SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Salary: USD 275,000 - 321,000 / annual
Atoms is building Physical AI—real-world robots for industries including food, mining, and transport. The company integrates hardware, software, AI, operations, manufacturing, and real estate to deploy and scale autonomous systems in complex physical environments.
As a Staff Machine Learning Engineer focused on Perception, you will be a foundational technical leader in Atoms' AI organization. You will define how the company's machines transform raw sensor data into rich representations of the physical world, working across camera, LiDAR, radar, audio, and other sensor modalities to build perception systems for complex, dynamic real-world environments.
This is a deeply technical individual contributor role with significant influence over Atoms' technical direction. You will help determine what the perception stack should become, working closely with AI research, robotics, autonomy, and engineering leaders to develop perception architectures that connect traditional perception capabilities with emerging approaches in world models and foundation models for physical AI.
Key responsibilities include:
- Define and help build the technical architecture for next-generation perception systems
- Develop ML systems that transform multimodal sensor inputs into useful representations of the physical environment
- Advance multi-sensor and multimodal fusion approaches across complex real-world operating environments
- Design deep learning architectures for detection, segmentation, tracking, classification, scene understanding, and 3D perception
- Explore how modern foundation models and world models can complement or replace components of traditional perception pipelines
- Build perception systems designed to serve downstream reasoning, planning, and action models
- Make architectural decisions spanning data, model design, training, evaluation, inference, and deployment
- Develop systems that balance model quality with latency, efficiency, reliability, and compute constraints
- Establish evaluation methodologies and metrics that accurately measure perception performance in real-world operating environments
- Partner closely with researchers and engineers across world models, action models, robotics, autonomy, and ML infrastructure
- Provide technical leadership through architecture reviews, mentorship, technical direction, and hands-on engineering
- Help establish the technical bar for the growing perception organization and participate in identifying and assessing exceptional talent
The role is based in San Francisco with onsite work five days per week.
Requirements:
- Deep expertise in machine learning and computer vision with substantial experience building sophisticated perception systems
- Strong understanding of modern deep learning architectures and their application to visual and multimodal perception
- Experience working with one or more real-world sensor modalities such as cameras, LiDAR, radar, audio, depth sensors, or related systems
- Experience with sensor fusion, multimodal learning, 3D perception, scene understanding, or related areas
- Strong understanding of the full ML lifecycle, including data strategy, training, model architecture, evaluation, optimization, and deployment
- Experience designing ML systems that operate under real-world latency, compute, reliability, and safety constraints
- Strong software engineering fundamentals and ability to remain deeply hands-on in Python and/or C++
- Demonstrated ability to make consequential technical and architectural decisions in ambiguous problem spaces
- Track record of technical leadership and influencing engineering direction beyond immediate projects or team
- Ability to communicate complex technical ideas clearly and collaborate across research and engineering disciplines
