Machine Learning (Imitation, RL) and Robotics Research Engineer — 🔒 Filled · Accepting expressions of interest

The role We're looking for Machine Learning and Robotics Research Engineers who live and breathe robot learning. If you get excited about imitation learning, behavior cloning, and pushing the boundaries of embodied AI—and terms like ACT, Diffusion Policy, and VLA are already in your vocabulary—we want to hear from you! You'll need a solid grasp of general ML concepts and practices and hands-on experience with with modern deep learning algorithms, ideally along with experience with imitation learning and reinforcement learning, or the ability to pick up those concepts quickly. The core of this role involves training, adapting, and developing imitation and reinforcement learning algorithms that enable robots to learn behaviors from human demonstrations. You'll work on adapting learning pipelines to new tasks and robot configurations, as well as novel algorithm design to improve learning efficiency and sample complexity. Beyond algorithm development, you'll also contribute to building robot simulation frameworks, developing software for robot integration and human demonstration data collection, and maintaining the broader supporting codebase. About Imitation Machines Imitation Machines has an ambitious goal of bringing the latest advances in Imitation and Reinforcement Learning, Robotics and AI into an integrated platform that enables anyone in the world to create robot new and custom behaviours. Imagine the state of future technology where anyone can teach any robot, by demonstration, to solve any task they need, and create intelligent robotic workforce that meets unique needs across industries, robot forms, and environments. We want this future to happen. We see that all the necessary technological prerequisites are in place, and we're building that future. The readiness of reinforcement learning, imitation learning, and robot hardware has reached a tipping point, positioning Imitation Machines to transform how robot behaviours are created. Rather than relying on conventional programming, we can radically accelerate the process and start building systems where robots are taught by human teachers through demonstration. We're taking a practice-first approach: real-world robot deployments with our industry partners guide our research and development. By working directly with industries, robot developers, and system integrators, we position ourselves at the frontier of practical imitation learning—tackling the challenges that emerge when these technologies meet real applications. This makes us early integrators of imitation learning in production environments, and it's how we're paving the way for broader adoption of robotic automation. Imitation Machines is a pre-seed funded startup moving quickly toward our next milestone. We're laser-focused on proving our vision and building a team of machines learning and robotics engineers who want to push the boundaries of what's possible in robot learning. What we’re looking for Strong understanding and hands-on experience with deep learning algorithms and architectures, especially in reinforcement and imitation learning, ideally applied to robotics Solid grasp of machine learning concepts Experience in robot systems integration — ROS, robotics simulation environments Ability to develop ML pipelines: data collection, training, evaluation, and deployment Solid software engineering practices — version control, testing, debugging, modular design, and documentation Self-driven builder mentality – experimenting, making things happen, thriving on challenges, and breaking new ground Position details Level: Entry to mid-level Type: Full-time Location: Theoretical, algorithms, and in-simulation work can be done remotely and we do consider fully remote candidates. While based in Sydney is preferred for hardware testing and integration, we welcome strong interstate candidates. Compensation: As an early-stage startup, we can't always match big tech salaries. What we can offer is the opportunity to work on genuinely hard problems in robot learning, significant autonomy in shaping our technical approach, and the chance to build something from the ground up. Primary location: Sydney, Australia