From decision traces to decision intelligence.
“Decision traces” is a term people now use when they talk about agentic AI. I ran into it a few months back writing about context graph architecture: the idea that fragmented decision traces can become governed, queryable institutional memory.
Turns out some people were building exactly that years before the term existed. Fred Laluyaux is one of them.
He co-founded Aera in 2016, before GenAI and LLMs, and has been CEO since. In our conversation, he traces his own path: from DJing in Paris, to building SAP’s finance line of business, to Aera’s decade-long bet:
Decisions, not dashboards, are the unit that matters.
We covered a lot: context graphs, agents, skills, memory, determinism vs. agentic reasoning, all through the lens of decision intelligence.
It all comes back to one call Laluyaux made early on: talk about what the technology can do, not how it does it.
His analogy for what that looks like in practice: he trusts a Waymo with his kids more than an Uber driver. Not because the ride is different, but because he can see exactly how each choice gets made.
