Robotics Hardware & Data Engineer

Calder•Redwood City, CA

About The Position

Calder is solving the data bottleneck for embodied AI and world models by capturing, processing, and delivering high-fidelity human-generated spatial data to train generalizable models for manipulation, navigation, and interaction. The Robotics Hardware & Data Engineer will design hardware and software interfaces for robot teleoperation and large-scale data collection. The role owns systems that enable human operators to control robots and convert each session into clean, high-fidelity training data. It spans mechatronics, embedded software, and machine-learning data pipelines, with hands-on work across motor drivers and training loops. The engineer will collaborate with engineering, operations, and research teams to ensure captured data supports model needs, and will feed learnings back into hardware and catalog design. The role involves rapid prototyping, debugging real hardware, and iterating on rigs, controllers, and capture setups.

Requirements

  • Background in mechanical or mechatronics engineering, robotics, or a closely related field.
  • Hands-on experience with motors, robotic control, and signal processing.
  • Familiarity with machine learning and robot learning, including understanding how data becomes a trained policy.
  • Comfort working across hardware and software, including embedded systems, real-time control, and data pipelines.
  • A builder's instinct and interest in making physical systems work reliably in the real world.

Nice To Haves

  • Experience with teleoperation systems or large-scale robot data collection.
  • Familiarity with the robot learning research community (e.g. CoRL) and modern imitation or reinforcement learning approaches.
  • Experience designing data schemas or catalogs for ML training.

Responsibilities

  • Design and build hardware interfaces for robot teleoperation, mapping human input to robot control in real time.
  • Develop data-collection systems that record synchronized, high-quality sensor, actuation, and video streams for downstream robot learning.
  • Work across motor control, signal processing, and low-latency communication between operator hardware and robots.
  • Build tooling and pipelines to validate, clean, and structure collected data for ML training.
  • Partner with engineering, operations, and research teams to ensure captured data meets model needs and feed learnings back into hardware and catalog design.
  • Prototype quickly, debug real hardware, and iterate on rigs, controllers, and capture setups.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service