Good science demands transparency, reproducibility, and rigour. The software underpinning it should be no different. In labs, hospitals, and research institutes, Rust is beginning to appear where it matters most: places where correctness and clarity aren't just nice-to-haves, but the foundations of trustworthy research.

This talk explores what it means to write scientific software that lives up to the standards we expect of science itself. We'll look at how Rust's emphasis on explicitness and safety aligns naturally with the principles of open, reproducible research, and how we can go further by treating tests as proof, documentation as methodology, and readable code as a form of scientific communication.

Drawing on examples from epidemiology, synthetic data, and biomedical infrastructure, we'll examine how to build tools that are auditable, maintainable, and built to last. We'll also reflect on how the choices we make today, in our dependencies, our environments, and our defaults, shape whether the next generation of researchers can understand, verify, and build on our work.