Machine Learning Engineer, Foundation Model at DiDi Labs

San Jose, California, United States -

Full Time

Start Date

Immediate

Expiry Date

25 Jun, 26

Salary

247038.0

Posted On

27 Mar, 26

Experience

5 year(s) or above

Remote Job

Yes

Telecommute

Yes

Sponsor Visa

Skills

Machine Learning, Deep Learning, Robotics, PyTorch, JAX, TensorFlow, Transformers, Diffusion Models, Flow-based Models, Decision Models, VLM, VLA, Python, C++, Large-scale Training, Data Pipelines

Industry

IT Services and IT Consulting

Description

About The Company DiDi's autonomous driving unit was established in 2016 with the mission of developing Level 4 autonomous driving (AD) technology to make transportation safer and more efficient. In August 2019, the unit became an independent company, DiDi Autonomous Driving, dedicated to advanced AD R&D, product application, and business expansion. We believe integrating AD technology into a shared-mobility fleet will generate immense social value. By leveraging DiDi's specialized technology, operational expertise, and integrated ecosystem, we are positioned to build and operate a highly efficient, user-oriented autonomous fleet. About The Role The Foundation Model Team focuses on building large-scale foundation models for multi-agent behavior prediction and autonomous vehicle planning. By leveraging DiDi Voyager’s unparalleled driving data, we train highly robust and generalizable deep learning systems that enable safe and intelligent autonomous driving at scale. Our models serve as the core intelligence of the autonomous driving stack, enabling vehicles to understand complex traffic scenarios, anticipate agent behavior, and make safe and efficient driving decisions. We operate at the intersection of large-scale machine learning, autonomous driving, and foundation model research, pushing the frontier of multi-agent prediction and planning. Responsibilities As a member of the Foundation Model Team, you will: Design and train large-scale deep learning models for: Multi-agent trajectory prediction Behavior and intent prediction Planning and decision-making Build foundation model architectures (Transformers, Diffusion, Flow-based models, Decision models, VLM/VLA) Develop scalable training pipelines across hundreds to thousands of GPUs Work with massive real-world datasets and build high-quality data pipelines Optimize models for latency, reliability, and on-vehicle deployment Collaborate closely with perception, mapping, simulation, and systems teams Drive research ideas into production systems used by real autonomous vehicles Qualifications Strong background in machine learning, deep learning, or robotics Experience with PyTorch / JAX / TensorFlow Solid understanding of modern neural architectures (transformers, diffusion, auto-regressive) Strong coding skills in Python and C++ Passion for building real-world, safety-critical AI systems Preferred Qualifications BS, MS or PhD in Computer Science, Machine Learning, Robotics, or a related field Experience in autonomous driving, robotics, or embodied AI Experience training large models on distributed GPU clusters Experience with trajectory prediction, planning, or decision-making systems Publications in top ML / robotics conferences (NeurIPS, ICML, ICLR, CVPR, RSS, CoRL, etc.) The base salary range for this position is $129,189-$247,038 annually in addition to bonus, equity and benefits. Our salary ranges are determined by role, level, and location. Within the range, individual pay is determined by work location and additional factors, including job-related skills, experience, and relevant education or training. I acknowledge that prior to submitting this application, I have read and accepted the Privacy Notice for California Residents which is available on https://v.didi.cn/AQnxlBa

Responsibilities

The role involves designing and training large-scale deep learning models for multi-agent trajectory prediction, behavior prediction, and planning, utilizing foundation model architectures like Transformers and Diffusion models. Responsibilities also include developing scalable training pipelines across many GPUs and optimizing models for deployment on autonomous vehicles.