We're looking for a reinforcement learning researcher or engineer who wants to apply deep thinking to real-world autonomy challenges.

You'll join a small, multi-disciplinary autonomy group working on high-level tactical decision-making using reinforcement learning. Your focus will be on getting policies from sim to hardware.

- Reward function design and environment shaping

- Policy development for long-horizon decision-making

- Integration with perception, SLAM, and low-level control systems

- Moving from simulated success to real-world performance

You'll be working alongside engineers and researchers from defence, robotics, and academia. There's a strong technical bar, but an equally strong bias toward iteration, delivery, and ownership.

You don't need a 10-paper publication record or a decade of industry deployment. What matters most is capability, systems thinking, and the drive to build.

- RL researchers with some industry context

- Robotics or autonomy engineers looking to deepen their ML capability

- Mid-level engineers with strong fundamentals and the hunger to learn

Experience with RLlib, Isaac Gym, PPO/SAC variants, or classical control systems is a bonus — but not a deal breaker. If you've worked on applied autonomy and know how to turn theory into decisions, you're likely in the right ballpark.

You'll need to be based in Australia or willing to relocate. Permanent residency or citizenship is required. This is an onsite role with real aircraft, not a remote lab job.