Kecheng Xu
Tech Lead Manager at Tensor Auto
Sunnyvale, California, United States
Summary
🤩
Rockstar🎓
Top School
Top expert inAutonomous Robotics and Computer Vision Development
Kecheng Xu is a tech lead manager with 8 years of hands-on experience building prediction, planning, and interactive decision systems for autonomous driving, currently leading end-to-end joint prediction and planning large-model efforts in Sunnyvale. He has driven production deployments at Baidu Apollo (a prominent open autonomous driving platform) and contributed as a top-three Apollo maintainer, improving planning modules and adding clarity to prediction code. His work spans classical planners (Lattice Planner, piecewise-jerk optimization) to deep learning models (CNN/LSTM/RNN-based trajectory and pedestrian prediction) and ML data pipelines on PyTorch, PySpark and Kubernetes. Kecheng also architected onboard/offboard training services, a siren detection system, and a reinforcement learning simulation stack, reflecting a rare blend of research citations and production impact. Trained in computational mathematics (Fudan, UT Austin), he pairs rigorous quantitative foundations with practical software engineering and open-source collaboration.
8 years of coding experience
8 years of employment as a software developer
Master of Science (M.S.) Computational Mathematics (was in Ph.D program), Master of Science (M.S.) Computational Mathematics (was in Ph.D program) at The University of Texas at AustinBachelor of Science (BS) Computational and Applied Mathematics, Bachelor of Science (BS) Computational and Applied Mathematics at Fudan University
Github Skills (7)
algorithm10
algorithms10
pathfinding10
c-language10
planning10
cprogramming-language10
computer-engineering9
Programming languages (3)
Github contributions (5)
ApolloAuto/apollo
Jul 2017 - Sep 2020
An open autonomous driving platform
Role in this project:
Back-end Developer
Back-end DeveloperContributions:3 reviews, 2062 commits, 1143 PRs in 3 years 3 months
Contributions summary:Kecheng primarily focused on enhancing the planning module of the open autonomous driving platform, as evidenced by contributions to the `dp_st_cost.cc` file with bug fixes and code style improvements. They also made significant refactoring changes to files associated with the planning module such as `qp_frenet_frame.cc`, and contributions to the `st_graph` and `dp_st_graph` files. In addition, the user introduced documentation and comments to the prediction module, which suggests a focus on improving code clarity and usability within the project.
autonomousmachine-learningautonomyautonomous-drivingapollo
kechxu/apollo
Dec 2018 - Sep 2020
An open autonomous driving platform
Contributions:6 PRs, 722 pushes, 125 branches in 1 year 9 months
autonomousmachine-learningautonomous-drivingmavlinkdriving
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.