About — Kanyo.AI

Teach AI to work like you do.

Most AI content shows you demos. Impressive for ninety seconds, useless on Monday morning. The gap between a demo and a system that does real professional work is engineering — and that engineering is what this site is about.

I build AI systems that carry real weight in a real job: work with deadlines, standards, and someone checking the output. Then I show you exactly how they're built, so you can do the same with your own expertise.

Who I am

I'm Kanyo — a chartered accountant and ex-auditor, now specialising in revenue accounting at a top-tier technology company in the UK. My day job includes building agentic AI systems that automate our day-to-day work: steady output, audit and SOX ready standards, and a faster period close.

My roots are the other way round: technical degrees in software engineering and data science, and years of building with data and automation before I ever touched a ledger. Most people learn a profession and later bolt on the AI technology. I came in carrying the engineering and learned the profession on top of it — which is why my systems get built like systems, not like prompts.

Why accounting shows up so much

Because it's my profession, and because it's an unforgiving test bed. Accounting work has hard deadlines, defined standards, and audit requirement. An AI system that survives that environment has solved problems every professional field shares: unreliable outputs, missing structure, no way to check the work.

But the method isn't about accounting only. Same architecture, your domain — the ontology changes, the method doesn't. If you're a lawyer, an ops manager, an analyst, an engineer of any kind: the builds are worked examples, and the method is the product.