Widgets

Every widget points at a dataset and renders a single view of the data. Add as many as you want, mix types freely, and resize individually.

KPI tile

A big-number tile for headline metrics: total revenue, active users, conversion rate. Optionally show a trend arrow compared to a previous period.

• Value column — what to display
• Aggregation — sum, avg, min, max, count
• Format — number, currency, percent
• Comparison — vs. previous period, vs. target

Table

A tabular view of the dataset. Sortable columns, pagination, search, and column visibility toggles for readers. Great for “Top 50 accounts”-style widgets.

Chart

DB AI Magic auto-picks a sensible chart type from the columns you select, but you can override:

• Line — trends over time
• Bar / Column — categorical comparisons
• Area — cumulative trends
• Pie / Donut — composition (use sparingly)
• Scatter — correlation between two numeric columns
• Heatmap — magnitude across two categories
• Stacked bar — composition over time or category

Per-chart configuration

For each chart you can configure axes, series, colours, legend position, and value labels. Charts inherit their colour palette from your workspace theme but you can override per-widget.

Text / Markdown

Drop a markdown block onto the grid to add context — a description, a methodology note, a link to a runbook. Text widgets render in public reports too, so they're a good place for the “data caveats” paragraph.

Filter slicer

A standalone widget that exposes one of the report's filters inline (instead of, or in addition to, the top filter bar). Useful for dashboards where the filter is a primary navigation control.

Best practices

• Group related widgets in the same row
• Put the most important KPI top-left — eye lands there first
• Use consistent colours for the same dimension across widgets
• Annotate non-obvious metrics with a text widget below them