinout building safer AI systems for high-stakes environments. We recently launched /Spec27.ai, a product for validating LLM applications and AI agents with reusable specs, automated test generation, and deeper capability analysis. I’m passionate about making AI safer and more dependable.