Use cases › Customer service & contact center

Commerzbank "Bene": how the agent works and how to build one like it

Summary: Commerzbank's chatbot "Bene" has handled 2M+ chats and resolves 70% of them, per the source.

Source: https://cloud.google.com/transform/101-real-world-generative-ai-use-cases-from-industry-leaders source page fetched

Why this use case works

At this volume even modest resolution rates remove a large amount of contact-center work. A 70% resolution figure in banking also implies the bank defined clear boundaries for what the bot can and cannot do, since regulated topics need precise answers. (Our analysis.)

Agent pattern: Support chat agent

A support chat agent: a retrieval layer over help-center and policy content, tool calls into the systems of record (billing, orders, accounts), and explicit escalation rules. The quality of the knowledge base matters more than the model choice.

How a smaller team could replicate it

A regional bank or credit union can launch on the highest-volume, lowest-risk intents (card blocks, statements, branch info, how-to questions) and track resolution per intent before widening scope. (Our analysis, not a claim made by the source.)

Want a rough business case first? Try the AI agent ROI calculator, or ask us to scope it.

Risks and guardrails

Risks: wrong answers on regulated products and customers stuck in loops. Guardrails: approved answer sources for regulated topics, an always-visible route to a human, and loop detection that escalates after repeated failed turns. (Our analysis.)

What to measure

Resolution rate per intent, containment versus transfer, repeat contact within 7 days, and compliance review findings.

Related Customer service & contact center deployments

Part of the AgentStack Atlas catalog of 52 sourced agentic AI deployments. Spotted an error? Email agentstackatlas@mail.grokbot.com.