Types of Integrations#
Integrations let Unryo connect with external systems, enabling end-to-end monitoring, alerting, and visualization. They also allow the AI agent to query data sources during its reasoning to uncover insights. Depending on your needs, you can choose from:
- Data Sources — connect to existing observability tools and telemetry
- Infrastructure Integrations — capture data directly from devices, hosts, and networks
- Outbound Integrations — send alerts, notifications, or data to external destinations
- Custom Integrations — create tailored connectors for data not covered out-of-the-box
Data Sources#
These integrations connect Unryo to your observability tools, such as Prometheus, InfluxDB, VictoriaMetrics, OpenSearch/Elasticsearch, and others.
- Unryo does not duplicate telemetry data: it fetches only required metadata, metrics, and events as needed.
- After adding a data source:
- Dependencies are discovered
- Root cause insights and AI recommendations are computed
- Existing tools keep owning the data
Infrastructure Integrations#
Infrastructure integrations are built-in instrumentation (MLPacks) for direct monitoring of devices, hosts, networks, etc.
These integrations can:
- Collect key metrics
- Automatically discover topology and dependencies
- Enrich data with tags (location, role, etc.)
- Ingest logs and events
- Compute KPIs and detect anomalies through alert policies
- Display ready-to-use dashboards
Use infrastructure integrations when you want long-term storage, detailed device-level metrics, or coverage where telemetry tools don’t already reach.
Outbound Integrations#
Outbound integrations are for sending data, alerts, or notifications from Unryo to external systems or services. Examples include:
- Alerting tools (Slack, PagerDuty, email)
- Remote logging or storage
- External workflows triggered by value thresholds
Use these when you need Unryo to alert external systems, or integrate with existing workflows for incident management.
Custom Integrations#
If your data is not covered by existing “Data Sources” or “Infrastructure” integrations, you can build your own via the Configuration UI:
- Define custom metric/event ingestion
- Map custom logs or telemetry
- Tag or transform data as needed
- Create dashboards or alerts for it
Use custom integrations when you have proprietary or specialized monitoring needs, or using tools not yet supported out-of-the-box.