Senior ML Engineer, Action Models for Physical AI

Atoms; San Francisco (CA); On-site; USD 273,000 - 321,000; Full time

Benefits offered by this job

Health insurance

401(k) plan

Equity

Job summary

Atoms is building real‑world robots to power the next era of progress. We seek a Senior Staff Machine Learning Engineer focused on Action Models to lead AI architecture and translate learned representations into actions in physical environments.

This deeply technical IC role influences research direction and long‑term AI architecture, with open-ended problems across perception, world models, and Robotics AI. SF office onsite.

Qualifications

• Deep expertise in ML with decision-making or embodied intelligence.
• Experience applying DL to sequential decision-making in physical systems.
• Background in reinforcement/imitation learning and planning.
• Ability to connect representations to downstream actions.
• Strong sequential/temporal modeling skills.

Responsibilities

• Define architecture for action models and learned decision-making.
• Develop models translating representations into decisions, plans, actions.
• Explore reasoning, planning, control for complex environments.
• Develop policies for real-world robotics and autonomous systems.
• Connect perception and world models to decision-making.
• Research imitation and reinforcement learning and other data-driven methods.

Skills

ML expertise

DL architectures

reinforcement learning

imitation learning

planning and control

robot learning

embodied AI

sequential/temporal modeling

large-scale data training

ML lifecycle

research-to-engineering translation

Python/C++

ambiguous research spaces

technical leadership

cross-disciplinary communication

Tools

Python

C++

Job description

Atoms is building real‑world robots to power the next era of progress. We seek a Senior Staff Machine Learning Engineer focused on Action Models to lead AI architecture and translate learned representations into actions in physical environments.

This deeply technical IC role influences research direction and long‑term AI architecture, with open-ended problems across perception, world models, and Robotics AI. SF office onsite.