What challenges did Commerzbank face before AI implementation?

Commerzbank AG needed to modernize digital customer support, unify various outdated systems, ensure end-to-end regulatory compliance, elevate fraud protection, and maintain data security and transparency.

Source

MicrosoftProductionEvidence: Medium65/100

Commerzbank AG uses Microsoft Azure AI Foundry to build Ava, an AI-powered banking agent

Use case typeCustomer service automationUpdated Jun 13, 2026

Commerzbank AG modernized digital customer support by building Ava, an AI-powered banking agent leveraging Microsoft Azure AI Foundry and associated Azure AI services. They unified outdated systems, ensured regulatory compliance, enhanced fraud protection, and improved customer experience with Ava managing 30,000+ monthly conversations and resolving 75% autonomously. Ava uses Azure OpenAI for natural language understanding, Azure AI Content Safety for security and compliance, Azure Cosmos DB for synchronizing conversation context, Azure Kubernetes Service for scalable deployment, and Azure AI Search for trusted data retrieval. The solution delivers 24/7 personal banking experience, operational efficiency by reducing response times, and a scalable AI agent framework expanding to multifunctional banking workflows.

Organization: Commerzbank
Industry: Finance
Location: Germany

Why do we believe this?Outcome claims, sources, and evidence checks

Normalized claim

Quantified impact: 75%

Microsoft Customer StoriesCustomer storyInferred claimMedium evidence strength

Ava handles over 30,000 monthly customer conversations, resolving 75% autonomously, significantly reducing human agent workload.

Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks

Customer: Commerzbank
Provider: Microsoft
Maturity: Production
Linked source: Microsoft Customer Stories

The solution delivers 24/7 personal banking experience, operational efficiency by reducing response times, and a scalable AI agent framework expanding to multifunctional banking workflows

Customer identity supportedSource describes one deploymentMaturity supported

Primary read

Use case focus

Showing 3 of 3

1Conversational AI
2Autonomous Agents
3Customer Support Automation

Built Ava, a conversational AI banking agent on Microsoft Azure AI Foundry platform.
Implemented agentic orchestration and persistent memory for compliance and security.
Leveraged Azure OpenAI for advanced natural language processing to understand and respond empathetically.
Filtered sensitive data with Azure AI Content Safety to prevent fraud and ensure compliance.
Used Azure Cosmos DB to maintain conversation context across sessions.
Deployed on Azure Kubernetes Service for scalable, elastic resource management.
Integrated Azure AI Search to fetch accurate, compliant, and contextually relevant information for customer inquiries.
Added Azure Speech in Foundry Tools for text-to-speech avatar enabling voice conversational experience.

Technologies

Azure AI Foundry Azure OpenAI Azure AI Content Safety Azure Cosmos DB Azure Kubernetes Service Azure AI Search Azure Speech

Reduced response times and improved operational efficiency.

Architecture

Architecture leverages Azure AI Foundry Agent Service with agentic orchestration on Azure Container Apps, Azure OpenAI for language understanding, Azure AI Content Safety for fraud detection, Azure Cosmos DB for shared context, Azure Kubernetes Service for scalability, Azure AI Search for data retrieval, and Azure Speech for avatar voice interface.

Sources & evidence1

Evidence: Medium65/100Evidence strength