Senior VLA Pre-training & World Models Engineer - London

hirehire; Greater London; On-site; GBP 90,000 - 135,000; Full time

Benefits offered by this job

Equity

30+ days off

Private healthcare

Pension

Free meals in office

Visa sponsorship & relocation

Job summary

Humanoid in London is seeking a senior VLA Pre-training Engineer / Deep Learning Engineer to join two tracks: VLA Pre-training and World Models. On-site at King’s Cross, visa sponsorship and relocation to London offered.

The role covers pre-training on diverse trajectory data, policy training with cloning/RL, scalable data pipelines, and evaluation tools for robot deployment in real time.

Qualifications

• 3+ years building deep learning systems in industry or research.
• Hands-on experience with at least one: LLMs, VLMs, image or video generative models (architecture, training, inference).
• Trained models yourself: pre-training runs, distributed infrastructure, APIs, fine-tuning not enough.
• Experience with deep learning infrastructure: streaming datasets, checkpointing, state management, distributed training strategies.
• Strong Python and PyTorch or JAX; able to profile, optimize data loaders and write maintainable research code.
• Strong coding fundamentals; LeetCode-style coding round.
• Empirical rigour: clear baselines, honest evaluations, precise experiment docs.
• Proactive; able to tackle open, messy problems without a spec.
• Working English.

Responsibilities

• Pre-train base models on a diverse multi-embodiment corpus of trajectories and synthetic data.
• Post-train policies with behaviour cloning and RL; own data-to-deployment loop.
• Collaborate with Data Collection: specify good data, identify failure modes, ensure coverage.
• Run pre-, mid- and post-training on the VLA stack; explore new modalities and architecture changes.
• Build and maintain continuous pipelines: ingest synthetic data, version datasets, surface failures for retraining.
• Scale distributed training and optimize models for real-time edge inference on robots with MLOps/data platform teams.

Skills

Deep learning systems

LLMs / VLMs / video models

Python & PyTorch / JAX

DL infra & distributed training

English proficiency

Tools

PyTorch

JAX

Isaac Sim

Job description

Humanoid in London is seeking a senior VLA Pre-training Engineer / Deep Learning Engineer to join two tracks: VLA Pre-training and World Models. On-site at King’s Cross, visa sponsorship and relocation to London offered.

The role covers pre-training on diverse trajectory data, policy training with cloning/RL, scalable data pipelines, and evaluation tools for robot deployment in real time.