Defence Applied ML Lead — Onsite in London, Equity

Cosine; Greater London; On-site; GBP 110,000 - 160,000; Full time

Benefits offered by this job

Equity / share options

30 days holiday + bank holidays

Pension

Dog-friendly office

Weekly team breakfast & lunch

Monthly socials

High-quality equipment

Job summary

Cosine, a UK AI lab based in London, seeks an Applied ML Technical Lead to develop and deploy advanced AI capabilities for Defence and national security. This hands-on leadership role spans ML research, deployment, and direct customer delivery with stakeholders across Defence.

Clearances and close collaboration with engineering teams are central to success. You will own technical direction, work across model development, evaluation, and deployment, and translate operational needs into concrete

Qualifications

• Writing production-quality Python and ML code.
• Experience with PyTorch and ML frameworks.
• Ability to design, train and evaluate ML models for defence/secure envirnoments.
• Understanding deployment constraints: latency, compute, on-prem / secure environments.

Responsibilities

• Lead technical direction and day-to-day progress of applied ML work.
• Coordinate ML researchers, engineers and stakeholders to deliver capabilities.
• Translate operational objectives into technical goals and experiments.
• Lead from prototype to deployment while maintaining high technical depth.
• Engage with Defence stakeholders to understand operational needs and drive delivery.

Skills

Python

PyTorch

Model deployment

ML engineering

GPU compute

Git

Linux

Containers

Tools

Git

Linux

Docker

Job description

Cosine, a UK AI lab based in London, seeks an Applied ML Technical Lead to develop and deploy advanced AI capabilities for Defence and national security. This hands-on leadership role spans ML research, deployment, and direct customer delivery with stakeholders across Defence.

Clearances and close collaboration with engineering teams are central to success. You will own technical direction, work across model development, evaluation, and deployment, and translate operational needs into concrete