AI-powered process mining resets insurance operations at Allianz
Process mining used to be a niche topic. Today it's a practical way to turn messy operations into clear, measurable flows that teams can fix fast. For Allianz SE, the shift is tangible: less friction for customers, fewer internal handoffs, and faster outcomes across core processes.
Speaking at Celosphere 25, Jan Malmendier, chief operating officer of Allianz, said the company is moving past "customer satisfaction" and aiming for genuine enthusiasm. "We want customers to love our products and processes. That's where process mining helps us a lot."
From years to days: a claims overhaul
Allianz applied AI-driven process mining to its accident claims. The focus: cut touchpoints, trim cycle time, and reduce cost to serve. "In the past, it took years until we were really ready to pay," Malmendier said. "Now, we can pay in most cases after a few days - and that really makes a difference."
That improvement didn't come from a single bot or a new portal. It came from seeing the end-to-end flow, spotting blockers, and removing them with data-backed changes.
Home purchases, simplified
Transferring a property to a new owner used to involve a "complex triangle" of documents and steps. Allianz stripped away the paperwork and flattened the process. The result: faster transactions and a cleaner customer experience.
How they did it: process mining + systems thinking
Process mining turns raw event data into a living map of what actually happens - not what the SOP says should happen. If you're new to the concept, this overview is a useful primer: Process mining.
Allianz pairs that capability with systems thinking, supported by Celonis, to connect customer touchpoints with back-office reality. "We can build all our transactions into a digital twin and then check if the processes we envision … really happen in real life," Malmendier noted. Learn more about the concept here: Digital twin.
What insurance leaders can do next
- Pick one high-volume, high-friction flow (e.g., FNOL to payout, endorsement changes, subrogation) and map the actual path with process mining.
- Instrument core systems to capture event logs with timestamps, case IDs, and user/system actions so you can measure reality, not opinions.
- Define the customer outcome in numbers: target cycle time, touchpoints, straight-through rate, and rework rate.
- Stand up a small digital twin pilot to test scenarios (policy complexity, fraud flags, vendor capacity) before rolling changes into production.
- Close the loop with alerts and automations that trigger on risk thresholds (aging claims, missing docs, idle queues) and track impact weekly.
The mindset shift
Allianz didn't just implement a tool. It shifted focus from internal convenience to customer excitement. As Malmendier put it, the goal is simple: products and processes people actually love - and operations that back that promise every day.
If you're building these capabilities in-house, curated learning paths can speed up adoption for claims, underwriting, and operations teams. Explore job-based options here: Complete AI Training: Courses by Job.
Your membership also unlocks:
