AI accelerated legacy modernisation - Fabrice Bernhard - DDD Europe 2025

Fabrice Bernhard discusses leveraging AI to accelerate modernization of legacy systems, highlighting the risks associated with outdated technology through examples like Citigroup's massive erroneous transfer and the WannaCry ransomware attack. He shares a case study involving a company migrating multiple systems to a unified European platform, detailing how AI-driven dependency mapping and automated translation agents can significantly speed up the modernization process. Bernhard emphasizes the importance of crafting specific prompts for AI, making target architecture standards explicit, and fostering a culture of continuous improvement. The discussion also covers challenges like legacy archaeology, reliability rupture, feature freeze, and migration mutiny, along with strategies to address them using AI and lean thinking principles. The talk concludes with a Q&A session, addressing topics such as the relationship between existing and new systems, validation methods, and documenting learnings from AI prompting.