
Staff AI Research Engineer
Posted Aug 25

Posted Aug 25
This is a fully remote position, open to applicants in New York.
• Design and develop cutting-edge ML models and training pipelines.
• Utilize innovative machine learning techniques for various robotics applications.
• Create effective data management and training strategies.
• Implement frameworks for model evaluation and metrics tracking.
• Lead the development and refinement of models through swift experimentation and prototyping of new architectures.
• Enhance model performance and troubleshoot issues.
• Develop strategies for transfer learning and fine-tuning.
• Establish robust evaluation and debugging systems.
• Examine model behavior and identify failure modes.
• Deploy interpretability tools and visualization frameworks.
• Monitor and enhance model metrics.
• Work collaboratively with the engineering team to refine training infrastructure and deployment processes.
• Over 7 years of AI modeling experience specifically in the self-driving car sector, or a PhD with at least 3 years of AI modeling experience in the self-driving car field.
• Demonstrated success in developing and deploying large-scale ML models.
• Professional experience with diffusion and autoregressive models.
• Knowledge of training, inference, and infrastructure pipelines from camera input to trajectory output for self-driving or robotics applications.
• Experience with reinforcement learning (RL).
• Solid understanding of contemporary ML architectures and training methodologies.
• Proven experience in model debugging, optimization, and performance enhancement.
• Background in implementing ML research papers and adapting academic studies.
• Proficiency in PyTorch is essential.
• Familiarity with Python, NodeJS/Typescript, and Docker.
• Employee Stock Options for all Full-Time Employees.
• Paid time off.
• Medical insurance.
• Dental insurance.
• Vision insurance.
• Life insurance.
• Disability insurance.
• 401(k).
Improbable
Recursion
accesa.eu
TRAX
Get handpicked remote jobs straight to your inbox weekly.