How analysis works

Querio develops analysis in a notebook, where you can follow the work from your question to the queries, calculations and results behind an answer.

Work with the agent, edit the code yourself, or combine both approaches.

From a question to a result

When you ask a question, the agent uses your selected datasources and context to understand the task and develop the analysis.

It can write SQL to query your data, use Python for calculations, and present the results as tables or charts. You can inspect that work in the notebook and ask for changes as you go.

For example, you might start with monthly revenue, then ask for a regional breakdown or investigate an unexpected change. The analysis can develop in the same notebook as the conversation progresses.

A notebook keeps the analysis connected

A notebook contains cells for queries, calculations, text, charts and interactive controls.

Cells can build on each other. A query might supply order data, a calculation turn it into monthly totals, and a chart display the trend. Querio tracks those dependencies so it can run the affected calculations and charts together.

This lets you develop an analysis in understandable pieces and trace a result back to its inputs.

More about notebooks

An Explore brings the work together

An Explore combines a notebook with its conversations. Its board layout presents selected notebook outputs for your team.

Use Edit to develop the analysis, change cells and arrange the board. Your notebook changes and editing conversation are saved with the Explore.

Use View to explore the presentation, interact with controls and ask follow-up questions. When you want to develop a follow-up independently, continue in a new Explore with the relevant cells and conversation.

View conversations belong to that live viewing session. To keep one as editable work, use Explore from a board.

Work from the app, CLI or MCP

Each interface works with Querio's notebook analysis:

• App: develop the notebook with the agent, inspect results and build boards.
• CLI: work from your terminal or let a coding agent create, inspect and edit notebook cells.
• MCP: connect a compatible AI assistant to Querio's notebook and data tools.

Workspace and datasource permissions apply through each interface.

Get started with CLI | Get started with MCP

Return to saved work

Saved work includes notebook code, layout, stored outputs and the editing conversation. You can reopen it, inspect the analysis and continue working.

A displayed result can come from an earlier execution. After changing code, check the new execution and output. When you need an updated answer, rerun the analysis against the relevant data.

Check and correct an answer explains how to review the logic, inputs and results. For live execution and saving behavior, see Environments, sessions and runs.

Try your first analysis

Choose a datasource, ask a focused question and inspect the notebook behind the answer.

Get started in the app