AI Business Analyst

AI Business Analyst (Applied AI) — 6-month contract

Hybrid working – 3 days per week in London

Reports to: Transformation Director | Team: AI COE

Role

Hands-on AI Business Analyst translating business intent into build-ready requirements for applied AI (GenAI/LLMs, ML, automation). You’ll partner with the AI Product Lead, Data and Engineering to shape use cases, define functional + data requirements, and ensure delivery is feasible, data-ready and measurable.

This is not a workshop-only BA role — you should be able to explain how the AI solution works, contribute to delivery decisions, and help diagnose issues when outcomes miss expectations.

What you’ll do

• Convert AI opportunities into clear problem statements, scope, assumptions and success measures.
• Produce build-level requirements: user stories, acceptance criteria, edge cases and test scenarios.
• Define field-level data needs (datasets, key fields, metadata) and assess availability, quality and readiness.
• Work with Engineering on feasibility, integrations, data flows and platform dependencies (build vs buy).
• Support evaluation with the AI Quality/Evaluation lead: define “good”, track performance and iterate.
• Support backlog refinement and agile ceremonies to keep delivery aligned to priorities.

What we need

• Proven BA experience in product/data/AI delivery (not just coordination).
• Strong stakeholder skills: facilitation, influence without authority, clear comms across technical/non-technical.
• High judgement in ambiguity: structures problems fast and knows what not to specify.
• Strong requirements craft: crisp stories/AC, edge cases, and testability mindset.
• Practical understanding of how data + systems/APIs (and, where relevant, prompts/retrieval/models) connect in real workflows.
• Comfortable discussing applied AI patterns used in delivery (e.g., LLMs/RAG/embeddings) at a business and delivery level.
• Evidence of hands-on problem solving when outcomes underperform (root-cause thinking, not just escalation).

Nice to have

• Familiarity with enterprise knowledge/search patterns (permissions, source-of-truth, content quality).
• Experience supporting AI testing/evaluation (expected behaviour, UAT, failure modes, iteration).
• Change/adoption experience (enablement, onboarding, feedback loops to drive sustained usage).