Dated notes on building sovereign machine learning: what we believe, what we're learning, and why we build the way we do. Written as we go, kept honest by the date on each one.
Building software got cheap, so value moved. The thirty-second test, the value equation, and the value-versus-effort map from our guest lecture at the InFiNet Code bootcamp, with the full field guide as a free download.
Our founding thesis: the machine learning that matters most should run on infrastructure you own, under a jurisdiction you can audit, and Sweden's west coast is the right place to build it.
Thirty minutes, no commitment. Tell us the problem and the constraints, and we'll tell you honestly whether, and how, machine learning fits.