Machine Learning Resident

“If you are interested in the application of Machine Learning for object manipulation with robots, this is the right opportunity for you. Be a part of the team of research and machine learning scientists building AI algorithms from the ground up and get mentored by some of the best minds in AI during the process.” Abdul Wahab, Machine Learning Scientist, Amii

DescriptionAbout The Role

This is a paid residency that will be undertaken over a 12-month period with the potential to be hired by our client, Sarcomere Dynamics, afterwards (note: at the discretion of the client). The Resident will report to an Amii Scientist and regularly consult with the client team to share insights and engage in knowledge transfer activities. Successful candidates will be members of a cross-functional project team with backgrounds in ML research, project management, software engineering, and new product development. This is a rare opportunity to be mentored by world-class scientists and to develop something truly impactful.

About The Client

Sarcomere Dynamics is a Canadian physical AI company at the forefront of revolutionizing automation. By combining high-dexterity robotics with AI, Sarcomere is addressing global labor shortages and increasing safety across industries. Sarcomere’s mission is to develop synthetic labor solutions that replicate human dexterity and adaptability, making complex automation accessible and efficient.

About The Project

Sarcomere is partnering with Magna International to bring physical AI onto the manufacturing floor. Magna is evaluating which of their work cells and manual tasks can be automated across their facilities, and Sarcomere is leading the robotics and technical side of that effort, from assessing candidate tasks to building and deploying the systems that perform them. The dexterous robotic hands give Magna reach into work that conventional grippers and fixed automation can’t handle. This role will sit at the center of making that work, developing the learning-based control and perception stack that turns a real production task into a robot that can do it reliably.

Required Skills / Expertise

Are you passionate about building great solutions? You’ll be presented with opportunities to both personally and professionally develop as you build your career. We’re looking for a talented and enthusiastic individual with a solid background in machine learning, specifically in robotics for object manipulation.

Key Responsibilities

• Develop, evaluate and improve policies for robot manipulation tasks, with a focus on improving task reliability and cycle time in industrial assembly settings.
• Build sim-to-real learning pipelines that augment real-world robot demonstrations with synthetic data generated in simulation.
• Design and implement domain randomization and data augmentation strategies across visual, physical, sensor, object, and environmental parameters to improve policy robustness and transfer to physical hardware.
• Analyze the distribution gap between simulated and real-world demonstrations and develop methods for distribution alignment, dataset balancing, filtering, or representation alignment where appropriate.
• Train, fine-tune, and benchmark robot manipulation policies using real-only, simulation-only, and blended real/synthetic datasets.
• Work with imitation learning, behavioural cloning, reinforcement learning, and/or vision-language-action approaches as appropriate to the problem and available data.
• Work with multimodal robotics data including vision, tactile signals, proprioception, robot joint states, actions, and teleoperation trajectories.
• Collaborate with robotics and controls engineers to ensure learned policies integrate effectively with the broader robot stack.
• Deploy and evaluate trained policies on physical robotic hardware in a controlled setting, iterating between simulation, offline evaluation, and real-world testing.
• Develop reproducible training and evaluation pipelines, including experiment tracking, dataset versioning, model checkpoints, benchmark definitions, and quantitative reporting.
• Communicate research findings, experimental results, limitations, and recommendations to both technical collaborators and project stakeholders.
• Stay current with emerging approaches in robot learning, dexterous manipulation, imitation learning, synthetic data generation, foundation models for robotics, and sim-to-real transfer, and assess their relevance to the project.
• Engage in regular client meetings