Reimagine mainframe workloads with a highly automated workflow using AWS Transform and Devin

This article describes how AWS Transform and Devin work together to automate mainframe modernization with full traceability from source code to requirements to cloud-native services.

• AWS Transform reverse-engineers COBOL, PL/I, and JCL to extract business rules with source-line traceability and technology-agnostic requirements
• Devin reads AWS Transform output to propose domain-aligned service decomposition and generates code, tests, and infrastructure with requirement identifiers intact
• Requirements flow through service specifications, Jira epics and stories, and generated code—each step maintains traceability back to original mainframe source lines
• Devin integrates with Jira, GitHub, Slack, and Teams; runs builds and tests locally; and opens pull requests for human review
• Parallel execution via Dynamic Workflows allows Devin to work on multiple independent tasks simultaneously across separate VMs
• Automated validation checks ensure every requirement maps to a specification and code, with acceptance criteria becoming automated tests

This workflow transforms mainframe modernization from years of manual reverse-engineering into an auditable, automated process where business rules remain verifiable throughout the migration.