Location: hybrid/remote Reports to: CTO (permanent role) - No Visa Sponsorship The company An early-stage legal tech SaaS business. Its platform automates document-heavy legal workflows for law firms and institutional clients. It has product-market fit, a growing customer base and a lean engineering team. The role You will lead the development of AI features in the core product. This is a hands-on role. - Design and build ML models for document parsing, search, classification and summarisation - Integrate LLMs into product workflows with guardrails, evaluation and monitoring - Optimise model serving and inference for latency and cost on AWS (ECS, Lambda, S3) - Build data pipelines for training and evaluation (PostgreSQL, Redis, OpenSearch) - Work with engineering, product and QA in Scrum squads, with CI/CD and infrastructure as code already in place Requirements - 5+ years as an AI/ML engineer delivering NLP and ML models into production SaaS - Production LLM experience: RAG, fine-tuning, evaluation, monitoring - Strong Python, including building APIs and integrating models into applications - NLP frameworks such as Hugging Face Transformers, PyTorch, spaCy or TensorFlow - AWS and distributed systems - Data pipelines and orchestration for model training and evaluation Nice to have - Legal tech or another document-heavy domain - Vector databases and semantic search (OpenSearch, Pinecone, FAISS) - Explainability and compliance work on deployed AI - Knows React or TypeScript Right to work No visa sponsorship is available. Applicants must already hold the full right to work in the UK. Please get in touch if you think you would be suitable for this role