5+ years of experience in deep learning for robotic manipulation in industrial settings
Master's or PhD in Robotics, Computer Science, or a related field
Strong understanding of algorithms in behavior cloning and reinforcement learning
Proficiency in Python, C++, or Rust for software development
Experience with cloud computing and large-scale datasets
Responsibilities
Design and implement advanced robot learning models for dexterous manipulation
Own the ML manipulation pipeline and roadmap for industrial robustness
Collaborate with teams to design data collection and sensor strategies
Leverage existing manipulation models and develop new architectures
Deploy trained models to robotic hardware ensuring real-time performance
Stay updated with the latest in ML research and development
Develop evaluation pipelines for testing manipulation models
Benefits
Full-time employment with a focus on impactful work
Opportunity to work on cutting-edge humanoid robotics
Collaborative environment fostering team growth
Engagement with state-of-the-art technology and research
Potential for travel to enhance project execution
Full Job Description
Job Title: Staff Engineer, Deep Learning Manipulation
Department: Software Engineering
Employment Type: Full-Time
FLSA Status: Exempt
Location: Houston, TX
Travel: 10%
About the Role
We're looking for a Staff Deep Learning, Robot Manipulation Engineer to help build and execute Persona AI's robot learning capability and roadmap. We're looking for exceptional talent who dream big, thrive on challenges, and love seeing their efforts make an impact on Persona's humanoid robots doing real work in real-world environments.
You will drive and implement Persona's learning-based manipulation direction, research and establish the best learning formula for Persona's humanoid use cases such as welding and grinding, and continuously improve the robot fleet's performance to achieve industrial-level robustness.
If you want to push the boundary of humanoid robot capability performing real-world tasks, this role is for you.
What You Will Be Doing
Design and implement advanced robot learning models and training procedures to achieve dexterous manipulation for humanoid robots with high DOF and multi-fingered hands.
Own the ML manipulation pipeline and roadmap, combining different learning paradigms to achieve real-world, industrial-level robustness.
Work with the Teleoperation and Data team to design data collection, sensor modularity, and data mixture strategies.
Leverage existing state-of-the-art manipulation models and contribute to the development of new architectures for emerging complex tasks.
Deploy trained manipulation models to robotic hardware, ensuring real-time performance, safety, and integration with control systems and sensors.
Keep up to date with the state of the art in ML research and development.
Develop and execute evaluation pipelines to rigorously test learned manipulation models, including real-world trials and simulation, measuring performance, robustness, and generalization across tasks and environments.
Collaborate in attracting, nurturing, and growing the machine learning team.
What We Are Looking For:
5+ years of experience applying deep learning techniques to robotic manipulation in industrial settings.
Master's or PhD in Robotics, Computer Science, or a related field.
Courage and grit to tackle some of the most challenging problems in robot manipulation.
Strong understanding and proficiency with state-of-the-art algorithms and best practices in behavior cloning, reinforcement learning, vision-language-action models, foundation models, etc.
Experience with cloud computing and large-scale datasets.
Capable of writing high-quality software in Python, C++, or Rust.
Bonus Skills:
Experience with other aspects of ML applied to robotics, including robot simulation, locomotion, computer vision, robot hardware, and sensors.
Published papers at top ML/Robotics conferences (ICML, ICRA, CoRL, RSS, NeurIPS, etc.).
Have trained models and deployed them to robots in production environments.
