Legacy AI

Development of an AI-based system to continue using outdated but still necessary programming languages. Publicly available code as well as company-specific code will be used as training data. The system should be integrated into common development environments via plugins and enable the automated migration of entire programs to higher versions of the programming language. The development involves an AI pipeline for the automated refinement of a pre-trained Large Language Model (LLM) with company-specific training data, allowing the underlying model to be replaced when necessary. New training data can be easily added, even by non-AI experts. The system analyzes user behavior and continuously improves through self-learning. Intelligent cloud scaling enables parallel access by several thousand users.

Project Lead

Prof. Dr. Munir Georges
Prof. Dr. Munir Georges
Phone: +49 841 9348-2331
Room: K305
E-Mail:
Prof. Dr. rer. nat. Stefan Kugele
Phone: +49 841 9348-2481
Room: K305
E-Mail: