My mission as a computer scientist is to formally understand the underlying principles of real-world computer systems, thereby helping programmers to write and reason about those systems. For my Ph.D., I have primarily focused on formally understanding non-blocking concurrent programming, which is an essential ingredient for exploiting parallelism and is becoming more and more important since the slowdown of Moore's law. In the future, as an independent researcher, I would like to apply the understanding of concurrency to building practical verification/analysis tools for concurrent programs. I believe my research will have far-reaching impact on the way the IT industry develops system software. I am maintaining the Crossbeam project, which serves as the de facto standard concurrency library for the Rust programming language. I am also writing a compiler for Furiosa AI's MadRun deep learning accelerator.