Published at: 2026-10-15
Knowledge Sources and Ingestion Methods Reference
Review the knowledge sources and ingestion methods in the Knowledge Base 1.0 customized backend.
⚠️
Standard Knowledge Base 2.0 prefers the frontend system data import feature. The table below applies only when the enterprise has built customized backend ingestion rules based on Knowledge Base 1.0.
| Knowledge source | Main use | Key rules |
|---|---|---|
| Knowledge articles | Maintain text, rich text, and attachment knowledge in the 1.0 customized CRM backend | When a sync rule runs, one knowledge article generates one AI knowledge entry |
| Fxiaoke network disk files | Use files in the company network disk as knowledge | The system initializes the “Knowledge Base Files” directory; sync the specified directories and their files |
| CRM objects | Ingest existing knowledge or manuals in enterprise custom objects | Data can be filtered by field conditions and field mapping can be configured; only data added after the rule is enabled is synced |
| Work circle sharing | Turn sharing with a specified topic into knowledge | The “Knowledge Base” topic is available by default; each matching sharing generates one AI knowledge entry |
| Training assistant courses | Turn courses, video transcripts, materials, and attachments into knowledge | The course channel structure is synced as directories; large videos may need to wait for scheduled system processing |
General Checks
-
Confirm that the corresponding ingestion rule is running.
-
Confirm that the synced data scope and directory structure match expectations.
-
Confirm that required fields such as title, content, or attachments are mapped.
-
Check new, updated, or failed records in the ingestion log.
Related Topics
- Knowledge Ingestion Rules
- Import System Data