Why it matters
Agentic memory architecture is a live design problem for anyone building AI systems that need to reason over growing, changing knowledge instead of re-reading everything each time.
The tokenmaxxing angle
A local-first semantic graph is essentially a context-efficiency system: instead of stuffing full documents into every prompt, agents query structured memory, cutting redundant tokens on repeat lookups.
From the organizers
Speaker Tim Rayburn live-demos ingesting papers into a system using Open Knowledge Format and Obsidian, then querying the resulting semantic graph; sponsored by Coder.