Data Scientist - Outside IR35

2x Data Scientist£475 - £500 outside IR35 p/d6 months initial - likely to extendOnce a week into LondonSR2 are supporting a major cross-government digital initiative focused on transforming how citizens interact with government services through AI. The team is developing an intelligent tool that understands when a user requires human assistance and seamlessly connects them to the most appropriate support route across departments.This project sits at the intersection of agentic AI, conversational systems, and government service delivery, and aims to reduce the knowledge burden on users by creating a single, intelligent entry point for government support.The RoleAs a Data Scientist, you'll play a pivotal role in designing, experimenting, and validating new AI models that enhance the performance and human-routing logic of the platform. You'll work in a multidisciplinary environment alongside developers, researchers, and policy experts, helping shape the direction of a national-scale AI service.Key Responsibilities

• Identify and define key technical research questions, and run rapid experiments to validate assumptions in the AI domain.
• Develop, evaluate, and optimise conversational AI models, including Retrieval-Augmented Generation (RAG) architectures.
• Implement and refine LLM-as-a-judge evaluation frameworks for continuous model improvement.
• Collaborate closely with cross-functional teams to translate business challenges into analytical and data-driven solutions.
• Produce clear, visually compelling analytical outputs to inform decision-making at senior levels.
• Develop Python-based analytical pipelines and prototype models in a cloud environment (AWS preferred).
• Contribute to integration efforts between AI systems and existing digital service platforms.

Skills & Experience Required

• Proven experience as a Data Scientist, with hands-on capability in Python and applied AI experimentation.
• Experience in cloud-based environments (AWS preferred).
• Practical experience with conversational AI and RAG system optimisation.
• Strong stakeholder engagement skills, able to explain complex AI/ML concepts to non-technical audiences.
• Demonstrable ability to design analytical outputs that support business and policy decisions.
• Background in collaborative, agile environments delivering data-driven value.

Desirable

• Experience working with agentic AI frameworks in an experimental or applied context.
• Familiarity with Model Context Protocol (MCP).
• Exposure to CRM systems or human support integration use cases.