Alexander Belyaev
Staff Software Engineer at Google
Bavaria, Germany
Summary
🤩
Rockstar🎓
Top SchoolAlexander Belyaev is a Staff Software Engineer based in Bavaria with eight years of experience building compiler and ML runtime infrastructure at Google. He specializes in MLIR/LLVM toolchains, lowering and optimizing Linalg and vector operations, and integrating compiler backends for TensorFlow, IREE, and XLA. His contributions span high-impact open-source projects—fixing complex build-system bugs in LLVM, implementing numerically stable lowering for complex operations, and shipping passes that improve JIT and CPU codegen. Earlier work on Google Cartographer shows hands-on robotics and SLAM experience, including proto-driven landmark propagation and ROS visualization. With a strong academic background (MS and PhD in mathematics), he brings rigorous numerical thinking to practical compiler engineering problems. Colleagues rely on him for robust, low-level fixes that prevent subtle runtime regressions while advancing performance-sensitive ML stacks.
8 years of coding experience
12 years of employment as a software developer
Master of Science (MS), Industrial Mathematics, sehr gut, Master of Science (MS), Industrial Mathematics, sehr gut at Technische Universität KaiserslauternBachelor of Science (BS), Applied Mathematics and Informatics, excellent, Bachelor of Science (BS), Applied Mathematics and Informatics, excellent at Far Eastern Federal University
Russian, English, German
Github Skills (40)
robotics10
c-language10
protocol-buffers10
llvm10
3d-mapping10
mlr10
rviz10
proto10
localization10
compiler-compiler10
code-generation10
lg10
tensorflow10
bazel10
compiler-optimization10
Programming languages (10)
Github contributions (5)
cartographer-project/cartographer
Nov 2017 - Jun 2020
Cartographer is a system that provides real-time simultaneous localization and mapping (SLAM) in 2D and 3D across multiple platforms and sensor configurations.
Role in this project:
Back-end Developer & Robotics Engineer
Back-end Developer & Robotics EngineerContributions:146 commits, 281 PRs, 140 pushes in 2 years 8 months
Contributions summary:Alexander primarily contributed to the development and improvement of the Cartographer system, focusing on the implementation of core features and data structures related to Simultaneous Localization and Mapping (SLAM). They implemented new data structures and protocols, particularly the 'LandmarkData' structure and its associated proto definitions. Furthermore, the user worked on propagating landmark data to the pose graph and sending landmark data via gRPC. Their work also involved refactoring and refactoring code organization to improve performance and code structure.
3dlocalizationmappingroboticsself-driving
openxla/xla
Jul 2019 - Jan 2023
A machine learning compiler for GPUs, CPUs, and ML accelerators
Role in this project:Back-end Developer
Back-end DeveloperContributions:45 reviews, 137 commits, 2 PRs in 3 years 7 months
Contributions summary:Alexander primarily contributed to the XLA compiler's CPU and GPU backends, focusing on implementing and refining features related to fast math and code generation within the XLA ecosystem. Their work included adding flags to control and optimize CPU-specific functionalities like approximating functions, division handling, and the enablement of fast math features. Furthermore, the user addressed issues in the GPU backend, resolving asymptotic behavior problems in specific mathematical functions and optimizing fusion of operations.
compilermachine-learning
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.