主要职责/Your Responsibilities:

Technical key words:

• Responsible for the design, development and delivery of the multi-modal AI algorithm module in autonomous driving systems.
• Participate in the construction of core algorithms for the data closed-loop and toolchain throughout the autonomous driving algorithm R&D process, including:
  • Assist in sorting out scenario suites, designing simulation use cases, optimizing data annotation and regression, to support rapid algorithm iteration.
  • Participate in the construction of AI infrastructure; assist in organizing and archiving basic equipment management, data assets and model assets, to ensure the sustainability and stability of the infrastructure.
• Based on actual autonomous driving business requirements and in response to the technological transformation of AI Agent, design and develop AI efficiency-enhancing tools and algorithm models to boost the efficiency of development, testing, validation and other work; assist in conducting AI-driven business compliance evaluation and release verification to support the implementation to SOP.
• Participate in the construction of the model-based closed-loop simulation platform; assist in developing simulation functions that meet the regulatory requirements for L3-level autonomous driving, to provide support for the technical validation of high-level autonomous driving.
• Track the cutting-edge technological trends in the fields of perception and AI; participate in the research and validation of advanced algorithms, explore technological innovation points, and assist the team in outputting intellectual property achievements such as academic papers and patents.

岗位要求/Required Qualification:

• Education: Bachelor's or master's degree in Computer Science, Electronic Engineering, Automation, Robotics, Photogrammetry and Remote Sensing, Computer Vision, Natural Language Processing, Artificial Intelligence or other related fields.
• Professions:
  • Master the basic theories of computer vision, good understanding of core visual perception tasks such as object detection, semantic segmentation, and multi-object tracking.
  • Be familiar with the fundamental concepts of deep learning, and understand common network architectures including CNN and Transformer, with basic knowledge of AI algorithm applications in autonomous driving scenarios.
  • Understand the core components of AI infrastructure (data assets, model assets, and basic equipment). Preference will be given to candidates with a preliminary understanding of the technical validation process for ADAS/AD projects.
  • Understand the basic regulatory requirements for L3-level autonomous driving, or have a foundational grasp of the core logic of closed-loop simulation platforms and data closed-loop processes; such candidates will be preferred.
  • Be familiar with the principles, deployment, and fine-tuning techniques of cutting-edge algorithms such as MLLM (Multi-module Large Language Models).
• Technical Skills:
  • Proficiency in programming languages such as Python or C++, with solid capabilities in code writing and debugging.
  • Familiarity with at least one deep learning framework (PyTorch/TensorFlow) or computer vision library (OpenCV). Preference will be given to candidates with relevant project experience, course assignments, or competition participation records.
  • Understanding of development tools including ROS, Docker, and MATLAB; familiarity with the usage or secondary development logic of at least one mainstream autonomous driving simulation platform (e.g., 51sim, CARLA, CarMaker, CarSim, etc.). Preference will be given to candidates with experience in simulation scenario construction and use case writing.
  • Basic practical experience in AI Agent related technologies: understanding of key technologies such as Prompt Engineering, Function Calling, and Multi-Agent collaboration. Preference will be given to candidates with experience in building simple AI Agent prototypes (e.g., task automation Agent, intelligent testing Agent) based on frameworks such as LangChain, LlamaIndex, and AutoGPT.
  • Basic data processing, analysis, and visualization capabilities, with the ability to assist in organizing algorithm test data, analyzing simulation results, conducting performance evaluation, and archiving assets. Preference will be given to candidates with an understanding of simulation data generation and scenario automatic generation technologies.
  • Understanding of application scenarios of AI tools in process optimization, or preference will be given to candidates with experience in using automated testing tools, data management tools, and simulation intelligentization tools (e.g., automatic scenario traversal, fault injection Agent).
• Preferred Qualifications
  • Priority is given to candidates with project experience (e.g., course design, competition projects, laboratory research projects) in parking scenario perception algorithms, camera calibration, multi-sensor data fusion, AI infrastructure construction, or closed-loop simulation platforms.
  • Priority is given to candidates who have participated in AI-driven process optimization, compliance evaluation, or related technical research.
  • Priority is given to candidates with academic paper publications, patent applications (including pending ones), or awards in relevant technical competitions.
  • Priority is given to candidates with a certain understanding of scenario suite development and technical verification processes for ADAS/AS SOP.
  • Doctoral degree is preferred.