Key Responsibilities
- Responsible for architecture design and optimization of intelligent driving core algorithms (perception, prediction, planning, control and other modules).
- Lead the design and implementation of multi-sensor fusion (Camera, LiDAR, Radar) algorithms to solve key technical problems such as target detection, tracking, and scene understanding.
- Develop high-precision positioning (SLAM), behavior prediction, path planning and motion control algorithms to adapt to L3/L4 autonomous driving requirements.
- Deploy algorithm models to vehicle embedded platforms to optimize real-time performance and resource usage.
- Build a simulation test framework (based on CARLA, etc.), design a scenario library and verify algorithm performance, and support the reproduction and tuning of real-car road test problems.
- Break through the algorithm bottleneck in long-tail scenarios (such as bad weather, unprotected left turns, and dense pedestrian interactions) and propose innovative solutions.
- Research end-to-end autonomous driving technology and the application of Transformer large models in perception and decision-making.
- Design, build, and maintain the foundational autonomous driving development platform, including Kubernetes cluster deployment and operations.
- Orchestrate cloud-native services using Infrastructure-as-Code tools such as Pulumi and Helm to ensure environment consistency and reusability.
- Lead the deployment, customization, and operation of workflow engines such as Flyte,nabling large-scale distributed training and simulatioPlatform & Infrastructure
- Design, build, and maintain the foundational autonomous driving development platform, including Kubernetes cluster deployment and operations.
- Orchestrate cloud-native services using Infrastructure-as-Code tools such as Pulumi and Helm to ensure environment consistency and reusability.
- Lead the deployment, customization, and operation of workflow engines such as Flyte, n task management.
- Build and optimize CI/CD pipelines with Argo CD and GitOps principles to achieve fully automated deployment and configuration management.
- Promote the fusion of simulation and real-world data to enhance system-level closed-loop validation capabilities.
- Build and manage a high-performance team, foster an agile engineering culture, and coordinate cross-functional collaboration across platform, algorithm, and test teams.
- Closely collaborate with teams across automotive electronics architecture, perception systems, and system validation to ensure seamless R&D integration.
- Take ownership of team development, including technical roadmap planning, personnel growth, and delivery quality control.
Qualifications
•Education: Master degree or above in computer science, automation, vehicle engineering, robotics and other related majors.
•Work experience:
-More than 5 years of experience in autonomous driving algorithm research and development, and has led the development of at least 1 mass production project module. Fully participate in the full cycle of L2+ to L4 projects, with patents or top conference papers (CVPR/ICRA/IROS, etc.). proven experience with end-to-end project delivery is preferred.
-Proficient in cloud-native architecture and tools, including Kubernetes, Docker, and Istio.
-Hands-on experience in deploying and optimizing workflow orchestration systems, such as Flyte.
-Familiarity with at least one CI/CD toolchain (e.g., Argo CD, GitLab CI/CD).
-Hands-on experience with autonomous driving simulation tools (e.g., CARLA, RoadRunner) and the ROS framework.
-Solid understanding of ADAS algorithm validation workflows; experience in hybrid (real + virtual) testing architecture is a plus.
-Familiarity with the application of deep learning models such as Transformers and Diffusion models in perception and reconstruction tasks is a strong plus.
-Experience in building data feedback platforms or data-driven development pipelines is highly desirable.
•Language: Proficient in English, can read technical literature and participate in international technical exchanges.
•Bonus points
- Familiar with the application of large models (such as BEV perception, DriveGPT) in autonomous driving;
- Experience in transplanting algorithms for automotive-grade chips (Orin/Xavier, TI TDA4);
- Participated in the construction of an autonomous driving data closed-loop system and is familiar with the entire process of data mining-labeling-training-deployment.