AI Robotics Research Engineer

present of AI; San Francisco (CA); On-site; USD 140,000 - 230,000; Full time

Benefits offered by this job

Health insurance

Stock options

401(k)

Job summary

Nimble is seeking a scientist/engineer to build and train the AI models that let warehouse robots see, understand instructions and act. You’ll work on diffusion policies, vision-language models and reinforcement learning for groups of robots, including collecting and labeling data and moving finished models into live systems.

A strong candidate has a graduate degree in robotics or computer science, hands-on time with real robot hardware, and a record of published research or shipped real-world

Qualifications

• Graduate degree in robotics or computer science.
• Hands-on time with real robot hardware.
• Record of published research or shipped real-world work.

Responsibilities

• Build and train AI models for warehouse robots.
• Develop diffusion policies and vision-language models.
• Collect and label data; move models to live systems.

Skills

Robotics

Machine Learning

Reinforcement Learning

Vision-Language Models

Data Labeling

Simulation

Education

Graduate degree in robotics or CS

Tools

ROS

Gazebo

PyTorch

TensorFlow

Job description

This role is about building and training the AI models that let warehouse robots see, understand instructions and act, including diffusion policies, vision-language models and reinforcement learning for groups of robots. It also covers collecting and labeling data and helping move finished models into live systems. A strong candidate has a graduate degree in robotics or computer science, hands-on time with real robot hardware, and a record of published research or shipped real-world work.

How it fits the AI era

The core of this job is AI itself, so the work shifts from hand-coding robot behavior to training models on data and testing them in simulation and on real machines. Judgment stays human: deciding what data to collect, spotting why a model fails on a physical robot, and weighing what is safe and reliable enough to put into a working warehouse.

How it builds AI experience

Someone here would get direct experience with robotic foundation models and reinforcement learning applied to autonomous fulfillment centers, from data pipelines through to production. To build toward it, train manipulation policies in simulation, try them on a real arm if possible, and build a small custom dataset for a warehouse-style task like picking and moving packages.

Worth showing

• Real hardware experience
• Custom dataset collection
• Published papers or deployed systems

This is our own read, written from the employer's posting and the company's Spotlight. The employer's posting is the source of truth for the role.

About Nimble

Nimble provides fully autonomous fulfillment centers powered by AI robotics, per the company.