Dimitar Asenov
Senior Software Engineer at Google
Zurich, Zurich, Switzerland
Summary
🤩
Rockstar🎓
Top SchoolDimitar Asenov is a senior software engineer with 14 years of experience bridging research and production-grade systems, currently building high-performance GPU compilation and ML tooling at Google. He combines deep academic training—a PhD from ETH Zurich—with hands-on expertise in C++ and Java to design developer tools, compilers, IDE features, and embedded software. His open-source contributions include extending the JAX and TensorFlow XLA GPU stacks with new MLIR dialect features and async/GPU operations, reflecting a knack for low-level code generation and performance tuning. Comfortable working across test automation, visual programming and version-control-aware tools, he brings both rigorous research thinking and pragmatic engineering discipline. Colleagues rely on him for complex refactors and reproducible test coverage that make advanced compiler features production-ready.
14 years of coding experience
2 years of employment as a software developer
High School Diploma, Information Technology, High School Diploma, Information Technology at Technology School Electronic Systems (TUES), associated with Technical University-SofiaBSc, Computer Science, BSc, Computer Science at Jacobs University BremenMSc, Computer Science, MSc, Computer Science at ETH Zürichtransferred to Jacobs University, Computer Engineering, transferred to Jacobs University, Computer Engineering at University of Duisburg-Essen
English, German, Bulgarian
Stackoverflow
26,036reputation
831kreached
5answers
10questions
visual-studio-code
top-1%
Github Skills (26)
visual-studio-code10
c-language10
python10
operation10
tensorrt10
gpu-programming10
machine-learning10
mlr10
tensorflow10
gpu10
cuda10
xla10
configurations10
tensor10
jax10
Programming languages (5)
Github contributions (5)
tensorflow/tensorflow
Jan 2025 - Jan 2025
An Open Source Machine Learning Framework for Everyone
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:2 comments in 5 days
Contributions summary:Dimitar primarily contributed to the XLA/Tensorflow project, focusing on code related to GPU compilation, particularly within the XLA (Accelerated Linear Algebra) framework. Their work involved optimizing code generation for GPU execution, addressing performance bottlenecks, and refactoring components related to the handling of constants. The user also worked on code generation for custom kernel fusions.
pythondata-sciencedeep-learningmlmachine-learning
jax-ml/jax
Mar 2025 - Mar 2025
Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more
Role in this project:ML Engineer
ML EngineerContributions:1 comment in 1 day
Contributions summary:Dimitar primarily contributed to the Mosaic GPU MLIR Dialect, a component of the JAX machine learning framework. Their work involved extending the dialect with new features like fragmented layouts for GPU processing. They added new operations, such as `async_load`, `async_store`, and WGMMA, along with corresponding tests to validate their functionality and ensure proper integration within the broader JAX ecosystem. Furthermore, the user also contributed to code formatting and refactoring efforts within the test code.
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.