Manipulation Co-Op/Internship (Summer 2026)

Tuesday Lab • New York City, NY

11d

About The Position

Tuesday Tuesday Lab builds delightful robots. We are driven by a desire to make everyday life easier and more fun through cutting-edge robotics. If you love watching videos of robots helping out around the house, deploy new policies onto your lab’s robot just to see how they do, and take inspiration from sci-fi rather than real life, you should come work at the Lab. We are engineers, researchers, and builders who love living among our creations. Our mission is to make the magical world of animated objects accessible to everyone. Graduate Student Manipulation Co-Op/Internship We are seeking a talented graduate student for our Manipulation Co-Op Program to help us bring our robots to life. Over the 15-20 weeks you’ll spend with us, you will be responsible for fine-tuning language-conditioned robot policies, including Vision-Language-Action (VLA) manipulation models, and deploying them onto our physical robot hardware. This is a high-impact, hands-on role for someone who loves seeing their models actually work on a physical system, not just in simulation. You will be critical in bridging the gap between large-scale AI and real-world, helpful, and delightful robot behavior. If you have ever fine-tuned an open-source model to make a robot arm pick up a coffee mug just because you could, you are exactly who we're looking for.

Requirements

• Academic status: currently enrolled in a graduate-level (MS or PhD) computer science or robotics degree program.
• ML Skillset: Proven experience fine-tuning large-scale models (e.g., VLMs, VLAs, or large transformers) using PyTorch, JAX, or a similar framework.
• Robotics Skillset: Hands-on experience deploying models onto real-world robots .
• Familiarity with robotics middleware (like ROS/ROS2) and the challenges of working with physical hardware (sensors, actuators).
• A strong understanding of the sim2real gap and experience with techniques to bridge it.
• A portfolio of projects (e.g., from GitHub, a personal blog, or academic papers) demonstrating hands-on robotics and ML integration.

Nice To Haves

• Experience specifically with imitation learning , diffusion policies, or other modern manipulation-focused models.
• Familiarity with on-device optimization (e.g., TensorRT, quantization, pruning).
• Experience with data collection at scale, especially using teleoperation.

Responsibilities

• Assist with data collection , writing software and helping to operate our robot as it performs various delightful tasks around the office
• Fine-tune language-conditioned robot policy models (such as VLAs and diffusion policies) on our custom robotics datasets to achieve specific manipulation behaviors, working closely with our ML engineers.
• Test and validate model performance on the physical robot, troubleshooting issues that span from the model's "understanding" to its physical execution.
• Collaborate with our hardware and software teams to integrate the ML-driven behaviors into the robot's autonomy stack.

Benefits

• The opportunity to play a critical role in developing the future of home robotics.
• You will be building an important piece of infrastructure and your design decisions will make the difference between a delightful companion and a dysfunctional boondoggle.
• Work alongside people with deep expertise who love playful applications and are excited about building the future.
• Freedom to rapidly iterate, contribute to the technical direction of the company, and contribute novel and impactful solutions to important problems in robotics.

What This Job Offers

• Career Level: Intern