Applied Machine Learning Engineer

Pantograph•San Francisco, CA
•$150,000 - $260,000•Onsite

About The Position

Pantograph is building general-purpose intelligent robots that can safely and autonomously accomplish useful work in the real world. The Applied Machine Learning Engineer will help turn increasingly capable AI models into increasingly capable robots by developing tools, interfaces, and features that enable frontier models to perform real-world tasks. The role involves understanding what models need to observe, reason, and act; identifying performance breakdowns; and testing improvements on Pantograph’s robot fleet. The work includes building harnesses that connect models to robot observations and actions, designing useful tools, improving context and memory, and developing evaluations to measure progress. The engineer will explore approaches to organizing robot intelligence, including direct control by frontier models, coordination with faster specialized policies, and models writing their own tools or control code. They will assess tradeoffs in capability, reliability, latency, and system interpretability and supervisability. The role works closely with research, robot software, and product teams, creating a tight feedback loop between experiments and experience on real hardware. Pantograph seeks someone with strong experimental judgment and intuition for model behavior who enjoys building systems to test ideas and investigating whether failures require better observations, different tools, orchestration changes, or other system fixes. The team works in person in San Francisco.

Requirements

  • Excellent experimental design and interpretation skills, including separating competing explanations, identifying confounders, and understanding what evidence does and does not establish.
  • Strong intuition for model behavior, including what models understand, what confuses them, and how tools, context, and feedback affect behavior.
  • Experience building and improving AI systems that people use, with clear evidence of what made them better.
  • Comfort working across model APIs, data pipelines, inference, and application software.
  • Enjoys investigating failures and model behavior beyond aggregate metrics.
  • Ability to build reliable software around models, including tools, harnesses, and testing infrastructure.
  • Pragmatic approach to methods and focus on what works on real tasks.
  • Clear communication skills and enjoyment of cross-functional research and engineering work.
  • Ability to use AI tools 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 in classical robotics, including controls, motion planning, or perception.
  • Experience in robot learning, imitation learning, or reinforcement learning.
  • Experience with evaluation design, failure analysis, or safety testing.
  • Experience with inference optimization, including latency, throughput, and cost.
  • Experience learning from fleet data or operating ML systems across varied real-world environments.

Responsibilities

  • Develop tools, interfaces, and features that help frontier models accomplish real-world tasks with robots.
  • Build harnesses that connect models to robot observations and actions.
  • Design useful tools and improve model context and memory.
  • Develop evaluations to measure progress.
  • Explore and evaluate ways to organize robot intelligence, including direct model control, coordination with specialized policies, and model-generated tools or control code.
  • Evaluate tradeoffs in capability, reliability, latency, and system interpretability and supervisability.
  • Test improvements on the robot fleet and connect experiments with experience on real hardware.
  • Investigate model failures and determine whether they call for better observations, different tools, orchestration changes, or fixes elsewhere in the system.
  • Work across research, robot software, and product teams.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service