Next Steps

You have Prelude Collector installed, configured, licensed, and collecting from a first device. The rest of the documentation breaks down by topic so you can jump to whichever part of the collector you need next.

Learn the model

• Concepts - the canonical reference for Device, Protocol, Subscription, Model, Field, Mapping, Transform, Output, and Snapshot. Read this once before you go deep.

Build real collections

• Tutorials - end-to-end walkthroughs that go further than the first-collection page: writing your own Models, building Mappings, shaping data with Starlark Transforms.
• Best Practices - patterns the team recommends for production: subscription sizing, model versioning, output fan-out, and operational hygiene.

Connect to your devices

Each protocol page covers connection setup, supported features, and known caveats:

• gNMI - streaming telemetry from Cisco, Juniper, Nokia, Arista, and other platforms with OpenConfig support.
• NETCONF - structured config and state retrieval over SSH.
• SNMP - poll-based collection against MIBs, including the bundled vendor MIB browser.
• CLI - screen-scraping with parsers, for devices that expose nothing better.

Send data downstream

Outputs decide where Snapshots go after collection:

• NATS - the default streaming bus; use this unless you have a specific reason not to.
• Prometheus - scrape-based metrics with label control.
• InfluxDB - line-protocol writes for time-series storage.
• Kafka - durable streaming for multi-consumer pipelines.
• Webhook - HTTP POST per Snapshot, useful for ad-hoc integrations.
• File - JSONL on disk, mostly for debugging and offline analysis.

Drive the collector programmatically

• API Reference - the full REST API rendered with Scalar, including request and response schemas for every endpoint you saw in the walkthrough and many you did not.

Stuck?

If something is not behaving the way the docs describe, see Support for how to reach the team.