SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Berkshire Grey is a leader in AI and robotics, providing innovative solutions for e-commerce, retail replenishment, and logistics. The company automates complex pick, pack, and sort operations using advanced robotic systems.
As a Senior Software Engineer on the Core software team, you will develop learning-based approaches that help robots reason about 3D environments, anticipate outcomes of physical interactions, and make effective decisions over sequences of actions. The work combines perception, planning, and manipulation, with emphasis on reliable performance under uncertainty and real-world constraints. You'll explore both data-driven and model-based approaches, evaluate them in simulation, and bring promising methods onto physical robotic systems.
Key responsibilities include:
- Develop learning methods for spatial reasoning, sequential decision-making, and robotic manipulation
- Combine learned policies and predictive models with planning and optimization to solve physically constrained tasks
- Design training and evaluation workflows that measure task success, robustness, and execution efficiency
- Quickly prototype solutions and create demos for stakeholders
- Serve as subject matter expert for transitioning prototypes to product teams
- Identify high-impact areas for improvements to robotic systems
- Utilize and extend simulation software environments to develop and test manipulation behaviors
- Stay current with latest advancements in robotics and evaluate applicability to company challenges
- Collaborate with external research partners
- Mentor junior engineers and interns
- Communicate technical priorities and status
REQUIREMENTS:
Minimum qualifications:
- Master's degree in Robotics, Machine Learning, Computer Vision, Computer Science, or closely related field
- 4+ years of software development experience with focus on robotics manipulation or related areas
- Strong development expertise in Python and C++
- Experience with major deep learning frameworks such as PyTorch
- Experience with data science tools & libraries (numpy, pandas, scipy, matplotlib, scikit-learn)
- Experience working with ROS or ROS2
- Demonstrated experience training and adapting existing machine learning architectures for robot learning in domains such as manipulation, locomotion, or navigation
- Demonstrated ability to: develop on and troubleshoot real robotic systems, determine and communicate technical priorities, rapidly prototype and iterate on solutions, work independently while maintaining focus, work in fast-paced environments with changing priorities, mentor junior engineers
- Strong verbal and written communication skills, capable of explaining complex ideas clearly to technical and non-technical stakeholders
- Ability to provide technical leadership on key projects
Preferred qualifications:
- PhD in Robotics, Machine Learning, Computer Science, or closely related field
- Demonstrated technical proficiency in spatial reasoning and sequential decision-making in robotics
- Experience with: applying machine learning to hardware interacting with real world, real and simulated data capture, sim-to-real for robotic manipulation, real-time perception-based control, incorporating tactile sensor information into manipulation behaviors, robot simulators (OpenRAVE, Isaac Sim, MuJoCo), robot learning frameworks (Isaac-Gym, Orbit/Isaac Labs), combining model-based and data-driven approaches, Docker/cloud computing, experiment tracking and dataset management (Weights & Biases), database systems like MongoDB, parallel/distributed systems and asynchronous/concurrent programming
- Knowledge of LLM usage within robotics context
