About

I am Dan Hesketh. I have spent twelve years building AI systems, and the last few of them building one specific thing: architecture that makes AI behave like it actually knows you, and keeps it that way.

At reMarkable I built rM Brain, a context system the company adopted across seven-plus departments as its own AI infrastructure. The pattern underneath it never changes. Give the model a real, structured foundation of who you are and how you think, keep it fresh, and stop it quietly drifting back into a generic stranger. Most AI fails for a boring reason: the context decays. I build the thing that prevents that.

Then I built it for one person. A professional speaker with a large audience now publishes in his own voice at a scale he never could before, with repeat paid bookings and, across thousands of subscribers, not one reader who has clocked it as AI. That is the proof that this works on a real reputation, with money on it, not just on a slide.

I also teach it. I have run an AI masterclass for rooms of smart, skeptical professionals, five times in the last few weeks. The thing they keep writing down afterwards is the thing this product is built on: use your voice, build a living context that is genuinely yours, and never let the machine fake what it does not know.

This is that, built by hand, one person at a time. If that person should be you, I would like to hear from you.