AI Research Scientist

This role is for you if: instead of building in an AI lab, you want to do R&D grounded in data with learnings from real world deployments. You’ll believe that iterative development and learning by doing reduces technical risk and increases chances of success. No two days will be the same, but you can expect to:

• Develop cutting edge navigation, manipulation and human-robot interaction models.
• Identify and validate adjacencies (cross task / cross sector) that your research unlocks.
• Build iteratively to generalisable robotics tasks whilst delivering real-world impact from deployed robots, today.

About Paddington Robotics: At P9R7, we’re building robots that can handle the messy, unpredictable and chaotic real world! We believe that by deploying today, we reveal second‑order effects and edge cases that can’t be seen in the lab or the sim. Our mission: become the UK’s go‑to applied AI robotics team - and we’ll go above and beyond to get there.

What we actually do:

• Build AI models and hardware that let robots work safely alongside humans – lifting, moving, and getting stuff done better and faster.
• Spend lots of time on Customer sites, to better understand and solve for the messy chaos of most environments.
• Think 10+ years ahead, reimagining manual labour and creating robots that become complementary assistants for labour‑intensive and physically demanding tasks.

We’re a small, dynamic team from Amazon, InstaDeep, Imperial, UCL, and Cambridge to name a few, united by one of the founders of Magic Pony (acquired by Twitter).

We’re hiring an AI Research Scientist to leverage the data from our real‑world deployments to build the AI and software stack required for reliable robotics. Depending on your experience and interest, you might be problem‑solving across:

• Understanding
• Manipulation
• Navigation
• Human‑robot behaviour

About you: You follow the latest in AI robotics research: data‑efficient learning of moving, grasping, seeing, sensing, all the things that people do naturally in a world created by people for people. You have developed and deployed cutting‑edge models for use in the real world: robustness against the mess of the world, sensor noise, ever‑changing dynamic environments. You have experience with what works in practice and what is more speculative. You’re highly collaborative, and love working with others: sharing knowledge, learning from and teaching others, and embracing feedback. You’re a fast learner, and low ego: and ask all the questions! Fast iteration speed means you do not get attached to specific solutions. You’re a practical problem solver: you can take your expertise and tackle engineering challenges pragmatically in messy, real‑world environments.

To thrive here: Highly ambitious: comfortable taking a project from concept to completion – you do what is necessary to achieve our goals. Self‑directed and proactive: you thrive in a dynamic, unpredictable, multi‑disciplinary environment. Enthusiastic team players: you ask questions, help others, and communicate effectively.

What will impress us: We love proactive doers. Show us what you’ve built – whether it’s interfaces, projects, or experiments. Tell us why our mission excites you. Be bold, be creative, and don’t feel embarrassed to stand out from the crowd.

Location: Our London office is right by Portobello Road. We work full‑time, and largely on‑site (4 days a week). Expect vibrant startup energy with a team that knows when to laugh, and when to buckle down.

Paddington Robotics is an equal opportunity employer and encourages individuals from all backgrounds to apply. We offer competitive compensation, the friendliest of teams, an insanely stimulating work environment, and the opportunity to be a part of something massive.