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

# About integration

Below is an overview of how to integrate with the Engage platform. It covers:

* Available integration patterns
* The authentication model
* Key considerations for data synchronization and error handling

Use this section to consider the right integration approach before diving into the implementation details.

## Integration patterns

The Engage platform supports multiple integration patterns depending on use case, your data volume, and your latency requirements.

* **API**: Synchronous, request-based access to platform data and operations. Suitable for real-time and transactional workflows where your system actively reads or writes data.

* **Webhooks**: Event-driven notifications sent from Engage to your system. Used for reacting immediately to changes without the need for polling.

* **Technology partners**: Using a pre-made integration with one of Engage's many technology partners.

* **File-based imports**: Scheduled or on-demand FTP batch imports using files.

* **File-based exports**: Batch exports delivered as files via FTP, suitable for downstream systems that consume data from files.

* **Delta sharing**: Incremental data access designed for analytics platforms and data lakes. Requires a dedicated setup and is updated once per day.

* **BI export**: Curated analytical datasets, updated once per day. Used for datasets not yet available via delta sharing. Not intended for operational integrations.

## Authentication

Authentication in Engage works like this:

* All integrations authenticate use *API keys*.
* The API key is included in the request header for each API call.
* Multiple keys can be created, but all keys currently have the same level of access.
* Token-based authentication and scoped permissions are not supported.
* Different integration patterns provide different guarantees concerning data freshness and consistency.
* Engage does not support request-level idempotency. Duplicate prevention is handled through **configurable duplicate control**, where you define which customer attributes must be unique. Integrations should be designed with this in mind, particularly for retries or reprocessing data.

## Error handling

Error handling depends on the integration method used.

* APIs require synchronous error handling per request.
* Webhooks require validation and acknowledgement of incoming events.
* Imports and file-based exports require monitoring of job execution and result files.
* Delta sharing and BI export require validation of data completeness per refresh cycle.

More detailed guidance is provided in the respective sections.