Technical Lead - Robot Learning

Want to build a robot learning stack from zero, on hardware that customers already pay for?

Tired of choosing between a startup’s freedom and a business that actually makes money?

Have you trained policies that work on real robots, not just in simulation?

The opportunity

This is a confidential search for a London robotics company whose dexterous manipulation hardware is used by some of the world’s leading AI labs. They’re profitable, independent and proudly scrappy: a small team, no funding-round clock, and real customers from day one.

Now they’re building the intelligence layer behind their robots, and they want one person to own it.

This is a genuine 0-to-1 build. You’ll shape it, not inherit it.

What you’ll own

• The robot learning layer, end to end: data, telemetry, models and sim-to-real transfer
• Training learning-based policies for contact-rich manipulation and getting them working on physical hardware
• The technical roadmap for learning and AI, from first experiments to capability customers can use
• The data foundations: how robot data is captured, stored and turned into better models
• Setting technical direction alongside the leadership team, with a real voice in what gets built next

You’ll be right for this if you have

• Around 3+ years of hands-on robot learning experience, with work that ran on real robots
• Reinforcement learning or learning from demonstration applied to manipulation, ideally contact-rich tasks
• Sim-to-real experience using Isaac Lab, Isaac Sim, Isaac Gym or MuJoCo, including domain randomisation
• Evidence that you’ve set technical direction or owned a roadmap, not only executed someone else’s
• Strong Python and the engineering discipline to build systems others can rely on

Nice to have

• A PhD in robotics, robot learning or a related field (a strong portfolio counts just as much)
• Experience with teleoperation data or robot data infrastructure
• Time at a robotics company taking research into product

This is not for you if

• Your AI experience is mainly LLMs, agents or RAG without robots in the loop
• Your robot learning work has stayed in simulation
• You want a large team, heavy process or a corporate atmosphere

Why it’s worth your time

• Real hardware, real customers and real revenue, so your work ships
• A blank page to design the learning stack your way
• A small, collaborative team with no egos and no room for grandstanding
• The rare chance to do 0-to-1 work without startup survival risk

The practicals

• London-based, working hands-on with the hardware
• You’ll need the right to work in the UK already, as this role can’t offer visa sponsorship
• Not suitable for recent graduates

How to apply

Apply here and Helen at HFBAC, the company’s hiring partner, will be in touch. We work Chemistry First, so your details go nowhere without a conversation first. Send your CV and a few lines on the most interesting policy you’ve got working on real hardware.

No agencies, please.