TLM, AI Evaluation Science

What we’re doing isn’t easy, but nothing worth doing ever is.

We envision a future powered by robots that work seamlessly with human teams. We build artificial intelligence that enables service robots to collaborate with people and adapt to dynamic human environments. Join our mission-driven, venture-backed team as we build out current and future generations of humanoid robots.

The TLM, AI Evaluation Science will lead the team responsible for advancing the state of the art to measure the performance of physical AI systems, and measuring and validating how our AI systems perform in the real world. This group defines requirements, builds metrics, and creates rigorous evaluation pipelines. This work ensures that our robots meet high bars for safety, reliability, task performance and human trust. You’ll own simulation, testing, labeling, and interpretability frameworks, making sure our robots not only work, but work safely, repeatably, and explainably.

This is a hands-on leadership role in a startup environment. You’ll be both strategist and player-coach: defining evaluation standards, coding tools and models, and building the team that ensures our embodied AI is ready for deployment.

Responsibilities

• Lead the AI Evaluation Science team, owning evaluation strategy for robot perception, planning, control, and multimodal models.
• Define metrics and benchmarks for AI performance across safety, reliability, user experience, and robustness.
• Develop and maintain large-scale simulation environments to test robot behaviors under diverse real-world conditions (edge cases, adversarial scenarios, rare failures).
• Design evaluation frameworks that cover offline experiments, simulation, and live deployments.
• Build scalable pipelines for test coverage, automated evaluation, and regression tracking.
• Oversee labeling and data curation pipelines to generate high-quality ground truth for training and validation.
• Drive interpretability and explainability in embodied AI models—ensuring failures are measurable, diagnosable, and improvable.
• Collaborate closely with AI/Robotics engineering teams to define product requirements, set acceptance thresholds, and close the loop between evaluation and development.
• Actively mentor engineers and scientists while contributing hands-on to code, experiments, and metrics design.

Skills and Experience

• MS or PhD in Computer Science, Robotics, ML, EE, or related field along with 8+ years of AI/ML experience.
• Proven leadership experience: built and managed technical teams in AI, simulation, or robotics evaluation.
• Hands-on expertise building and evaluating large multimodal ML models (vision, language, action).
• Strong background in defining and operationalizing metrics for AI/robotics systems (safety, robustness, reliability).
• Demonstrated success in designing end-to-end evaluation pipelines: from data labeling and test definition to automated reporting and regression tracking.
• Experience in evaluation, benchmarking, or safety in robotics, AVs, or similar domains.
• Experience with simulation platforms for robotics or AVs
• Technical depth in ML interpretability, error analysis, and data-driven model improvement.
• Ability to operate in a startup context: strategic, but hands-on in code and experimentation.
• Excellent communication and cross-functional alignment skills—able to articulate risks, metrics, and trade-offs to executives, engineers, and non-technical stakeholders.