For the complete documentation index, see llms.txt. This page is also available as Markdown.

Ask Dan

Ask questions in natural language, explore DinMo data, and turn analysis into clear, shareable insights with Ask Dan.

Turn warehouse data into clear decisions, directly from DinMo.

Ask Dan is your AI data analyst inside DinMo. Ask a question in natural language, bring the right workspace context, and get an answer grounded in the data available to your workspace.

Dan helps you explore questions, compare populations, surface patterns, and turn results into tables or visualizations that are easier to understand and act on. Because every analysis lives in a conversation, you can refine the question, challenge an interpretation, and continue exploring without starting again.

Ask Dan conversation showing a data analysis with a chart and referenced DinMo context
Move from a business question to an explained, visual result in one conversation.

From question to insight

You can open Ask Dan from the Home page or the Ask Dan panel.

  1. Ask a business question in natural language.

  2. Add relevant DinMo context when the question concerns a specific workspace object.

  3. Follow the progress of the analysis while Dan works.

  4. Review the answer, table, visualization, and method when available.

  5. Continue the conversation to refine the scope, compare another population, or investigate a result.

If you leave the conversation while Dan is working, an unread indicator lets you know when a new result is ready.

Bring the right context into the conversation

Use @ in the composer to reference relevant DinMo objects available to you. This gives Dan precise workspace context without requiring you to copy technical identifiers or describe the object manually.

References remain connected to the conversation, making follow-up questions easier to understand. The objects you can mention and open depend on your current workspace permissions.

Explore results in the clearest format

Dan adapts the presentation to the result. Depending on the analysis, you may receive an explanation, a table, a chart, or a combination of formats.

When several compatible presentations are available, you can change how a saved result is displayed without running the warehouse query again. Tables keep the complete result accessible, while charts focus attention on the most useful patterns. The Method view explains how the result was produced.

Collaborate without losing the thread

When conversation sharing is enabled for your workspace, you can share an Ask Dan conversation with other workspace members. Shared conversations preserve the questions, results, visualizations, and referenced objects needed to understand the analysis.

You can also ask about a precise part of a previous answer:

  1. Select the relevant excerpt in a Dan response.

  2. Choose Add to chat.

  3. Optionally add a comment explaining your question.

  4. Review, edit, or remove the attached excerpts above the composer.

  5. Send the message when the context is ready.

You can attach up to 10 excerpts to one message. Dan receives each excerpt, its position in the original answer, and any comment you added. Adding an excerpt never sends a message automatically.

Selected Ask Dan answer excerpts with comments attached above the message composer
Anchor a follow-up question to the exact passage you want Dan to reconsider or explain.

Trust what you see

Ask Dan checks the current state of referenced DinMo objects in personal and shared conversations:

  • Available: you can preview and open the object.

  • Deleted: the object can no longer be found. Its last known label remains visible, but navigation is disabled.

  • Not available in this workspace: you do not currently have access to the object, or it belongs to another workspace. This does not mean that the object was deleted.

  • Unable to check this object: a technical lookup failed and can be retried.

Dan respects your current permissions and keeps historical conversations readable without exposing raw object identifiers as fallback labels.

Questions to try

Start with a question tied to a decision you need to make:

  • “Which customer segments grew the most this month?”

  • “Compare conversion behavior between these two audiences.”

  • “Show the distribution of this metric and highlight unusual values.”

  • “Which categories contribute most to this result?”

  • “Explain how this analysis was calculated.”

Last updated