My research investigates how artificial intelligence, and especially Large Language Models, can be applied to core software engineering tasks: code generation, translation, testing, verification, and trustworthiness assessment. A central thread is rigorous benchmarking: building the evaluation frameworks needed to measure what AI tools can and cannot do reliably in real engineering contexts. This work is part of a broader agenda to establish Benchmark Engineering as a formal discipline. See the dedicated Benchmark Engineering page for the full research agenda. I also write opinion and perspective pieces on what the AI revolution means for software quality, human expertise, and professional responsibility (see my blog Echoes of Saudade).