About this role
ML / LLM Research Engineer, National Security
£80,000 to £90,000 + 10% bonus for DV clearance
Manchester | Hybrid / Secure Site Working | Permanent
Ncounter is supporting a growing technology consultancy looking for an experienced ML Research Engineer to join a specialist team working across AI, machine learning and secure government programmes. The immediate project is particularly interesting, combining hands-on engineering with genuine research into the security and behaviour of Large Language Models. Rather than simply building applications around existing AI tooling, you will investigate how models work internally, exploring areas such as how information is processed, where data moves, whether it could contribute to future training and whether models present risks when deployed within secure environments. This will involve taking ideas from academic research and turning them into practical experiments, primarily using Python.
What we’re looking for:
Active SC clearance as a minimum, with previous high-side experience particularly valuableStrong Python development skills, this is a hands-on engineering position and you must be comfortable codingStrong underlying knowledge of machine learning and Deep Neural Networks, including an understanding of how models operate internally Nice to have:
Research experience, with the ability to understand technical literature and turn findings into practical experimentsSpecific experience with LLMs is highly desirable, particularly research into their architecture, behaviour or securityA strong mathematical or STEM background, with MSc or PhD-level research particularly attractiveExposure to AI/ML security, model safety, explainability or similar areas would be beneficialBroader ML engineering skills, allowing you to move between research-led and more traditional engineering projects over time
The position offers more variety than a purely academic research role. Beyond the initial programme, you will contribute across the wider ML engineering lifecycle, including production-quality development, testing, documentation and code reviews.
The role is expected to require around three days per week onsite, potentially more when secure environments require it.
If you combine genuine ML research depth with strong Python engineering skills and want to apply that knowledge to complex, security-focused AI problems, contact Ncounter to discuss the opportunity.