Dashboards, templates and Auto-Insights
There are three ways to get a dashboard, from fastest to most custom:
| Way | Where | Best when |
|---|---|---|
| Templates | Dashboards → Templates | You want a ready-made setup for a common source |
| Auto-Insights | Dashboards → Auto-Insights | You want Anosys to look at your data and suggest what to watch |
| Build your own | Dashboards → All Dashboards + Visualizations | You know exactly what you want to see |
Templates
- Open Dashboards → Templates and search or browse.
- Pick a template and select the endpoint it should read from.
- Review what it will create and deploy it.
A template can set up more than a dashboard: it can also add process units, alert triggers, pipelines and macros that belong together. You can edit everything afterwards.
Auto-Insights
- Open Dashboards → Auto-Insights.
- Source: choose an endpoint and time range.
- Analysis: Anosys looks for patterns that matter for that kind of data.
- Recommendations: pick the insights you want and add them as widgets.
What it looks for depends on the source, for example:
| Source | Examples of insights |
|---|---|
| AI agents and coding tools | Expensive failures, failure patterns, tool failures and reliability, repeated tools, session outliers, model comparison, cost efficiency |
| Websites | Daily visitors, geographic breakdown, audience quality |
| Network data | Threat signals, targeted ports |
Run it again any time with Recalculate Insights.
Build your own
- Open Dashboards → All Dashboards and create a dashboard.
- Add widgets from Visualizations. Each widget reads from one endpoint and time range.
- Arrange, resize and save.
Visualizations
| Group | Widgets |
|---|---|
| AI and traces | Trace Timeline, Trace Tree, RCA Report, Anosys Copilot |
| Time and trends | Advanced Time Series, Simple Time Series, Sparkline, Calendar Heatmap |
| Comparison | Bar Chart, Pie Chart, Table, Value Chart, Gauge Chart |
| Flow and distribution | Funnel Plot, Sunkey Diagram, Sunburst Plot, Heatmap, Scatter Plot, Network Graph |
| Geography | World Heatmap |
| Analysis | Contribution Analysis, Hilbert Curve |
| Notes | Notes Widget, Markdown Widget |
You can also bring in a saved chart with Visualizations → Import Graph.
Tips
- One dashboard per audience (for example engineering health, team spend, website traffic) stays readable.
- Name fields in Collecting → Field Mapping so chart labels make sense.
- Pair every summary chart with a Trace Timeline or Table so you can click through to the actual runs.
