Benchmarking is among the most consequential activities in computer science, yet it has never matured into a discipline of its own. For over two decades, my research has examined how benchmarks are designed, what makes them fail, and what principled foundations would look like across domains including security, dependability, and AI. This page brings that work together under a single agenda: Benchmark Engineering. I also write opinion and perspective pieces on these challenges on my blog, Echoes of Saudade.