Notes on building AI-first systems.

Constant recursive self evaluation and improvements as you collaborate and use flostate.

How a persistent knowledge graph gives agents context that survives across projects, teams and time.

A practical look at goal-driven agents that reason over your tools and data, then check their own work.

Why every task is routed to the best available model, and what that means for quality, cost and lock-in.

White paper: tracing every answer to its sources and every change to its moment, without slowing teams down.