Jaroslav Sevcik
Principal Software Engineer at NVIDIA
Munich, Bavaria, Germany
Summary
🤩
Rockstar🎓
Top SchoolJaroslav Sevcik is a Principal Software Engineer based in Munich with 11 years of professional experience and a PhD in Computer Science. He has driven low-level engineering and developer tooling across industry leaders—most recently at NVIDIA on JAX and previously as a Staff Engineer at Google leading the Chrome DevTools JavaScript debugger and Stadia CPU developer tools. His work spans JVM/LLDB integration, JIT compiler internals, performance profiling, and debugging UX, with notable open-source contributions to V8’s debugger and Chrome DevTools UI. He also added rigorous ML runtime tests and memory-offloading coverage to JAX, showing breadth from systems and compilers to ML infrastructure. Colleagues rely on him for shipping production-quality, security-conscious optimizations (including Spectre mitigations) and scaling debuggers to AAA workloads. He pairs deep systems-level expertise with a practical bent for improving developer productivity and observability.
11 years of coding experience
21 years of employment as a software developer
PhD, Computer Science, PhD, Computer Science at The University of EdinburghZakladni skola Kridlovicka, BrnoMgr., Compute Science, Mgr., Compute Science at Charles University
Github Skills (41)
debugging10
debug10
javascript10
c-language10
python10
testing10
gpu-programming10
machine-learning10
typescript10
virtual-machine10
ui-design10
interpreter10
sourcemaps10
tpu10
ml10
Programming languages (9)
Github contributions (5)
ChromeDevTools/devtools-frontend
Aug 2021 - Jan 2023
The Chrome DevTools UI
Role in this project:
Front-end Developer
Front-end DeveloperContributions:65 commits in 1 year 5 months
Contributions summary:Jaroslav primarily contributes to the Chrome DevTools UI, as evidenced by their commits to the `devtools-frontend` repository. Their work involves modifying UI elements, such as filter search cancel buttons, and implementing features like inline variable value display. The user demonstrates expertise in JavaScript, TypeScript, and related UI technologies, as seen in changes to files related to UI components and source map handling. A significant portion of their work focuses on debugging tools and features, with many commits targeting the improvement of debugging experiences such as source map integration and inline variable display.
instrumentationreactjavascriptchromecdp
jax-ml/jax
Jan 2024 - Feb 2025
Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more
Role in this project:ML Engineer & QA Engineer / Test Automation Engineer
ML Engineer & QA Engineer / Test Automation EngineerContributions:13 reviews, 12 PRs, 41 comments in 1 year
Contributions summary:Jaroslav primarily focused on enabling and testing features related to memory offloading and rematerialization (remat) within the JAX framework. Their contributions involved adding tests for GPU and TPU backends, particularly related to the `device_put` operation and scan functionality. Furthermore, they were involved in improving the testing infrastructure and test coverage for layout and sharding features.
pytorchpythonjitautomatic-differentiationgpu
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.