> Markdown version of [/videos/1531-migrating-from-cobol-with-ai-a-moonshot-demo?t=402](https://www.wearedevelopers.com/videos/1531-migrating-from-cobol-with-ai-a-moonshot-demo?t=402). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Migrating from COBOL with AI: A Moonshot Demo

Still terrified of touching massive legacy mainframes? Discover how a fault-tolerant, multi-agent AI framework systematically reverse-engineers and translates complex COBOL programs into modern Java.

- **Speakers:** [Julia Kordick](https://www.wearedevelopers.com/@julia-kordick)
- **Event:** World Congress 2025
- **Published:** August 20, 2025
- **Duration:** 8:23
- **URL:** https://www.wearedevelopers.com/videos/1531-migrating-from-cobol-with-ai-a-moonshot-demo

## Summary

Modernizing legacy mainframe systems is notoriously complex and costly, especially when dealing with massive COBOL codebases that organizations rely on for critical operations. Instead of attempting a risky, full-scale mainframe exit, enterprise engineering teams can leverage multi-agent AI to selectively migrate specific COBOL programs into modern languages like Java. An open-source modernization framework, transitioning from Python-based tools to the robust orchestration capabilities of Semantic Kernel, enables this targeted approach. This allows financial institutions and large conglomerates to incrementally reduce mainframe dependencies while safely preserving core system functionality.

The migration process itself is driven by a specialized multi-agent architecture powered by Azure OpenAI. A dedicated analyst agent handles comprehensive reverse engineering, systematically deciphering complex legacy routines that heavily utilize unconditional "go to" commands across deeply nested program levels. Simultaneously, a dependency agent maps interactions between copybooks, files, and obsolete utility classes. Recognizing the immense compute time required for legacy code—often taking hours for small batches of files—the system automatically logs intermediate states into markdown files. This fault-tolerant design guarantees that if an orchestration failure occurs, the AI agents can seamlessly resume progress from the latest checkpoint without losing valuable deep-analysis context.

Ultimately, this is designed as a highly flexible AI framework rather than a rigid one-to-one translation tool. By simply adjusting agent personas and targeting specific prompts, engineering teams can configure the source and destination languages to fit any tech stack requirement. The process concludes by outputting clean codebase translations alongside comprehensive migration reports and Mermaid diagrams, giving technical leads immediate architectural visibility. This modular approach turns legacy modernization from an opaque, monolithic hurdle into an observable, incremental software engineering workflow.

**Keywords:** multi-agent ai modernization, cobol to java migration, legacy system modernization framework, semantic kernel orchestration, mainframe dependency reduction, legacy code reverse engineering, automated codebase dependency mapping, azure openai integration, fault-tolerant ai logging, ai agent persona generation, automated architecture diagramming, incremental mainframe exit, legacy codebase transition, ai-assisted software localization

## Chapters

1. **Multi-agent AI for migrating legacy COBOL to Java** (00:06) — An open-source project uses multi-agent AI to help financial institutions modernize legacy mainframe codebases.
1. **Accessing the repository and collaborating on COBOL problems** (01:08) — Developers can access the project resources on GitHub and collaborate on solving complex legacy codebase problems.
1. **Setting up the environment and executing the migration tool** (01:58) — The required setup involves configuring an AI endpoint and using a provided dev container for seamless execution.
1. **Understanding the multi-agent flow and COBOL code analysis** (03:19) — The initial analyzer agent performs reverse engineering and preparation steps on legacy source files to begin migration.
1. **Parsing dependencies and complex COBOL call chains** (04:03) — The dependency analyzer agent maps unconditional jumps and intricate call chains between legacy files.
1. **Orchestrating AI agents with the Semantic Kernel framework** (05:07) — The project utilizes Semantic Kernel in C# instead of Autogen to manage complex agent logs and orchestration effectively.
1. **Storing intermediate results for fault-tolerant AI migrations** (05:52) — The system stores intermediate state in markdown files to gracefully handle errors during long-running code conversions.
1. **Extending the framework beyond COBOL to Java migrations** (06:42) — The flexible framework allows developers to target different source and destination languages by adjusting agent personas.
1. **Reviewing generated Java code and migration documentation** (07:20) — An inspection of the output directory reveals the newly generated Java code, comprehensive migration reports, and architecture diagrams.

## Related Moments

- [Modernizing legacy COBOL mainframe systems using AI agents](https://www.wearedevelopers.com/videos/1365-wearedevelopers-live-the-weekly-developer-show-with-chris-heilmann-and-daniel-cranney) (from " WeAreDevelopers LIVE - the weekly developer show with Chris Heilmann and Daniel Cranney")
- [Migrating legacy COBOL systems using generative AI tools](https://www.wearedevelopers.com/videos/1766-devs-vs-marketers-cobol-and-copilot-make-live-coding-easy-and-more-the-best-of-live-2025-part-3) (from "Devs vs. Marketers, COBOL and Copilot, Make Live Coding Easy and more - The Best of LIVE 2025 - Part 3")
- [Analyzing COBOL mainframe migration challenges with AI](https://www.wearedevelopers.com/videos/100320-when-agents-meet-legacy-never-change-a-running-system) (from "When Agents Meet Legacy: Never Change a Running System")
- [Deploying AI agents for enterprise legacy code modernization](https://www.wearedevelopers.com/videos/100256-can-this-elephant-dance-ibm-bob-and-the-future-of-ai-first-software-development) (from "Can This Elephant Dance? IBM Bob and the Future of AI-First Software Development")
- [Leveraging artificial intelligence to accelerate system migration](https://www.wearedevelopers.com/videos/1444-crew-management-system-for-airlines-plan-duties-for-pilots-flight-attendants-worldwide) (from "Crew Management System for Airlines: Plan duties for pilots & flight attendants worldwide")
- [Evaluating AI agents for iterative legacy code modernization](https://www.wearedevelopers.com/videos/100320-when-agents-meet-legacy-never-change-a-running-system) (from "When Agents Meet Legacy: Never Change a Running System")

## Related Articles

- [MLOps And AI Driven Development](https://www.wearedevelopers.com/magazine/82-mlops-and-ai-driven-development)
- [What is Agentic Programming and Why Should Developers Care?](https://www.wearedevelopers.com/magazine/625-what-is-agentic-programming-and-why-should-developers-care)
- [How we Build The Software of Tomorrow](https://www.wearedevelopers.com/magazine/120-how-we-build-the-software-of-tomorrow)
- [Navigating the AI Shift](https://www.wearedevelopers.com/magazine/629-navigating-the-ai-shift)

## Related Jobs

- [Senior AI Agent Software Engineer (Go, Python) (m/f/x)](https://www.wearedevelopers.com/jobs/48277-senior-ai-agent-software-engineer-go-python-m-f-x) at **Dynatrace**
- [Staff Software Engineer, Copilot Experiences](https://www.wearedevelopers.com/jobs/ext/164361-staff-software-engineer-copilot-experiences) at **GitHub**
- [AI Software Engineer (Germany)](https://www.wearedevelopers.com/jobs/48317-ai-software-engineer-germany) at **Sunhat**
- [Principal Software Engineer, Enterprise AI Platform](https://www.wearedevelopers.com/jobs/ext/1467292-principal-software-engineer-enterprise-ai-platform) at **GitHub**
- [Senior Engineer, Infrastructure Platform](https://www.wearedevelopers.com/jobs/ext/328836-senior-engineer-infrastructure-platform) at **Intercom, Inc.**
- [AI Operations Manager (all genders)](https://www.wearedevelopers.com/jobs/48263-ai-operations-manager-all-genders) at **envelio**