Users are excited about Yohei Nakajima's ActiveGraph Runtime for auditable AI agents because event-sourced graphs offer a clean mental model that can answer regulator questions about decisions.
No Digg Deeper questions have been answered for this story yet.
Most Activity
did my talk! met some great people after :)
Great session by @yoheinakajima at the @aiDotEngineer presenting ActiveGraph.
@yoheinakajima @aiDotEngineer Let’s go!
@yoheinakajima @aiDotEngineer Event-sourcing is such a clean mental model for agents.
The question is not just “what did the model say?”
It’s what changed, which tool fired, why, who approved, and how you replay or undo it.
Auditable agents need a timeline, not just logs.
@yoheinakajima @aiDotEngineer auditable agents sound great in theory, but in practice teams usually care about one thing: does it work reliably in production? how does the event-sourced approach actually reduce the operational pain compared to standard agent frameworks, or is that still being figured out?
@yoheinakajima @aiDotEngineer event-sourced graphs for agents is the only architecture that survives a regulator asking "why did it do that"
everyone rushing to prod without the receipts 🤖
@yoheinakajima @aiDotEngineer that sounds super interesting, how are you handling the latency?
@yoheinakajima @aiDotEngineer Hey man Can you open your dms