Job Description
Do you want to build AI agents that directly automate real-world claims decisions?
Have you experimented with LLMs beyond notebooks and shipped them into live products?
Are you ready to shape AI transformation inside a tech-driven insurer investing heavily in generative AI?
A leading UK insuretech (200–300 employees) is entering a major AI transformation phase following recent refinancing and dedicated AI investment. The business operates with a strong engineering mindset and is building a new AI function from the ground up. The vision is ambitious: fully agentic AI across the claims lifecycle. Early products are already live, including an internal AI productivity agent.
This AI Research Engineer role focuses on applied experimentation and fast iteration. You’ll prototype, evaluate and shape AI-first products, particularly LLM-powered agents, without being solely responsible for long-term platform ownership.
Key Responsibilities
• Rapidly prototype and benchmark LLMs, embeddings, and RAG pipelines
• Build AI agents automating claims logging through settlement
• Develop and refine evaluation frameworks (accuracy, completeness, cost, latency)
• Running on GCP (Vertex AI, Cloud Run, BigQuery)
• Collaborate with Claims Operations to demonstrate measurable business impact
• Extend internal AI tooling and agent capabilities
Key Details
• Salary: £60k-£75k
• Working model: Hybrid – 1-2 days per week in office (South West London, 20 minutes from Waterloo Station)
• Tech stack: Python, GCP (Vertex AI, Model Garden, BigQuery, Cloud Run), LLMs (OpenAI, Gemini, Claude, Llama), vector databases, RAG, Docker, CI/CD
• Sponsorship: Not currently available
Interested? Please apply below.