SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Physical Intelligence is developing foundation models and learning algorithms to power robots and physically-actuated devices. The company is seeking an Annotations Portfolio Lead to own and scale the data labeling operations that fuel model development.
In this role, you will manage end-to-end performance across a portfolio of annotation programs, driving quality, speed, and efficiency. You will partner directly with researchers to understand their data needs and translate those into concrete requirements and rubrics. You will identify and implement step-function improvements across the annotations function, balancing strategic problem-solving with tactical execution. You will foster clarity and collaboration across technical and nontechnical stakeholders, including vendor partners, researchers, and operations teams.
Key responsibilities include:
- Owning end-to-end performance for multiple annotation programs
- Partnering with researchers to define and refine data requirements
- Driving continuous improvement in operations and identifying scalable solutions
- Managing communication and alignment across remote, diverse stakeholder groups
- Bringing structure and organization to complex, multi-program workflows
Requirements:
- 5+ years of experience in program management, operations, technical project management, consulting, startup operations, or similar roles
- Strong problem-solving and systems thinking; ability to identify leverage points in evolving systems and diagnose problems with incomplete information
- Intuition for data quality and research needs, grounded in understanding of model training and evaluation
- Excellent communication skills; ability to clarify nuance, structure complex information, and work effectively with remote and ESL audiences
- Strong organizational skills to bring structure to chaos and balance competing needs across multiple programs
Bonus experience:
- Data annotations for ML/AI
- Robotics, egocentric video, or multimodal sensor data
- Using coding agents to fill gaps in tooling and infrastructure
