Core concept: Knowledge bases
A knowledge base is a searchable collection of your business content — the source of truth your assistant cites when answering document-based questions.
What goes in a knowledge base
Section titled “What goes in a knowledge base”| Source | Examples |
|---|---|
| Uploads | PDF, DOCX, TXT, HTML |
| URL import | Single page or help article |
| Website crawl | Entire site or path-limited crawl |
Each file is extracted, chunked, embedded, and indexed in Qdrant. See Processing and indexing.
How assistants use them
Section titled “How assistants use them”Attach one or more knowledge bases to an assistant to enable RAG (retrieval-augmented generation). At answer time Chatevo runs hybrid semantic + keyword search, passes top chunks to the model, and returns citations in the reply.
- Single KB — Attach knowledge bases
- Multiple KBs — Multi-KB strategy
Document classification
Section titled “Document classification”Every document carries a classification level that controls visibility at retrieval time:
| Level | Typical use |
|---|---|
| public | Customer-facing FAQs, marketing copy |
| internal | Team runbooks, internal policies |
| confidential | HR, finance, contracts |
| restricted | Highly sensitive — narrowest access |
Plan limits
Section titled “Plan limits”Knowledge base count, documents per KB, and storage vary by plan. Auto-retrain (re-index on content change) is available on Standard and above.
| Plan | Knowledge bases | Documents per KB |
|---|---|---|
| Free | 1 | 5 |
| Starter | 1 | 20 |
| Standard | 3 | 100 |
| Pro | 15 | 500 |
| Enterprise | Custom | Custom |
Full comparison: Plans compared.
Tune retrieval
Section titled “Tune retrieval”Per-assistant RAG settings (similarity threshold, max results) live on the assistant. Per-KB search settings are under Tune search settings.