Knowledge Base

Xentree supports knowledge-aware retrieval so responses can stay grounded in relevant context.

Use cases

retrieve supporting context for an AI answer

isolate knowledge by conversation or user scope

manage private and shared knowledge items

build more reliable AI workflows

Authentication and scopes

Knowledge Base search supports both bearer and API key authentication.

Recommended API key scopes:

rag:read for query/search usage

rag:write for ingestion/update workflows

Main endpoints

List documents

GET /api/rag/documents

Upload document

POST /api/rag/upload

Query knowledge (RAG)

POST /api/rag/query

Search the knowledge base (retrieval)

POST /api/knowledge/search

/api/rag/query returns a synthesized answer. /api/knowledge/search returns structured retrieval results (useful when you want to build your own grounding or UI around hits).

Example query request

{
  "query": "meeting notes about the launch plan"
}

Design principles

The knowledge layer is designed around:

privacy-aware access control

conversation isolation

relevance scoring

smart retrieval for grounded generation

When to use it

Use Knowledge Base APIs when your application needs better recall, context-aware search, or document-backed responses.

Use cases#

Authentication and scopes#

Main endpoints#

List documents#

Upload document#

Query knowledge (RAG)#

Search the knowledge base (retrieval)#

Example query request#

Design principles#

When to use it#