How The AA Cuts Routine Query Time 70% with Databricks AI/BI Genie in Microsoft Teams

The AA is the United Kingdom’s largest motoring organisation, providing roadside breakdown cover, insurance, and financial services to 14 million members. Its trading, marketing, and product teams make hundreds of data-dependent decisions daily — from tracking sales conversion rates to evaluating the impact of promotional campaigns. But data access was bottlenecked through a small team of specialists. Questions that should have taken seconds routinely took hours or days.

Analytics infrastructure at The AA was fragmented across multiple reporting tools, creating challenges in accuracy, timeliness, and governance. Data specialists spent most of their time fielding routine requests — weekly sales summaries, conversion breakdowns by device type, campaign attribution — rather than developing strategic analysis or building predictive models. Non-technical users had no viable self-service path: dashboards were complex, SQL expertise was required for anything custom, and integrating external data sources like Google Cloud’s BigQuery demanded substantial manual effort.

The AA implemented Databricks AI/BI Genie via its Conversation API, integrated directly into Microsoft Teams. Before launch, the team trained Genie on a curated set of “golden questions” — the queries trading teams asked most often — and validated every response against existing Power BI dashboards to ensure accuracy. Unity Catalog’s role-based access controls ensured users could only access the data they were authorized to see. Lakehouse Federation enabled Genie to query external sources including BigQuery through Unity Catalog, without requiring data movement or duplication. Critically, embedding Genie in Teams — not a new tool — eliminated the onboarding friction that typically slows adoption of analytics platforms.

The business impact was immediate. Routine query resolution time dropped by up to 70%. Trading teams can ask “what were Black Friday conversion rates by device type?” or “how did last week’s promotion affect sales?” and receive accurate answers within seconds, directly in Teams. Data specialists, freed from repetitive requests, shifted to higher-value work: predictive modeling, demand forecasting, and deeper strategic analysis. The head of data products called it a game changer — getting colleagues data in seconds instead of hours.

The AA is now expanding Genie across additional product lines and business units, replicating the initial rollout across new datasets. The team is also exploring Genie integrations with ML models for sales and demand forecasting — a step that would extend natural language data access from reporting into predictive decision support across the organization.