AWS re:Invent 2025 - Derisk your mainframe exit with agentic AI-powered refactoring ft BMW & Fiserv

AWS re:Invent 2025 - Derisk your mainframe exit with agentic AI-powered refactoring ft BMW & Fiserv

Overview

This presentation explores the challenges and solutions for enterprises looking to migrate and modernize their legacy mainframe systems. The focus is on leveraging agentic AI-powered refactoring to derisk the mainframe exit process, as demonstrated through case studies with BMW and Fiserv.

Key Challenges in Mainframe Modernization

• Complexity of legacy mainframe systems and applications
• Difficulty in extracting and understanding business logic embedded in COBOL and other legacy languages
• Lack of skilled resources familiar with mainframe technologies
• Risks and disruptions associated with a "big bang" migration approach
• Concerns around data integrity and application functionality during the transition

Agentic AI-Powered Refactoring

• Utilization of advanced AI and machine learning models to analyze and understand mainframe applications
• Automated extraction of business logic and data structures from legacy code
• Generation of cloud-native, microservices-based architectures as the target for modernization
• Continuous monitoring and adjustment of the refactoring process to ensure seamless transitions

Case Study: BMW's Mainframe Modernization

• Mainframe system supporting critical financial and supply chain operations
• Challenges included complex legacy applications, lack of documentation, and limited technical expertise
• Leveraged agentic AI-powered refactoring to:
  • Analyze and understand the existing mainframe applications
  • Automatically refactor the applications to a cloud-native, microservices-based architecture
  • Ensure data integrity and application functionality during the migration
• Achieved a 30% reduction in operational costs and a 50% improvement in system performance

Case Study: Fiserv's Mainframe Modernization

• Mainframe-based core banking system supporting critical financial services
• Faced challenges with scalability, flexibility, and integration with modern technologies
• Utilized agentic AI-powered refactoring to:
  • Decompose the monolithic mainframe application into a modular, cloud-native architecture
  • Automatically migrate and refactor the business logic and data structures
  • Seamlessly integrate the modernized system with new digital channels and services
• Resulted in a 40% reduction in time-to-market for new product offerings and a 25% improvement in customer satisfaction

Key Takeaways

• Agentic AI-powered refactoring can significantly derisk and accelerate the mainframe modernization process
• Automated analysis and understanding of legacy mainframe applications is crucial for successful migrations
• Transitioning to a cloud-native, microservices-based architecture enables greater flexibility, scalability, and integration with modern technologies
• Careful planning, continuous monitoring, and incremental migration strategies are essential for minimizing disruptions and ensuring a smooth transition