Large language models (LLMs) are increasingly used for code generation but still often introduce subtle vulnerabilities. This poses serious risks in security-critical contexts, where system failures can be catastrophic. We present TypePilot, an agentic AI framework that leverages the Scala type system to guide and verify LLM-generated code. By embedding type-driven constraints into the generation process, TypePilot mitigates issues such as input validation flaws and injection vulnerabilities. Our results show that structured, type-focused pipelines enhance the trustworthiness of automated code generation in high-assurance domains.

About the speaker

Alexander Sternfeld

Associate Researcher at HES-SO Valais-Wallis (IEM)

Alexander Sternfeld is an associate researcher at the Reliable Information Lab, at the Institute of Entrepreneurship and Management (HES-SO Valais-Wallis). In this position, he is currently focusing on the safety of Large Language Models, both considering their evaluation and training procedures. Additionally, he works in technology monitoring, focusing mostly on bibliometric methods. He has obtained bachelors in both economics and econometrics at the Erasmus University Rotterdam. Afterwards, he shifted his focus towards machine learning through a Master’s in Data Science at EPFL, Lausanne.
Read more …