Skip to main content

​
Atoti Intelligence Essentials

This is part of the Atoti Intelligence Essentials offer.
Visualize This is an AI assistant in the Atoti UI that turns natural-language requests into visualizations. Understanding what it sends to the LLM clarifies why configuration improves results.
To try Visualize This, first enable it: see Set up Visualize This in Java or Set up Visualize This in Python, then Configure Visualize This to supply cube context.

​
How a request is answered

  1. The user describes what they want in natural language in the Atoti UI (for example, “show revenue by country for last year”).
  2. Atoti builds the context sent to the LLM: the cube’s data model (dimensions, hierarchies, levels, measures) plus any descriptions and system prompt you configured.
  3. The LLM interprets the request against that context and produces a query and a widget definition.
  4. Atoti renders the visualization from the result, using the same query engine as the rest of the UI, so the numbers are consistent with everything else.

​
Why context matters

The LLM only knows what it is told about your cube. The more business meaning you provide — clear cube, dimension, and measure descriptions, and a focused system prompt — the more accurately it maps a request to the right members and measures. This is what the configuration step supplies. The LLM runs on the provider you configured (see Set up an LLM), so responses stay grounded in your data rather than general knowledge.