Research Engineer

Cubiq Recruitment; Greater London; On-site; GBP 130,000 - 230,000; Full time

Job summary

Cubiq Recruitment is assisting a London AI research company focused on biology to hire a Senior Research Engineer. You will own the systems powering frontier research: distributed training pipelines, GPU optimisation, and experiment frameworks, enabling rapid progress from idea to result.

Reporting to the Head of AI Research, you will design custom architectures on new data types and ensure tooling supports high-speed experimentation at scale.

Qualifications

• Industry experience at an AI lab where research engineers are key contributors.
• Experience with distributed training frameworks and GPU performance work at the CUDA/Triton/XLA level.
• Contributions to open-source ML frameworks or research codebases are valued.

Responsibilities

• Own distributed training pipelines and experiment frameworks for frontier-scale research.
• Develop and maintain tooling for training runs, ensuring robustness and scalability.
• Work within the research team and report to Head of AI Research to shape feasible experiments.

Skills

Distributed training

Megatron-LM

DeepSpeed

CUDA

Triton

XLA

Open-source contributions

Software engineering

Biology curiosity

Industry experience

Tools

GPU optimization

Job description

Senior Research Engineer | AI for Biology

Relocation support provided, including Visa Sponsorship for candidates working at a frontier lab offices based out of the UK.

Build the training systems behind foundation models trained on data no other lab has.

I'm working with a London AI research company that trains large foundation models for the life sciences. They own a biological dataset that is bigger and more diverse than anything public. They collected it themselves from environments most datasets never touch. Big Tech labs cannot copy their models by throwing compute at public data, because the data itself is the moat that took years to secure. This is truly a herculean task.

Compute is not a bottleneck. The research team runs like a frontier lab rather than a typical biotech and we're looking to add multiple hires.

The role

You will own the systems that power their research: distributed training pipelines, GPU optimisation, experiment frameworks, and the tooling that lets a team work at frontier scale. You sit inside the research team and report to the Head of AI Research. Your decisions shape what experiments are possible and how fast ideas move from whiteboard to result. Custom architectures on new data types. Training runs that push hardware limits. A pace of experimentation that demands robust, flexible tooling.

What they want

• Industry experience at an AI lab where research engineers are key contributors. They have been clear that this search favours people from industry over purely academic backgrounds.
• Full-stack depth: distributed training frameworks such as Megatron-LM, DeepSpeed or FSDP, GPU performance work, and experience down to the CUDA, Triton or XLA level.
• Contributions to open-source ML frameworks or research codebases. They genuinely value this.
• Strong software engineering habits. Clean code, good tests, a focus on performance, and systems other people can build on.
• Real curiosity about biology. You do not need bio experience, but the prospect should excite you.

Pay and process

£130k to £230k base depending on experience, plus meaningful equity. They will stretch above the top of the band for an exceptional candidate from a frontier lab. The interview process is rigorous so before the technicals candidates will have a more informal chat with the Head of Research before fully commiting. Expect a technical round on ML fundamentals, hands-on coding, and end-to-end systems design.