Agentic AI Meets PL/I: Modernizing Mainframes with AWS Transform

This article demonstrates how AWS Transform and Kiro AI tools modernize legacy PL/I mainframe applications into cloud-native microservices, using a credit card transaction system as a case study.

• PL/I accounts for ~5% of applications, primarily in finance, insurance, and government sectors
• Key modernization challenges: talent scarcity, manual conversion complexity, business risk, and technical debt
• AWS Transform extracts business logic, dependencies, and documentation from PL/I code automatically
• Kiro synthesizes extracted rules into service specifications and generates code and infrastructure
• Three-phase approach: reverse engineering, forward engineering, deploy and test
• Batch processing transformed to event-driven architecture using Kinesis, Lambda, and RDS
• Real-time reporting replaces overnight batch file generation via QuickSight dashboards
• CardDemo transaction modernization completed in approximately 12 hours
• Kiro's direct AWS access and vibe mode enable autonomous debugging and problem resolution

AWS Transform and Kiro reduce mainframe modernization time and risk by automating code analysis, specification generation, and infrastructure deployment while preserving business logic.