Robotics Research Intern (AI)

Location: London, UK; on-site in the London office

Compensation: Competitive pay and perks

Company

hirify.global develops commercially scalable hirify.global robots and software systems for real-world industrial applications, including the HMND-01 platform.

What you will do

• Train language-vision conditioned manipulation policies with reinforcement learning in simulation and real-world environments.
• Build manipulation task suites and reinforcement learning models using Isaac Sim and MuJoCo.
• Develop action-conditioned video prediction and dynamics models for long-horizon physical consistency.
• Use world models to evaluate policies offline, generate synthetic rollouts, and measure model fidelity.
• Post-train VLA models for production use cases and work with multimodal data, memory, and in-context learning.
• Profile, quantise, and optimise models for real-time inference on robot edge hardware and distributed GPU infrastructure.

Requirements

• Currently pursuing or holding a master’s degree or PhD in computer science, machine learning, robotics, or a related field.
• Strong machine learning foundations, strong Python skills, and hands-on experience with PyTorch or JAX.
• Interest in reinforcement learning, world models and generative video, VLA or multimodal models, or ML systems and inference optimisation.
• Experience running experiments and interpreting results rigorously.
• Ability to take ownership, iterate with guidance, solve problems, and learn quickly in a research-driven environment.
• Full-time, five-days-per-week on-site internship in the London office.

Nice to have

• Experience with simulation-to-real transfer for robotic policies.
• Experience with egocentric data, policy post-training, data retargeting, or world modelling.
• Experience optimising policies for faster inference or working with limited compute.

Culture & Benefits

• Work on real robotic systems and contribute from an early stage with guidance from experienced researchers and engineers.
• Collaborate with engineers, researchers, product experts, and founding leadership.
• Free daily breakfast, catered lunch, and snacks in the office.
• Internship duration: 12 to 24 weeks, with a flexible start date.

Hiring process

• Complete the intern challenge using personally collected data to drive a robotic manipulator in a simulation environment.
• Submit the solution as a public GitHub repository by 9 October 2026 at 23:59 BST, including run instructions, example outputs, and design notes.