Accelerating Legacy Code Migration with Artificial Intelligence

Abstract: Organizations relying on critical systems with decades-old code face significant challenges, making modernization an operational imperative due to issues like operational stability, security, and a lack of updated features. This process of transforming legacy code is challenging, but is now being accelerated and augmented by Artificial Intelligence (AI) and large language models (LLMs). This research investigates the use of various LLMs for legacy code translation, aiming not for a perfect solution, but to significantly assist senior software developers by accelerating the development process, enabling rapid prototyping and initial implementation. This approach allows senior engineers to refine and productize the solutions, ensuring quality and alignment with system requirements. The initial testing strategy involved evaluating small subsets of legacy code within the Motif Framework, with the ultimate goal to demonstrate AI’s role as an assistive tool for senior developers in accelerating code modernization efforts.

Keywords: Artificial Intelligence, Large Language Models, Human Systems Integration, Legacy Code Migration

DOI: 10.54941/ahfe1006223

Cite this paper:

​