Machine Learning Engineer – Neural Network R&D

London | Hybrid | £130,000

Most ML roles use models.

This one builds them.

You’ll be working on fundamental neural network development — taking architectures off the shelf and making them faster, more efficient, or entirely rethinking how they work.

This is deep, technical work:

Improving performance of neural networks at architecture level

Developing and experimenting with novel approaches (e.g. spiking neural networks)

Pushing models closer to real-world deployment constraints

The work sits within advanced AI programmes being applied to autonomous systems and defence environments - where performance actually matters.

They’re looking for someone who:

• Has built or significantly improved neural network architectures
• Strong experience with PyTorch / TensorFlow
• Deep understanding of deep learning fundamentals (not just using libraries)
• Can demonstrate research or advanced development work

London (hybrid)

Eligible for SC/DV

If you’ve gone beyond just “using models” and into actually improving how they work — this is your kind of role.