5 months ago
London, United KingdomSenior
Responsibilities
- Design and implement gym environments for dexterous manipulation tasks across industrial, service, and home domains.
- Define observation spaces, action spaces, and reward functions, and analyze and address reward hacking.
- Reduce the sim-to-real gap by tuning physics parameters, improving asset fidelity, and validating simulations against real-world data.
- Procure, validate, and integrate high-quality 3D assets from third-party vendors.
- Configure contact dynamics, friction, mass properties, and joint limits for reliable policy transfer to hardware.
- Identify and reduce simulation bottlenecks to improve training throughput and environment step rates.
- Improve simulation-based evaluation and reinforcement learning infrastructure for rapid iteration and scaling.
Requirements
- 3+ years of experience building simulation environments or game-engine-based interactive systems, with shipped products, published results, or equivalent artifacts.
- Deep hands-on experience with at least one physics simulator such as Isaac Sim, IsaacLab, MuJoCo, PyBullet, or Drake, or equivalent experience with Unreal or Unity focused on physically accurate interactions.
- Strong practical experience running large-scale parallel simulation on GPU clusters and familiarity with GPU-accelerated simulation infrastructure.
- Strong Python skills, including profiling bottlenecks, debugging physics issues, and writing maintainable research code.
- Familiarity with modern software engineering practices.
- Ability to document experiments clearly and communicate trade-offs crisply.
- Preferred qualifications include robotics or manipulation experience, reward-function design, reinforcement learning policy training, knowledge of reward hacking and sim-to-real issues, NVIDIA Isaac Sim and IsaacLab experience, domain randomization, system identification, sim-to-real transfer techniques, relevant publications, open-source contributions, or familiarity with robocasa and robosuite.
Benefits
- Competitive equity through stock options.
- 30+ paid days off, including 23 days of annual leave, UK bank holidays, and additional company closure days.
- Private healthcare with virtual and in-person care.
- Pension scheme with an 8% total contribution, comprising 5% employee and 3% employer contributions, on full earnings.
- Free daily breakfast, catered lunch, and snacks in the office.
- Opportunity to work with engineers, researchers, and product experts on humanoid robotics and AI.
- Direct access to founding leadership and meaningful input into product direction.
Tech Stack
PythonUnity
Categories
ML EngineeringRobotics
