What are the responsibilities and job description for the Research Engineer, Reinforcement Learning position at GRAM?
The Mission
At GRAM, we believe the next great leap for humanity will be physical, not digital. What AGI is to bits, Self-Replication (SR) is to atoms. As an SR research and deployment company, our mission is to make humanity galactic by enabling physical exponential growth. To achieve this, we are conducting foundational research across general-purpose robotics, artificial intelligence, biology, and materials science to enable SR within our lifetimes.
Our first step is to challenge the consensus that robots should look like us. We believe true progress comes from optimizing for function over form. We are creating a new embodiment for intelligence, designed from first principles to build the future physical economy, on and off-world.
The Role
A machine's body is only as capable as the mind that drives it. This role is for the engineers who breathe life into the hardware, teaching our Insectoid robot how to move, perceive, and act in the physical world. You will be responsible for creating the algorithms and software that form the robot's intelligence, from low-level motor skills to high-level reasoning.
This is an opportunity to solve foundational problems in AI and robotics on a novel morphology that has no precedent. You will work at the frontier of control theory, reinforcement learning, and computer vision to give our robot the ability to navigate and interact with complex, unstructured environments, creating the intelligent foundation for its ultimate mission.
Key Responsibilities:
You are a first-principles thinker who is energized by solving problems that have no playbook. You are drawn to ambitious, long-term missions and want to build technology that fundamentally changes our physical capabilities. You thrive in a high-agency environment where you are given the ownership and resources to tackle core challenges.
Basic Qualifications:
This is an on-site role at our research lab in El Segundo, California. We offer significant relocation assistance for exceptional candidates.
Interview Process
After submitting your application, we review your portfolio and any exceptional work you've shipped. If your application demonstrates the caliber we seek, you'll enter our interview process, which is designed for speed and substance. We aim to complete it within one week from start to finish.
Compensation
Your total compensation reflects both the significance of early-stage equity and competitive market rates. The following is the general framework for all roles at GRAM.
Why Join?
We are an early, focused team of scientists and engineers building the system that builds itself. GRAM is funded by leading investors and long-term visionaries. We operate without bureaucracy. Ownership and leadership flow to those who demonstrate exceptional capability, and every team member works directly on our core technology. We are based in El Segundo, the heart of America's industrial future, focused on solving one of the most important problems of our time.
At GRAM, we believe the next great leap for humanity will be physical, not digital. What AGI is to bits, Self-Replication (SR) is to atoms. As an SR research and deployment company, our mission is to make humanity galactic by enabling physical exponential growth. To achieve this, we are conducting foundational research across general-purpose robotics, artificial intelligence, biology, and materials science to enable SR within our lifetimes.
Our first step is to challenge the consensus that robots should look like us. We believe true progress comes from optimizing for function over form. We are creating a new embodiment for intelligence, designed from first principles to build the future physical economy, on and off-world.
The Role
A machine's body is only as capable as the mind that drives it. This role is for the engineers who breathe life into the hardware, teaching our Insectoid robot how to move, perceive, and act in the physical world. You will be responsible for creating the algorithms and software that form the robot's intelligence, from low-level motor skills to high-level reasoning.
This is an opportunity to solve foundational problems in AI and robotics on a novel morphology that has no precedent. You will work at the frontier of control theory, reinforcement learning, and computer vision to give our robot the ability to navigate and interact with complex, unstructured environments, creating the intelligent foundation for its ultimate mission.
Key Responsibilities:
- Develop and apply state-of-the-art reinforcement learning (RL) algorithms to solve GRAM's hardest robotics challenges, such as dynamic locomotion and contact-rich manipulation.
- Design reward functions and training curricula that enable the learning of complex, long-horizon behaviors.
- Build scalable, data-efficient RL training pipelines that leverage both simulation and real-world data.
- Research and implement novel solutions for sim-to-real transfer to ensure policies trained in simulation work robustly on our physical robots.
- Collaborate with the entire robotics team to integrate learned policies into the robot's autonomy stack.
You are a first-principles thinker who is energized by solving problems that have no playbook. You are drawn to ambitious, long-term missions and want to build technology that fundamentally changes our physical capabilities. You thrive in a high-agency environment where you are given the ownership and resources to tackle core challenges.
Basic Qualifications:
- MS in Computer Science, Robotics, or a related field with 3 years of hands-on experience applying RL to complex problems.
- Expert proficiency with Python and deep learning frameworks (e.g., PyTorch, TensorFlow).
- A strong theoretical understanding of modern RL algorithms (e.g., PPO, SAC) and their trade-offs.
- Experience building and managing large-scale model training pipelines.
- A PhD with a focus on reinforcement learning, preferably for robotics.
- A track record of publications in top-tier AI or robotics conferences (e.g., NeurIPS, ICML, ICLR, CoRL, RSS).
- Proven experience solving sim-to-real challenges for robotic systems.
- Expertise in imitation learning, offline RL, or model-based RL.
This is an on-site role at our research lab in El Segundo, California. We offer significant relocation assistance for exceptional candidates.
Interview Process
After submitting your application, we review your portfolio and any exceptional work you've shipped. If your application demonstrates the caliber we seek, you'll enter our interview process, which is designed for speed and substance. We aim to complete it within one week from start to finish.
Compensation
Your total compensation reflects both the significance of early-stage equity and competitive market rates. The following is the general framework for all roles at GRAM.
- Company-Wide Base Salary Range: $75,000 - $300,000 USD, calibrated to your impact potential
- Equity: Substantial ownership stake befitting founding team members
- Benefits: Health, dental, and vision coverage; all meals are paid for; relocation assistance
Why Join?
We are an early, focused team of scientists and engineers building the system that builds itself. GRAM is funded by leading investors and long-term visionaries. We operate without bureaucracy. Ownership and leadership flow to those who demonstrate exceptional capability, and every team member works directly on our core technology. We are based in El Segundo, the heart of America's industrial future, focused on solving one of the most important problems of our time.
Salary : $75,000 - $300,000