F
Flexion Robotics Tech Stack
Autonomy stack for humanoid robots using simulation and reinforcement learning
Flexion Robotics builds software for humanoid robot control—from perception and trajectory generation to dexterous manipulation and locomotion. The stack centers on PyTorch, NeRF, Transformers, Isaac Sim, and MuJoCo, with active development across learning-based controllers, multimodal trajectory models, and sim-to-real deployment pipelines. The organization is engineering and research-heavy (12 of 14 active hires), weighted toward senior roles, signaling focus on closing the simulation-to-reality gap and stabilizing field performance rather than pure scaling.
Tech Stack 26 technologies
What Flexion Robotics Is Building
◆Challenges
- Transition from prototypes to real-world deployment
- Transition from research prototypes to robust controllers
- Scaling technical hiring engine
- Building high-caliber pipelines for niche robotics roles
- Scaling software across customer robot fleets
- Stabilizing performance in the field
- Closing sim-to-real gap
- Scaling internal it operations
- Protecting company ip
- Maintaining secure cloud environments
▲Active Projects
- Learning-based controllers for dexterous manipulation
- Photorealistic 3d environments
- Learning-based manipulation controller deployment
- Manipulation stack definition
- State-of-the-art multimodal models for robot trajectory generation
- Deployment of learning-based trajectory generation for robotic systems
- State estimation solutions for humanoid robots
- Generative whole-body trajectory generation
- 3d environment perception
- Ai system integration with humanoid hardware
Hiring Activity
15 roles · 7 in 30d
Department
Seniority
Company intelligence
