Skip to content
Zenta
Menu

Success story.

Organized data and conversational AI so any team can ask the business questions

A retail group with two brands and a 100% AWS platform wanted to test, in a controlled manner, how Google Cloud could make information available to all its teams using conversational AI.

Image for the Organized data and conversational AI so any team can ask the business questions case

We take on any challenge

At Zenta there are no obstacles we cannot overcome.

  • Challenge

    With massive volumes of data from two brands, accessing the information required technical expertise. The test had to function without affecting existing systems, keep each brand's metrics separate, and provide reliable results.

  • Solution

    We built a governed data platform in BigQuery, with automatic daily updates and quality controls. On top of that, we created six specialized AI agents that answer business questions in natural language, each with its own data universe. We added dashboards in Looker Studio, cost control per query, and four workshops for the team to continue developing independently.

Results from our management

  • Daily load from end to end: 40 out of 40 boards in about 16 minutes.

  • 7 out of 7 natural language query use cases validated with real data.

  • 6 AI agents with 56 verified queries and their own business glossaries.

  • Brands isolated by design: each agent only has access to its own universe of data.

  • Roadmap with 24 prioritized actions for the next stage.