Skip to main content

Pickle Robot

Software Engineer V – Applied ML Research

Charlestown, MAfull-timemidAdded today

About this role

Pickle Robot seeks a Senior Software Engineer to lead applied ML research on robotic grasping and manipulation. You'll architect solutions combining classical robotics with deep learning to enable autonomous truck unloading, working on a two-person team with significant technical ownership.

What you'll do

  • Design and implement ML models for robotic grasp prediction and execution
  • Integrate neural networks with classical robotics approaches for real-world performance
  • Lead technical direction and strategy for grasp planning systems
  • Collaborate with robotics and hardware teams to validate solutions on physical robots
  • Own end-to-end development from research to production deployment
  • Mentor and guide junior team members on the project

What they're looking for

  • Machine learning (neural networks, computer vision, reinforcement learning)
  • Robotics (kinematics, grasping, manipulation, planning algorithms)
  • Python and systems programming for ML pipelines
  • Hands-on experience with real robot platforms
  • Data engineering and dataset management for ML
  • C++ or similar for performance-critical systems
  • Research-to-production software development
Apply with Autofill

Opens the application — the Jobs AI extension fills it for you. Set up autofill

Opens the official application on the employer’s site. No login required.

Pickle Robot

Pickle Robot builds warehouse automation robots and the software infrastructure to operate them at scale. The company is hiring simulation engineers, robotics software engineers specializing in ML and grasping, DevOps engineers, and backend engineers to develop cloud-based systems, robotic workflows, and human-robot interaction platforms.

View all jobs at Pickle Robot

Likely interview questions

  • Describe a project where you've successfully deployed ML models on physical robots—what were the key challenges and how did you overcome them?
  • How would you approach designing a grasp prediction system that needs to work reliably across diverse objects and warehouse conditions?