Peter Ekman

Focused on compact, sovereign AI, zero-gradient model optimization, adaptive memory outside model weights, and fast local reasoning.