Data Scientist

Cornucopia IT Resourcing; Greater London; Hybrid; GBP 70,000 - 90,000; Full time

Job summary

Cornucopia IT Resourcing is looking for a Data Scientist specialized in Applied AI to join a fast-growing AI startup in London. The role involves designing LLM prompt pipelines and working with complex claims while supporting clients and translating their feedback into product improvements.

The ideal candidate should have strong hands-on experience in the insurance sector, proficiency in Python, Git, and SQL, and be familiar with cloud environments such as Azure.

Qualifications

• Hands-on Data Scientist with strong insurance experience.
• Familiarity with cloud environments and SQL databases.
• Experience with LLMs or prompt-driven systems.

Responsibilities

• Design and iterate LLM prompt pipelines for claims.
• Work directly with unstructured claim documents.
• Build evaluation frameworks for AI outputs.
• Write Python for data pipelines and orchestration.
• Support client pilots and translate feedback into improvements.

Skills

Insurance and startup experience

Python programming

SQL (Postgres)

Cloud environments (Azure)

Experience with LLMs

Tools

Git

GitHub

AI coding tools (Copilot, Cursor)

Job description

Data Scientist – Applied AI (Commercial Insurance)

London / Hybrid

Our client is a fast-growing AI startup building next-generation decision intelligence tools for the commercial insurance market.

Their flagship platform analyses complex claims to identify leakage, litigation risk, fraud indicators, coverage issues and operational weaknesses — turning unstructured documents into structured, actionable insight. They operate at the cutting edge of applied GenAI, working directly with insurers, managing agents and TPAs in the Lloyd’s and US markets.

This is a hands‑on, high-impact role in a fast, iterative startup environment.

What You’ll Be Doing

• Designing and iterating LLM prompt pipelines (claims summarisation, audit, fraud/leakage detection, coverage check
• s)Working directly with long, unstructured claim documents (PDFs, reports, adjuster note
• s)Building evaluation frameworks to validate AI outputs against real eviden
• ceIdentifying hallucinations, bias and gaps in reasoni
• ngWriting Python for data pipelines, orchestration and evaluati
• onQuerying Postgres, working in Azure environments, using GitH
• ubSupporting client pilots and translating feedback into product improvemen
• tsUsing AI coding tools (Copilot, Cursor, LLMs) to accelerate developme
• nt

What We’re Looking

• Strong insurance and startup/greenfield experience as a hands on Data Scient
• istComfortable with Git, SQL (Postgres) and cloud environments (Azure a pl
• us)Experience working with LLMs or prompt-driven systems in product