Job Description
Do you want to work on LLM systems that ship to real users, not just papers?
Have you published at top AI venues and want to build in industry full time?
Are you ready to join a small, elite team with a genuinely high technical bar?
A fast-scaling, YC-backed LegalTech AI company is building GenAI systems across document processing, draft generation and information retrieval. Their platform is already used by hundreds of legal teams globally, and they’re now expanding their AI group as they scale in London and internationally.
This is not a generic graduate role. They’re looking for exceptional early-career research engineers who can combine strong academic foundations with practical engineering execution.
The role
You will work on applied research problems directly connected to production GenAI systems: improving retrieval, reasoning quality, grounding, evaluation, and system robustness across large, complex document corpora.
Key responsibilities
* Build and iterate on LLM + RAG pipelines over technical/legal documents
* Design evaluation methods to improve correctness, faithfulness, and reliability
* Improve retrieval quality (chunking, ranking, metadata strategies, failure analysis)
* Prototype and productionise applied research ideas into product workflows
* Collaborate closely with backend/full-stack engineers to deploy features end-to-end
What they’re looking for
* New Grad / Early Career, ideally MSc or PhD in CS/ML/NLP (or related)
* Strong applied ML research background, ideally with publications in top venues such as:
* NeurIPS, ICML, ICLR, ACL, EMNLP, AAAI (added bonus)
* Strong engineering fundamentals in Python
* Comfortable working in real codebases (not just notebooks)
* Product-minded: enjoys turning research into shipped systems
* Happy working onsite 5 days per week in Central London
Interested? Please apply below.