The Impact and Achievements of AI4SE in 2025 | The Research Blog
Blog post from JetBrains
AI4SE, a collaborative initiative between JetBrains and Delft University of Technology, aims to leverage AI advancements to transform software engineering practices, focusing on five key research tracks: testing and evaluating large language models (LLMs), adapting LLMs for coding tasks, creating interactive and aligned integrated development environments (IDEs), utilizing runtime information for development improvements, and developing intelligent teaching assistants for programming education. In 2025, significant progress was made, including the development of trigger models that optimize code completion processes, contributing to increased developer productivity. The program also supported graduate students' theses and facilitated impactful research projects, such as the creation of TreeRanker for code completion and AgentGuard for runtime verification of agentic AI systems. AI4SE's achievements in 2025, including award-winning publications and practical tools, underscore its role in reshaping software development and education, as it prepares to expand its impact in 2026 with new graduates and collaborative opportunities.