The intelligence lives in the workflow, not the model — and a model can't simply read it

FurtherAI’s Aman Gour says the common assumption — “the model is the intelligence, and the workflow is just scaffolding around it” — is backwards. In insurance, a lot of the intelligence lives inside the workflow itself: which risks get escalated, which loss signals matter, which appetite rule wins when two conflict, when a human must sign off. That logic is spread across SOPs, manager reviews, and underwriting philosophy — “not written down in a form a model can simply read.”

His construction is the agentic workflow: the workflow gives repeatability, auditability, and cost control; the agent handles variability and recovers when the happy path breaks; the human stays in the loop for the accountability calls. Over time “every escalation becomes a signal, every exception is feedback,” and the workflow “stops being a script and starts becoming the carrier’s operating memory.” This is the same enterprise-vs-startup harness split in Enterprise agents need deterministic structure while startups need autonomous loops — same models, different harnesses and the routing discipline of Production agents route routine cases through decision trees, reserving humans for complexity; the accumulated operating memory is what makes The context flywheel is a Day 90 moat — Day 0 comparisons are misleading real and what Persistent agent memory preserves institutional knowledge that walks out the door with employees is preserving.