DocumentationKnowledge Bases
Knowledge Bases
Document collections that agents search to answer from your own content (Smart RAG).
A knowledge base is a collection of documents (PDFs, Office files, web pages and more) that KNetAI splits, indexes and searches, so that Kue, K-Nets and bots answer from your own content.
Hybrid Search & Interactive Chatbots
KnowledgeBase combines graph, vector, and full-text search into an enterprise RAG pipeline:
- Interactive Bots: Open chat sessions directly with dedicated agent personas (such as
Analyst). - Collection Targeting: Type
@in the prompt composer to scope retrieval to specific document collections. - Multi-Modal Context: Attach supplementary files, browse the visual knowledge graph, or toggle live web search alongside your private documents.
- Prompt Engineering: Customize bot behavior with prompt overrides and model parameters directly from the chat interface.
Pages
| Page | What it is for |
|---|---|
| Chatbots | Interactive chat interfaces for querying collections with dedicated bot personas |
| Collections | Your knowledge bases. Add collection creates one; upload files or connect a data source |
| Marketplace | Browse and subscribe to published domain collections |
| Settings | Embedding models, chunking strategies, and retrieval thresholds for this workspace |
Typical tasks
- Ask Kue: "Make a knowledge base from these PDFs", then ask questions about them.
- Feed a collection from a website with a crawler.
See Smart RAG Retrieval for how search works.