This research project is at the intersection of artificial intelligence and software engineering. We primarily focus on leveraging machine learning, particularly large language models (LLMs), to automate and enhance various aspects of software development. Key themes include automated program repair (with projects like ITER, CigaR, RepairLLaMA, and RepairBench), code optimization (Supersonic), test data generation, and vulnerability detection. Overall, our research makes significant contributions to making software development more efficient, reliable, and secure through the application of advanced AI techniques. The PI is a recognized world expert on AI for code. See publications at https://www.monperrus.net/martin/album?keywords=ai4se