iLearn-Lab

Designs long-term user-centric memory for any agent using hierarchical implicit intent alignment (preference and routine), streaming updates, and dual filters (execution-like vs state-like). Use when building personalized agents, memory layers, user modeling, or adapting PersonalAlign/HIM-style ideas beyond GUI.