# Integrations overview

> How the Integrations screen in Unifuse works: adding a source from the connector catalogue, the connector drawer, health checks, syncing and disconnecting.

Source: https://docs.unifuse.ai/integrations/overview/

**Integrations** in the sidebar is where an institution connects the systems that feed
Unifuse: document sources for the knowledge base, and the student information, CRM and
service-desk systems that feed student and applicant records.

## Before you start

- Opening the screen needs *manage institution*; adding, syncing, changing and disconnecting
  sources needs *manage connectors*.
- Have the credentials for the system you are connecting. The
  [connector catalogue](/integrations/connectors/) lists what each one asks for.

## Add a source

1. Select **Add Source**.
2. Choose the **Connector**. Connectors that sign in with OAuth are marked.
3. Enter a **Name** that will identify this source, for example "Admissions Drive".
4. Fill in the connector's fields. Required fields are marked. Secret fields are entered as
   passwords and stored encrypted, separately from the rest of the configuration.
5. Select **Create Source**.

For an OAuth connector, the browser then goes to the provider's consent screen and returns
to Integrations with a success or failure notice. Other connectors are connected as soon as
they are created.

## The source cards

Each source is a card showing its status (Connected, Syncing, Disconnected or Error), the
connector, the last sync time and a health line: healthy with the last check time, a failed
check with the reason, or "not supported" for connectors that have no health probe.

Select a card to open the connector drawer, which shows the status, connection details,
last sync and last error, the configuration (editable, with **Name** and the connector's
fields), and the actions:

- **Connect**: complete authorisation for an OAuth source that needs it.
- **Test**: run the health check now.
- **Sync**: run a sync now.
- **Disconnect**: remove the source. For a document source this also purges every document
  and embedding it contributed to the knowledge base.

## When sources sync

- Document sources are re-synced on a schedule, by default every 24 hours; the Web Crawl
  connector has its own **Recrawl every** setting.
- Record sources (SIS and CRM connectors) sync when you select **Sync**. The nightly
  SIS and LMS extract over SFTP runs on its own schedule; see
  [SIS and LMS data](/integrations/sis-and-lms-data/).
- Health checks run hourly.

## Sensitivity

Most connectors have a **Sensitivity** field, default `internal`. It becomes a tag on every
document from that source. Only sources set to `public` can be used to answer anonymous
visitors through the web chat widget or the SMS line.

## Related

- [Connector catalogue](/integrations/connectors/)
- [Knowledge base](/enrollment/knowledge-base/)