AI Engineer, Model Distillation

Tesla AI is hiring an AI Engineer, Model Distillation to design and execute distillation pipelines that compress large teacher models into efficient students for autonomy and robotics.

Responsibilities

• Design and execute state-of-the-art distillation pipelines that transfer capabilities from large teacher models to ultra-efficient student models for FSD (Full Self-Driving), Optimus, and Digital Optimus
• Transfer reasoning, vision-language understanding, and action policies into compact students
• Develop and iterate on distillation methodologies across logit-level, sequence-level, and trajectory-level objectives
• Pioneer novel distillation recipes combining synthetic data generation, soft-label training, supervised fine-tuning (SFT), and Reinforcement Learning (RL)
• Run experiments to identify distillation scaling behavior, including teacher size, student capacity, student architecture design, data mixture, and compute allocation
• Measure how design choices impact capability retention, latency, and on-robot performance
• Build and maintain infrastructure for efficient teacher inference, distillation data curation, and distributed student training
• Resolve compute and memory bottlenecks end-to-end
• Evaluate distilled models for teacher fidelity and product metrics such as task success, robustness, and real-world behaviors
• Close performance gaps when students diverge from teachers
• Collaborate cross-functionally to ship distilled models to production while meeting performance, safety, and reliability standards
• Contribute tools and frameworks to keep distillation reproducible, measurable, and reusable across Tesla AI model families

Requirements

• Deep, proven expertise in deep learning fundamentals with experience in training, compressing, or distilling large-scale language, vision, or multimodal models
• Strong grasp of teacher-student interaction, synthetic data, and tradeoffs between capability, latency, and compute
• In-depth knowledge of loss design (example: KL / soft-target objectives) and modern architectures such as mixture of experts and hybrid attention
• Hands-on experience with post-training methods including distillation, supervised fine-tuning, and policy optimization / RL
• Expertise in distributed computing and large-scale training or inference pipelines
• Proficiency in Python and knowledge of software engineering best practices
• Experience with deep learning frameworks: PyTorch, TensorFlow, or JAX
• Demonstrated ability to work in a cross-functional team environment
• Strong problem-solving skills to troubleshoot complex issues across data, training, and deployment

Technologies

• Python
• PyTorch
• TensorFlow
• JAX
• Mixture of experts
• Hybrid attention

Compensation and Benefits

• Expected compensation: $124,000 - $558,000/annual salary + cash and stock awards + benefits
• Pay offered may vary based on market location and job-related knowledge, skills, and experience; total compensation may include other elements depending on the position offered
• Medical plans: plan options with $0 payroll deduction
• Family-building, fertility, adoption and surrogacy benefits
• Dental (including orthodontic coverage) and vision plans with options at $0 paycheck contribution
• Company paid HSA contribution when enrolled in the High-Deductible medical plan with HSA
• Healthcare and Dependent Care Flexible Spending Accounts (FSA)
• 401(k) with employer match, Employee Stock Purchase Plans, and other financial benefits
• Company paid Basic Life and AD&D
• Short-term and long-term disability insurance (90 day waiting period)
• Employee Assistance Program
• Sick and Vacation time (Flex time for salary positions, accrued hours for hourly positions) and Paid Holidays
• Back-up childcare and parenting support resources
• Voluntary benefits: critical illness, hospital indemnity, accident insurance, theft & legal services, and pet insurance
• Weight Loss and Tobacco Cessation Programs
• Tesla Babies program
• Commuter benefits
• Employee discounts and perks program

What to Expect

• Work on real-world autonomy at scale by training frontier-scale foundation models on multimodal telemetry, vision, language, and robotic action data
• Collaborate in an environment using driving, robotics, and human-interaction datasets spanning vision, language, action, and on-robot telemetry
• Study how capability transfers from large teacher models into compact students under strict latency, power, and reliability constraints
• Access high GPU resources per engineer to train frontier-scale teachers, generate large distillation corpora, and iterate on student architectures
• Run distillation experiments across language, multimodal, and policy models with high fidelity and scale, then deploy distilled models to FSD / Optimus in real-world settings