Humanoid robot control and manipulation software for industrial deployment
Persona AI is building the control and perception stack for humanoid robots, with a tech foundation spanning simulation (MuJoCo, Isaac Sim), robotics middleware (ROS, MoveIt), and deep learning (PyTorch, JAX, TensorFlow). The stack shape—heavy on physics simulation, manipulation algorithms, and teleoperation—reflects a company focused on solving the hard problems of getting humanoids to work reliably in manufacturing. Hiring skews heavily toward senior engineers and interns (11 of 16 roles), with almost no mid-level or leadership depth yet, suggesting a team still in rapid scaling mode while managing tight talent constraints.
Notable leadership hires: Head of Talent
Persona AI, founded in 2024 and based in Houston, develops control software, perception systems, and digital twins for humanoid robots intended for industrial work. The company's active project list centers on locomotion control, dexterous manipulation algorithms, teleoperation systems, and integration of manipulation with whole-body control. Manufacturing safety and the interplay between perception, manipulation, and control are documented challenges. The team is 11–50 people, primarily engineers, with active hiring in the United States.
MuJoCo, Isaac Sim, MoveIt, and OpenRAVE for simulation and middleware. Also ROS for robot operating system integration. PyTorch, JAX, and TensorFlow for perception and learning tasks.
Humanoid locomotion control, dexterous manipulation algorithms, perception systems, teleoperation interfaces, and digital twin development for manipulation tasks. Integration of manipulation with whole-body control is a current focus area.
Other companies in the same industry, closest in size