Autonomous Driving End-to-End Planning Algorithm Engineer

About Black Sesame Technologies

Founded in July 2016, Black Sesame Technologies is an AI digital imaging technology firm that creates solutions for real-world AI challenges. The company has developed a radically new chip and system to dramatically accelerate deep learning applications for autonomous.

We are innovating at every level of the stack – from chip, to new algorithms and network architectures at the cutting edge of ML research.  Our software stack is co-designed with the hardware and developed with scalability and quality in mind. Join us to create revolutionary Chips from the ground up.

Responsibilities

• Participate in the R&D and validation of end-to-end learning-based decision-making and motion planning models for autonomous driving, contributing to the development of next-generation advanced autonomous driving systems.
• Take charge of the design and training of end-to-end Planner models based on foundational architectures such as Transformer. Deeply explore and deploy generative paradigms—such as Autoregressive and Diffusion models—for trajectory generation and behavior prediction.
• Drive the application of Reinforcement Learning (RL) and Imitation Learning (IL) in autonomous driving decision-making and planning. Leverage cutting-edge algorithms (e.g., GRPO) to continuously enhance the model's reasoning and decision-making capabilities in complex, interactive, and game-theoretic scenarios.
• Build and refine the closed-loop simulation evaluation system. Promote large-scale testing, validation, and data closed-loop iteration of end-to-end models within simulation environments.
• Participate in the optimization and on-vehicle deployment of end-to-end models for mass-production projects. Drive the efficient utilization of high-quality data and successful engineering implementation.
• Collaborate with perception and prediction modules to optimize multi-modal feature fusion and model inference mechanisms, enhancing the stability and overall performance of the end-to-end decision-making chain.

Qualifications

• Master's degree or above in Computer Science, Automation, Robotics, Artificial Intelligence, or related fields. Possess a solid theoretical foundation in deep learning and strong mathematical modeling skills.
• In-depth, practical R&D and deployment experience in at least two of the following domains:
• End-to-End Decision & Planning Models: Familiar with foundational network architectures such as Transformer. Possess a deep understanding and practical experience in applying generative models (e.g., Autoregressive, Diffusion) to sequential decision-making.
• Reinforcement Learning & Imitation Learning: Solid theoretical foundation and hands-on system tuning experience in RL or IL.
• Closed-Loop Simulation & System Integration: Familiar with the overall workflow of end-to-end autonomous driving systems. Experience in building closed-loop simulation platforms and joint debugging across perception, planning, and control modules.
• Robust engineering implementation capabilities. Proficient in C++ and Python programming, with mastery of mainstream deep learning frameworks such as PyTorch.
• Familiar with ROS or common autonomous driving middleware. Deep understanding of core model evaluation metrics (e.g., closed-loop success rate, takeover/intervention rate). Candidates with a background in mass-production projects are highly preferred.
• Excellent engineering collaboration and systems thinking abilities. Capable of independently driving the R&D of cutting-edge models within a tightly coupled, multi-module environment.

Pay: $140,000.00 - $400,000.00 per year

Benefits:

• 401(k)
• Dental insurance
• Employee assistance program
• Employee discount
• Flexible spending account
• Health insurance
• Health savings account
• Life insurance
• Paid time off
• Referral program
• Retirement plan
• Vision insurance

Work Location: In person