Machine Learning Engineer - Fully Remote in London

Generative Engineering is bringing AI design into the real world by enabling generative engineering design for physical products. Our focus is creating millions more engineers globally and giving them the data and knowledge necessary to make efficient decisions quickly, one of the main challenges of the physical engineering industry today. Our team has a background in scaling software to millions of users and successfully disrupting industries, creating Unicorns and Decacorns along the way. We combine the advantages of an early-stage start-up with the ability to focus on creating high-quality, high-impact systems, without the distraction of fundraising.

We are looking for a Machine Learning Engineer to join the team — someone who can operate across the full spectrum from research to production. This role sits closer to the research end: you'll be pushing the frontier on generative models for physical design while also shipping real systems that engineers use every day. Please show both the quality of your past research and any production impact it has had.

• PhD in Machine Learning, Computer Science, Applied Mathematics, or a closely related field, with original contributions to deep learning, reinforcement learning, or generative models.
• Formal background in generative modelling — working knowledge of the transformer architecture, diffusion models, flow matching, and variational autoencoders: their evolution, their tradeoffs, and where they're going.
• Real world experience building ML/AI systems that reached production, not just research prototypes.
• Practical experience managing research projects end to end — from problem formulation through to evaluation and deployment.
• Knowledge of modern, larger-scale Python and the ML stack (PyTorch, JAX, or equivalent). You write research-grade code.
• Practical experience building large-scale data pipelines. We don't have data infrastructure — you'll help build it.

• Experience in a high-pace startup environment.
• Knowledgeable about physical engineering or related domains such as robotics or cognitive science.
• Experience working with PINNs (physics-informed neural networks) or graph neural networks for physics-based surrogate models.
• Experience owning or being involved longer-term in an open source project, ideally in a related field such as ML tooling or scientific computing.
• Experience with GPU cluster orchestration.
• Experience with vector embeddings, ideally retrieval-augmented generation (RAG) and multi-modal representations (e.g. CLIP).
• Experience with model fine-tuning.
• Experience with Markov chains or (partially-observable) Markov decision processes.

Just state the word 'Salmon' anywhere in your application, just to prove you can read a job advert :) We aim to improve all our colleagues' abilities and careers by exposing them to the bare bones of a tech start-up whilst giving them the opportunity to support the company in any way. If our people continuously improve, so does our product.