All case studies
Intelligent Enterprise·2025·Tier-1 European Insurer
An Intelligent Enterprise Operating System for Claims
Re-architected the claims business as a hybrid-intelligence operating system — humans, LLM agents and decision services sharing one nervous system. Cycle time fell 62% and the org learned to lead a mixed human/digital workforce.
−62%
Cycle time
47%
Straight-through
120+
Digital FTE deployed
2.3×
Human capacity
— Challenge
A €multi-billion claims P&L was running on disconnected rules engines, fragmented documents and tribal knowledge. Adjusters absorbed every exception by hand, leakage was rising, and the executive team had no operating model for AI beyond isolated pilots.
— Approach
- Reframed the operation as an Intelligent Enterprise: 14 decision moments, each with a clear contract between humans, agents and deterministic services.
- Stood up a digital workforce of LLM and vision agents for intake, triage, damage estimation and fraud — supervised through a human-in-the-loop cockpit.
- Built the hybrid-intelligence governance layer aligned to the EU AI Act: model registry, drift and bias monitoring, override audit trail, agent-level KPIs.
- Co-designed the leadership operating model — new roles (agent owners, exception leads), incentives, and a 90-day adoption ladder with operations.
— Outcome
- Cycle time down 62% across motor and property within 9 months.
- 47% of claims resolved straight-through by the digital workforce, fully auditable.
- 120+ digital FTEs deployed alongside a 2.3× more capable human team — no layoffs.
- Board-endorsed blueprint extended into liability, health and underwriting.
— Stack & Disciplines
Hybrid intelligenceAgentic LLM orchestrationMulti-modal visionEU AI Act governanceDigital-workforce operating model