Claude Code Modernizes a Legacy COBOL Codebase

Why It Matters

The demo shows how AI-driven tooling can rapidly de-risk and scale mainframe modernization—turning undocumented legacy code into maintainable modern services with repeatable verification, cutting time and hiring friction for enterprises with large COBOL estates.

Summary

Claude Code (via Cloud Code) was used to modernize a legacy COBOL-style (transcript: “Cobalt”) credit-card management codebase from an AWS mainframe demo by automating discovery, documentation, migration and verification. In phase one it scanned 94 files, produced more than 100 pages of documentation, generated indices and mermaid workflow diagrams, and mapped dependencies to reveal business logic such as interest calculations. In phase two it planned and executed a migration of a core batch interest-calculation program to idiomatic Java, recreating data models, IO layers, business logic and a dual test harness. Automated verification ran both COBOL and Java implementations against multiple test vectors and showed bit-for-bit fidelity across calculations, transformations and edge cases.

Original Description

Watch how Claude Code helps modernize a mainframe codebase.

Starting with code from an AWS Mainframe Modernization demo environment, Claude Code analyzes business logic, dependencies, and data flows, then helps refactor it into Java while preserving functionality. Claude Code handles the complex, multi-step work—generating documentation, developing migration plans, writing test suites, and iteratively validating the new code—directly from your terminal.