ML Engineer

What to expect

This role is for a machine learning engineer who wants to work on the models that give LeLamp its intelligence – the systems that let the robot perceive its environment, understand what's happening around it, and respond in ways that feel natural and intentional. You'll be developing and advancing the multimodal model architectures at the core of our robot, spanning vision, language, and action, and owning that work from early research through to deployment on real hardware. This is not a role where you train models in isolation and hand them off.

You'll see your work running in the robot, shaping how it behaves in the world. You'll work closely with the research and mechanical team at the frontier of embodied AI, where the problems are genuinely open and the decisions you make directly shape what the robot can do.

In this role, you will:

• Develop software systems for robotic platforms that power robot perception, intelligence, and behavior
• Write maintainable, production quality code for research and deployment
• Build multi-modal machine learning models for tasks such as computer vision, audio processing, and interaction understanding
• Design and implement pipelines for training and deploying machine learning models
• Integrate ML systems with robotic hardware and embedded systems
• Improve robot perception, responsiveness, and behavioral intelligence
• Collaborate with robotics engineers to build integrated robotic controls and architectures

You may be a good fit, if you:

• 3+ years (or equivalent) working on machine learning for robotics, controls, or perception
• Have strong programming skills in Python and/or C++ and understanding of algorithms and software engineering fundamentals.
• Have experience with machine learning frameworks such as PyTorch
• Have experience in one or more of the following areas: computer vision, robotics, reinforcement learning, or multimodal AI
• Have experience building and deploying machine learning systems preferably in Simulations (Issac Sim, MJLab, Mujoco, etc.)
• Show intentionality in design decisons and feel ownership over outcomes, even for systems you don't own.
• Think holistically about the entire system and approach complex problems with creative, outside-the-box solutions.

You will be a strong fit, if you:

• Absolutely love and are passionate about building intelligent physical systems.
• Move quickly and can do rapid iteration cycles. Share work early and iterate.
• Take ownership over the systems they design and build, and are curious, resourceful, and motivated to solve difficult problems.
• Work well in small, collaborative teams.
• Consider not just what immediately works, but how the consumer will engage with the robot.