Allianz Partners

Allianz Partners processes over 90 million cases per year across health, automotive, and travel insurance — a scale where operational inefficiency compounds into a material financial and competitive liability. A single claim took an average of 29 days to resolve end-to-end, routed across dozens of disconnected systems with no unified decisioning layer. In P&C, slow resolution directly erodes customer retention and inflates loss adjustment expenses. The operating model scaled linearly with volume: absorbing more cases required proportional headcount growth, leaving no leverage for margin expansion. As CEO Tomas Kunzmann noted, the risk was losing half the value chain if the company failed to act quickly.

Allianz Partners deployed Otera's Autonomous AI agents (formerly DeepOpinion) across the full claims lifecycle, going live in UK and DACH markets within 2.5 months before a broader global rollout — bypassing proof-of-concept cycles entirely and treating the platform as core operational infrastructure from day one. Agents autonomously review incoming claims, validate coverage against policy terms, assess liability, settle cases, generate customer-facing responses, and trigger payment across multiple product lines, languages, and regulatory environments. Complex cases escalate to human specialists; the rest run end-to-end without intervention. A second deployment covers autonomous provider invoice management across Allianz Partners' 16,500-strong partner network, handling ingestion, service validation against contractual terms, and payment decisions — eliminating the need for headcount growth to absorb partner volume increases.

Average claims handling time fell from 29 days to 3.5 days, with a meaningful share of cases resolving in minutes. Key outcomes include:

• 70% of claims now settle in under 12 hours
• Customer satisfaction scores improved significantly in live markets
• Provider invoice processing scaled across 16,500 partners without proportional staffing increases
• Production deployments live across UK, DACH, and additional global markets

These results anchor a transformation program targeting €300M in incremental annual profit by 2027, set against a revenue growth trajectory from €10 billion to €13 billion. The financial targets are grounded in production-scale outcomes already running in market, not projections from pilots.

• Decision accuracy compounds across process steps: at 80% accuracy per step over 10 steps, end-to-end correctness approaches zero; at 99%, it reaches roughly 90% — making precise agent scoping non-negotiable for production viability.
• Bypassing the PoC cycle is a strategic choice: Allianz Partners treated autonomous AI as platform infrastructure from the start, accelerating time-to-value and avoiding the organizational inertia of incremental validation.
• Removing legacy logic matters as much as adding AI: layering old decisioning rules on top of autonomous agents undermines their effectiveness — the organizational transition is as important as the technology.
• Speed and margin move together in claims: faster resolution drives retention while lower cost per claim expands margin, making autonomy a single lever that addresses both simultaneously.