Machine Learning Engineer (Robotics, Control Policies) - up to $10,000 + Bonus

TYSON JAY MANAGEMENT PTE. LTD.; Singapore; On-site; SGD 80,000 - 130,000; Full time

Job summary

Tyson Jay Management Pte Ltd in Singapore seeks a hands-on researcher to design and train reinforcement learning and imitation learning policies for movement and control tasks on real hardware. You will run experiments with physical robots and work to close the sim-to-real gap through debugging and domain adaptation.

The role involves building robust simulation environments and data pipelines to accelerate policy iteration, inspecting deployments for failure modes, and collaborating closely with

Qualifications

• 2–3 years of relevant experience; exceptional recent graduates with a genuinely strong portfolio and internship background will also be considered.
• Strong foundations in reinforcement learning or imitation learning, with hands-on experience training policies that run on real physical systems (not simulation only).
• Comfortable working directly with robots and hardware, not just simulators.
• Proficient in Python, with familiarity across standard RL/ML frameworks such as JAX, PyTorch, IsaacGym/IsaacLab, or MuJoCo.
• An empirical, debugging-first mindset - you care about what actually works on hardware.
• Able to move fast and switch between research problems and engineering tasks.

Responsibilities

• Design and train reinforcement learning and imitation learning policies for movement and control tasks
• Run experiments on physical hardware and close the sim-to-real gap through systematic debugging and domain adaptation
• Build and maintain simulation environments and data pipelines that support fast policy iteration
• Instrument deployments and analyse failure modes, feeding what you learn back into training
• Work closely with hardware and firmware engineers to understand physical constraints and improve policy robustness

Skills

Reinforcement learning

Imitation learning

Python

Hardware experience

Policy training on real hardware

Tools

JAX

PyTorch

MuJoCo

IsaacGym

Job description

Responsibilities

• Design and train reinforcement learning and imitation learning policies for movement and control tasks
• Run experiments on physical hardware and close the sim-to-real gap through systematic debugging and domain adaptation
• Build and maintain simulation environments and data pipelines that support fast policy iteration
• Instrument deployments and analyse failure modes, feeding what you learn back into training
• Work closely with hardware and firmware engineers to understand physical constraints and improve policy robustness

Requirements

• Around 2 to 3 years of relevant experience; exceptional recent graduates with a genuinely strong portfolio and internship background will also be considered
• Strong foundations in reinforcement learning or imitation learning, with hands-on experience training policies that run on real physical systems (not simulation only)
• Comfortable working directly with robots and hardware, not just simulators
• Proficient in Python, with familiarity across standard RL/ML frameworks such as JAX, PyTorch, IsaacGym/IsaacLab, or MuJoCo
• An empirical, debugging-first mindset - you care about what actually works on hardware
• Able to move fast and switch between research problems and engineering tasks

Tyson Jay Management Pte Ltd | EA License No.: 24C2479 Ivan Lim | EA Personnel No.: R1109856