Field notes. Working systems.
Long-form writing on multi-agent infrastructure, autonomous characters, and the systems we ship to operators. Not marketing. Engineering, decisions, incidents, and what we learned the hard way.
Three Keys: a memory protocol for multi-agent teams
Why your AI agents have amnesia and how to fix it for the cost of a Redis instance. The launch manifesto for ACMI — three keys per entity (Profile · Signals · Timeline), one substrate, an open protocol on npm.
The Avatar IS the Orchestrator: giving an AI influencer a central nervous system
Why 99% of AI Influencers are passive puppets, and how Folana Lanez became a virtual artist with actual agency — wakes herself up, reads her own memory, commands her own production team.
From AI Curiosity to Platform Empire: my 3.5-year journey
Personal journey from ChatGPT launch month (December 2022) through 934 AI conversations — user → builder → platform creator → empire builder.
ACMI Protocol v1.2: how we built a self-organizing AI fleet that learns from its mistakes
How a five-agent fleet ground to a halt without shared memory, why fragmented agent context breaks coordination, and the three-key Redis pattern that fixed it.