Robotics AI Research Engineer - CoRL

Maven Robotics•Santa Clara, CA
•$175,000 - $350,000

About The Position

Maven Robotics is seeking AI researchers and engineers to architect the neural backbone of its general-purpose robots. The role focuses on designing Vision-Language-Action (VLA) frameworks or Action World Models that enable robots to reason, adapt, and perform tasks in complex industrial environments. Maven is hiring across career levels, from early-career through staff and principal, and seeks candidates with world-class expertise in at least one core area, such as model architecture, reinforcement learning formulations, or high-scale data systems. The position combines advanced research with real-world engineering, developing models and learning methods for physical manipulation and deploying them in industrial robotics systems. Compensation is an annual base salary of $175,000–$350,000 USD plus equity, with comprehensive health benefits, a 401(k), and other employee benefits.

Requirements

  • MS or PhD in Computer Science, Robotics, Machine Learning, or a related field, or equivalent industry experience.
  • Advanced understanding of transformers, multimodal alignment, and mapping perception to high-frequency motor control.
  • Proven ability to innovate in one or more areas: VLA models, real-world reinforcement learning, or large-scale data infrastructure.
  • Expert-level Python skills and deep familiarity with PyTorch or JAX.
  • World-class excellence in at least one core factor, such as model architecture, reinforcement learning formulations, or high-scale data systems.

Nice To Haves

  • High-impact publications in venues such as NeurIPS, ICRA, RSS, or CVPR, or significant open-source contributions.
  • Experience with large-scale distributed training and model compression.
  • Experience deploying models to edge devices such as NVIDIA Jetson or Orin for real-time inference.
  • General knowledge of robotics principles, including kinematics and dynamics.

Responsibilities

  • Design embodied model architectures, including VLA models and World Models, to achieve high task success with minimal human demonstrations.
  • Develop co-training strategies and efficient learning algorithms that use diverse data sources, including Internet-scale video and human interventions.
  • Build models that can adapt zero-shot or few-shot to new robot configurations and maintain success as hardware and actuation systems evolve.
  • Formulate and deploy reinforcement learning and policy extraction methods for real-world physical manipulation.
  • Collaborate on data collection systems that capture human intervention data for model bootstrapping.

Benefits

  • Equity
  • Comprehensive health benefits
  • 401(k)
  • Other employee benefits
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service