Sergei Solonets is a researcher and hands-on engineering leader with 11+ years building real-time visual systems, SLAM engines, and C++/CUDA tooling. He led and scaled a computer vision team, shipping a production-grade Direct Sparse Odometry derivative that is 30% faster than DSO and powers road-asset analytics for multiple US DOTs. Now pursuing a PhD at TUM under Daniel Cremers, he combines academic impact — including an ICLR 2024 Oral on direct image alignment — with practical deployments from national financial infrastructure integrations to nation‑wide ITS projects. His contributions to projects like tiny-cuda-nn and Hyperledger Iroha show a rare mix of low-level optimizer and robust backend engineering across ML and distributed systems. Comfortable mentoring, teaching, and making tradeoffs between research and product, he thrives where mathematical rigor meets production constraints.
11 years of coding experience
7 years of employment as a software developer
Study Abroad, Computer Science, Study Abroad, Computer Science at Seoul National University
Master of Science - MS, Computer Science, Master of Science - MS, Computer Science at Technical University Munich
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at Technical University of Munich
High School, Physics and Mathematics, High School, Physics and Mathematics at Lyceum 30
Bachelor's Degree, Applied Mathematics and Computer Science, Bachelor's Degree, Applied Mathematics and Computer Science at Peter the Great St.Petersburg Polytechnic University
Bachelor of Science - BS, Computer Science, Bachelor of Science - BS, Computer Science at Innopolis University
Contributions:39 commits, 44 PRs, 127 pushes in 8 months
Contributions summary:Sergei primarily worked on refactoring and updating the Iroha decentralized ledger project. They focused on replacing `nullptr` with `Result` to enhance error handling in core components such as storage. Their contributions involved modifying core files related to storage implementation and simulator integration and tests, demonstrating a focus on improving the robustness and maintainability of the system. Furthermore, they contributed to the codebase, updating tests to accommodate the changes in return types.
Contributions:2 reviews, 7 commits, 4 PRs in 25 days
Contributions summary:Sergei primarily contributed to the `tiny-cuda-nn` framework by implementing and modifying optimizers, particularly for neural networks. Their work involved addressing gradient flow issues and integrating new optimizer features. The commits demonstrate a focus on enhancing the framework's capabilities, particularly in the area of composite optimizers, which suggests a deep understanding of neural network training methodologies.
cudapytorchmlpcppopencl
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.
Request Free Trial
Sergei Solonets - Researcher at Technical University of Munich