Robotics Machine Learning Engineer

Pantograph•San Francisco, CA
•Onsite

About The Position

Pantograph is building general-purpose intelligent robots that can safely and autonomously accomplish useful work in the real world. The Robotics Machine Learning Engineer will design and build systems that turn capable AI models into useful robots, combining frontier models, learned policies, and classical controllers. The role connects models to robot observations and actions, develops the tools, skills, controllers, and evaluations those systems need, and tests them in simulation and on Pantograph’s robot fleet. The engineer will investigate how models should act, coordinate with faster policies or classical controllers, and potentially create tools or control code. Decisions should account for capability, reliability, latency, safety, understandability, and supervision, with real-world deployments informing the work. The role works closely with research, robot software, and product teams to connect experimentation with experience on real hardware, and involves diagnosing failures across the system. Pantograph is a small, fast-moving team that works together in person in San Francisco. The company values clear thinking, careful execution, understandable and trustworthy systems, and practical results over credentials or complexity.

Requirements

  • Deep robotics experience across areas such as control, perception, planning, and learned policies.
  • Experience making robots perform useful work on real hardware.
  • Experience using simulation to develop and test systems, and understanding how simulation diverges from reality.
  • Experience designing and shipping robotic or AI systems used by people, with clear evidence of what improved them.
  • Strong experimental design and interpretation skills, including separating competing explanations, identifying confounders, and understanding what evidence establishes.
  • Strong intuition for model behavior, including how tools, context, and feedback affect it.
  • Enjoy investigating failures and understanding system behavior beyond aggregate metrics.
  • Pragmatic approach that prioritizes performance on real tasks over novelty or complexity.
  • Ability to build reliable software across model APIs, inference, robot software, and testing infrastructure.
  • Clear communication and ability to work across research and engineering.
  • Ability to use AI tools extremely effectively.

Nice To Haves

  • Experience with agent harnesses, tool use, memory, or long-horizon tasks.
  • Experience with systems where models write and execute code or develop their own tools.
  • Experience fine-tuning vision-language-action models or other large robot policies.
  • Experience with teleoperation, robot data collection, or learning from fleet data.
  • Experience with evaluation design, failure analysis, or safety testing.
  • Experience optimizing inference latency, throughput, and cost.

Responsibilities

  • Design and build systems that connect capable AI models to robots performing useful work.
  • Determine how frontier models, learned policies, and classical controllers should work together.
  • Connect models to robot observations and actions, and build the tools, skills, and controllers they rely on.
  • Develop evaluations in simulation and on the robot fleet.
  • Investigate when models should act directly, coordinate faster policies or classical controllers, or create their own tools or control code.
  • Assess tradeoffs in capability, reliability, latency, safety, understandability, and supervision, using real deployments to guide decisions.
  • Work with research, robot software, and product teams to create a feedback loop between experiments and real-hardware experience.
  • Design experiments, interpret results, and investigate system failures, identifying whether improvements are needed in observations, tools, orchestration, policies, controllers, or elsewhere.
  • Build reliable software across the stack, from model APIs and inference to robot software and testing infrastructure.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service