Algorithm & Software Architecture Engineer

该职位来源于猎聘 Position Overview We are looking for an Autonomous Driving E2E Algorithm & Software Architect with solid experience in autonomous driving algorithms, software architecture, and production development.The role focuses primarily on End-to-End autonomous driving algorithm architecture, while also requiring practical knowledge of modular architectures across perception, prediction, planning, and control.The ideal candidate should be able to bridge algorithm design and production software implementation, covering model architecture, inference deployment, software integration, performance optimization, and production delivery.

Key Responsibilities

• Design the software architecture and drive the engineering deployment of end-to-end autonomous driving algorithms (One-Model end-to-end, VLA, World Model, etc.), connecting the full pipeline from trained model output to in-vehicle production deployment.
• Design and develop the software architecture for modular algorithm components (perception, prediction, or planning), ensuring efficient integration and stable operation of each module within the overall vehicle system.
• Own model compression, quantization, and chip adaptation (e.g., NVIDIA Orin/Thor, Horizon Journey series) and real-time performance tuning, ensuring on-vehicle deployment meets real-time and compute-resource constraints.
• Conduct hands-on technology research, POC validation, and implementation for candidate solutions, and produce technical evaluation materials to support team decisions.
• Build and maintain the software frameworks, deployment toolchains, and SDKs required for algorithm engineering, improving iteration efficiency and deployment consistency.
• Support the build-out and integration testing of data-closed-loop and simulation-closed-loop platforms to enable fast validation and continuous algorithm iteration.
• Work closely with algorithm, system, and test teams to identify and resolve engineering issues (performance bottlenecks, accuracy loss, latency, etc.) in end-to-end algorithm production deployment. Qualifications
• 8 years of experience in software development, system design, or software architecture.
• 5 years of experience in autonomous driving/ADAS development with proven production experience.
• Hands-on experience in at least one L2 /L2 production program.
• Practical experience in E2E autonomous driving algorithm development, architecture design, or deployment.
• Strong understanding of modern E2E driving architectures.
• Solid knowledge of Transformer, Attention, temporal modeling, and multimodal fusion.
• Good understanding of BEV, Occupancy, Prediction, and Planning architectures.
• Hands-on experience deploying AI models to vehicle-side compute platforms.
• Strong programming skills in Python and C .
• Solid Linux development experience.
• Strong debugging and problem-solving capabilities across algorithms, models, software, and systems.

  Preferred

  Qualifications

• Experience with production E2E autonomous driving systems.
• Experience with Planning-oriented E2E or Unified E2E architectures.
• Experience with Diffusion, Flow Matching, World Models, VLM/VLA, or driving foundation models.
• Hands-on experience with BEV, Occupancy, E2E Planning, or trajectory prediction.
• Experience with TensorRT, ONNX Runtime, CUDA, or similar inference runtimes.
• Experience in model quantization, distillation, compression, or inference optimization.
• Experience deploying and optimizing E2E models on vehicle-side hardware.
• Experience with ROS/ROS2, AUTOSAR, or automotive software frameworks.
• Experience working across autonomous driving algorithms and production software.