> ## Documentation Index
> Fetch the complete documentation index at: https://docs.bondata.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Dashboards

> Group charts onto a single page, each one running its own query

A dashboard is a set of charts arranged on a grid. Each chart holds its own SQL query and runs it when the dashboard opens, so the numbers reflect whatever your integrations hold at that moment. Nothing is stored as a snapshot.

Dashboards belong to a project. Click **Dashboards** in the sidebar to see the ones in the project you're working in.

<Frame caption="The Dashboards list for a project">
  <img src="https://mintcdn.com/bondata/6jXa9WimaGAbIZRK/images/guide/dashboards/dashboards-list.png?fit=max&auto=format&n=6jXa9WimaGAbIZRK&q=85&s=074aa255a906b2668e47d71bd448a474" width="1568" height="311" data-path="images/guide/dashboards/dashboards-list.png" />
</Frame>

## Create a dashboard

Click **New Dashboard** and give it a name. The description is optional and appears under the title.

You can also describe one in **Chat**. Switch to **Analyst** mode and say what you want to see — for example, *"a revenue dashboard with monthly trend, top products, and a breakdown by region"*. BonData writes the queries, picks chart types, runs each query to check it returns rows, and saves the dashboard. Anything it produces can be edited afterwards.

## Add a chart

Click **Add Widget** to open the chart editor.

<Frame caption="The chart editor: query, chart setup, and preview">
  <img src="https://mintcdn.com/bondata/6jXa9WimaGAbIZRK/images/guide/dashboards/chart-editor.png?fit=max&auto=format&n=6jXa9WimaGAbIZRK&q=85&s=2da9b6de36d8f44a0c7e10628c8418c2" width="1568" height="692" data-path="images/guide/dashboards/chart-editor.png" />
</Frame>

1. Write a query and click **Run**. Reference your data as `"Integration".Entity` — for example `"Demo Shop".orders`. The columns it returns are listed under **Result Columns**.
2. Choose a chart type.
3. Drag columns from **Result Columns** into the chart's slots. A pie takes a **Category** and a **Value**; a bar takes an **X-Axis** and a **Y-Axis**.
4. Set the value format: currency, percentage or number, decimal places, and compact notation (1.2K / 1.2M).

The preview updates as you go.

To change a chart without writing SQL, use **Edit with AI** and describe the change.

## Chart types

| Type                  | Use it for                                                                  |
| --------------------- | --------------------------------------------------------------------------- |
| **KPI**               | A single number, optionally with a trend arrow or sparkline                 |
| **Trend**             | A measure over time                                                         |
| **Bar**               | Comparing categories, vertical or horizontal                                |
| **Combo**             | Two measures on shared axes, such as orders as bars and customers as a line |
| **Pie**               | Share of a total, as a pie or donut                                         |
| **Funnel**            | Drop-off through sequential stages                                          |
| **Gauge**             | One value against a range                                                   |
| **Table**             | Sortable rows with per-column formatting                                    |
| **Pivot**             | One dimension cross-tabulated against another                               |
| **Scatter**           | Relationship between two measures                                           |
| **Heatmap**           | Density across two dimensions                                               |
| **Treemap**           | Nested proportions                                                          |
| **Histogram**         | Distribution of one measure                                                 |
| **Radar**             | Several measures compared on one shape                                      |
| **Text**, **Divider** | Headings and separators                                                     |

### Turn a KPI into a trend tile

Return one row per period — a month column and a value column — and drag the date column into the KPI's **X-Axis** slot. The tile then shows the latest period's value, a sparkline of the series, and the change against either the first period or the previous one. Give the tile some extra height so the sparkline has room.

## Move and resize charts

Drag a chart by its header to move it. Drag its bottom-right corner to resize it. The grid is 12 columns wide, and the layout saves as you go.

<Tip>
  Layout is saved per dashboard, not per person. Rearranging affects everyone who opens it.
</Tip>

## Dashboard filters

Dashboard filters sit above the charts and apply to every chart whose query uses them, so one control moves the whole page. A date range, a region, or a customer segment are typical.

Click **Add filter** and describe what you want to filter by. BonData creates the filter and adds it to the relevant queries. Once a dashboard has at least one filter, the button becomes **Manage filters**, where you can rename them, reorder them, and set defaults.

A filter only affects charts whose query references it. If a chart doesn't move when you change a filter, that chart isn't wired to it — ask AI to add it.

## Chart filters

Individual charts have their own **Filter** button. Pick from the values already present in that chart, or drag across a time axis to narrow to a range.

Chart filters aren't saved and apply only to the chart you're looking at. Use them to inspect something; use a dashboard filter when you want it to persist.

## Chart actions

Hover a chart and open the **⋮** menu.

<Frame caption="Per-chart actions">
  <img src="https://mintcdn.com/bondata/6jXa9WimaGAbIZRK/images/guide/dashboards/widget-menu.png?fit=max&auto=format&n=6jXa9WimaGAbIZRK&q=85&s=7bb695922b3cbefeabf5d5a1221ca8e2" width="535" height="282" data-path="images/guide/dashboards/widget-menu.png" />
</Frame>

| Action           | Result                                        |
| ---------------- | --------------------------------------------- |
| **Edit**         | Opens the chart editor                        |
| **Edit with AI** | Describe the change in plain language         |
| **Expand**       | Full-screen view                              |
| **Duplicate**    | Copies the chart onto the same dashboard      |
| **Refresh**      | Re-runs the query, ignoring any cached result |
| **Download CSV** | Exports the rows behind the chart             |
| **Delete**       | Removes the chart                             |

The **`</>`** button beside the menu shows the chart's query and lets you edit it directly.

## Export a dashboard

<Frame caption="Dashboard actions">
  <img src="https://mintcdn.com/bondata/6jXa9WimaGAbIZRK/images/guide/dashboards/dashboard-toolbar.png?fit=max&auto=format&n=6jXa9WimaGAbIZRK&q=85&s=bbf3186214a7c045d11f8b058741b61a" width="1568" height="98" data-path="images/guide/dashboards/dashboard-toolbar.png" />
</Frame>

**Download PDF** captures the dashboard as it appears on screen, including charts below the fold.

**Ask AI** opens Chat beside the dashboard, so you can ask about what you're looking at or change charts and watch them re-render.

The **⋯** menu holds **Manage Filters**, **Duplicate Dashboard**, and **Delete Dashboard**.

<Frame caption="A dashboard with KPI tiles, a time series, and category and regional breakdowns">
  <img src="https://mintcdn.com/bondata/6jXa9WimaGAbIZRK/images/guide/dashboards/dashboard-overview.png?fit=max&auto=format&n=6jXa9WimaGAbIZRK&q=85&s=26c667fa0907c70e8e7a3b41f9bbdbb0" width="1915" height="868" data-path="images/guide/dashboards/dashboard-overview.png" />
</Frame>

## Further reading

* [AI Chat](/guide/chat/ai-chat) — building and editing dashboards from a conversation
* [Connecting your integrations](/guide/getting-started/connecting-integrations) — the sources charts query
* [Bondata Cloud](/guide/data-warehouse/bondata-cloud) — where queried data is stored
