Senior Applied Scientist, Robot Capability Evaluation, Fauna

The Robot Age; Sunnyvale (CA); Hybrid; USD 130,000 - 210,000; Full time

Job summary

Fauna is seeking a senior applied scientist to define how robot capabilities are measured, stress-tested, and improved. You will shape evaluation metrics, data collection protocols, teleoperation workflows, and task definitions to drive model development for real homes.

The role sits at the intersection of research infrastructure and product impact, ensuring benchmarks determine when robot behaviors are ready for deployment and real-world use.

Qualifications

• Applied science experience in robotics, machine learning, evaluation, or human-in-the-loop systems.
• Ability to turn ambiguous capability goals into measurable tests.
• Comfort with experimental design and systems constraints.
• Experience connecting evaluation data to model or policy development.

Responsibilities

• Design evaluation frameworks for robot tasks, edge cases, and real-world scenarios.
• Define benchmarks, task structures, and success criteria for embodied policies.
• Shape data collection protocols that support model improvement.
• Iterate on teleoperation and operator-facing workflows for reliable evaluation.

Skills

Applied science in robotics

Machine learning

Evaluation design

Human-in-the-loop systems

Job description

What the role actually is

Fauna is hiring a senior applied scientist to define how robot capabilities are measured, stress-tested, and improved. The posting ties evaluation to data collection protocols, teleoperation workflows, task definitions, and model development.

This is a research infrastructure role, but it is close to product reality because the metrics determine which robot behaviors are considered ready for real homes.

What you would work on

• Design evaluation frameworks for robot tasks, edge cases, and real-world scenarios
• Define benchmarks, task structures, and success criteria for embodied policies
• Shape data collection protocols that support model improvement
• Iterate on teleoperation and operator-facing workflows for reliable evaluation

What they are asking for

• Applied science experience in robotics, machine learning, evaluation, or human-in-the-loop systems
• Ability to turn ambiguous capability goals into measurable tests
• Comfort with both experimental design and systems constraints
• Experience connecting evaluation data to model or policy development

Why this one is worth a look

Evaluation is becoming one of the core bottlenecks in embodied AI. This role is worth tracking because whoever owns the benchmark effectively shapes what the robot is allowed to become.