This research introduces an autonomous system of LLM-powered agents that work together to upgrade legacy web applications without human intervention.

• Tackles the persistent challenge of maintaining outdated but business-critical applications
• Employs a multi-agent architecture where specialized LLMs handle different aspects of the modernization process
• Addresses both functionality preservation and security enhancement during upgrades
• Reduces costs and complexity barriers that typically prevent legacy system modernization

For engineering teams, this approach offers a scalable solution to technical debt while minimizing risk and human resource requirements in the modernization process.

Autonomous Legacy Web Application Upgrades Using a Multi-Agent System