AI Data Scientist - LLMs for Insurance Underwriting

CFC; Greater London; On-site; GBP 65,000 - 90,000; Full time

Job summary

CFC is building a production-grade Data & AI stack in London to power underwriting decisions with AI. As a mid-level Data Scientist, you will design experiments, run evaluations, and collaborate with ML and software engineers to deliver live, reliable AI solutions.

You will work across data, models, and infrastructure, owning the data science component of end-to-end production systems and validating outcomes in a fast-paced environment.

Qualifications

• Experience in Data Science, Applied Machine Learning, NLP or LLM roles.
• Strong Python and SQL with production code experience.
• Hands-on experience evaluating, deploying and monitoring ML/LLM applications.
• Solid understanding of experimentation, model evaluation, and A/B testing.
• Knowledge of cloud-based AI platforms, ideally Azure.
• Understanding how AI/ML systems are operationalised, monitored and maintained in production.

Responsibilities

• Own the data science side of production LLM systems end-to-end, partnering with stakeholders to build prototypes.
• Design evaluation frameworks and measure agent quality for live systems.
• Collaborate with ML engineers and software engineers to productionise experiments and findings.
• Develop and maintain analytical pipelines: prompt design, data-label calibration, bias checks, and validation.
• Translate data science insights into concrete engineering requirements for production services.

Skills

Python

SQL

Experimentation

NLP/LLM

Production ML

Stakeholder communication

Cross-functional teamwork

Azure

Tools

Databricks

Docker

Terraform

Azure Service Bus

Job description

CFC is building a production-grade Data & AI stack in London to power underwriting decisions with AI. As a mid-level Data Scientist, you will design experiments, run evaluations, and collaborate with ML and software engineers to deliver live, reliable AI solutions.

You will work across data, models, and infrastructure, owning the data science component of end-to-end production systems and validating outcomes in a fast-paced environment.