Customer Success

From manual driver-hunting to a refreshable executive narrative

Executive Summary

Zurich Insurance Group Reinsurance (Group Re) function produces executive-level performance reporting each quarter, a work that demands speed, accuracy and clear explanations of what changed and why.

Historically building the executive narrative involved extensive analysis across numerous data points, identifying key trends and preparing clear explanations, often resulting in significant time and effort from the team.

After discovering Zebra AI, Pablo Trachsler, Head of Business Development & Innovation, Group Re, set up Zebra AI on Zurich’s structured financial data and rebuilt the quarter-close executive summary workflow. In a parallel run Group Re significantly compressed the effort required to produce an executive story while maintaining the clarity and standards expected at the leadership level.

This was more than a productivity gain. It represented an operating model upgrade: turning quarter-close storytelling into a repeatable narrative workflow designed for executive decision-making.

“Zebra AI workload in Fabric has been a true game changer for us. Critical, recurring analytical processes that once consumed significant time and resources across multiple platforms are now completed in a fraction of the time. The gains in efficiency, speed and insight have been remarkable.

Having reviewed the broader market, it is clear to me that Zebra AI is significantly more mature than any alternative. From my perspective this solution is truly a game changer and delivers outstanding value to our organization.”

Customer

The Challenge

Quarter close in reinsurance isn’t simply producing numbers. It’s explaining them fast, consistently and in a way senior leaders can act on.

• That means breaking results down into true drivers. Even with modern BI in place, the last mile often remains manual. Teams still need to translate results into executive commentary that is coherent, defensible and decision-ready.
• Narrative formation is difficult to standardize, automate and refresh without breaking governance or losing interpretability.

What Zurich Did with Zebra AI

Pablo tested Zebra AI on structured finance data (~50,000 rows × 27 columns, later up to ~200,000 rows / ~3M datapoints) and rebuilt the executive summary workflow to standardize how close stories are produced and refreshed each cycle:

Results (Parallel Run)

• Close effort compressed dramatically
• Executive-ready narrative generated quickly once data is loaded/updated
• Reduced manual copy/paste and fewer error-prone production steps
• Repeatable workflow: update data → refresh story → review/refine → publish
• Validated on larger datasets up to ~200k rows / ~3M datapoints