---
title: "Knowledge Source Configuration"
description: "Chunking, embedding, vector index, sources. RAG setup and ingestion pipeline."
canonical: https://www.norcaster.com/docs/knowledge
---
<!-- Generated at build time from the rendered page by portal/scripts/generate-llms.mjs. The page is the source of truth; edit it, not this file. -->

# Knowledge Source Configuration

Configure RAG: chunking profiles, embedding profiles, vector indexes, and knowledge sources. Ingest from upload, S3, or web.

## Portal

**Dashboard → Knowledge** — Manage chunking profiles, embedding profiles, vector indexes, knowledge bases, and sources. Add sources (upload, S3, web), upload documents, trigger ingestion.

## Setup flow

1. **Provider** — Add a provider with embeddings (e.g. Voyage AI)
2. **Chunking profile** — Strategy (fixed, markdown, semantic), chunk size, overlap
3. **Embedding profile** — Provider model for embeddings
4. **Vector index** — Backend (pgvector, Pinecone, Weaviate, Qdrant)
5. **Knowledge base** — Container with default profiles + vector index
6. **Sources** — Add upload, S3, or web source
7. **Ingestion** — Upload docs, trigger ingestion job

[Follow the full RAG deployment guide →](https://www.norcaster.com/docs/rag-deployment)

## Source types

- **upload** — Upload files via portal or API
- **s3** — S3-compatible storage (MinIO, Scaleway, etc.)
- **web** — Fetch from URL

## Pipeline

Ingestion: fetch documents → chunk → embed → index into vector store. Documents and chunks are stored in Postgres; embeddings go to the vector backend (pgvector, Pinecone, etc.).