Document Ingestion

Ingestion is the process of reading your documents, splitting them into chunks, embedding them as vectors, and storing them in Qdrant for search.

Web UI

Open the Web UI at http://127.0.0.1:3235/ui/ingest. You can:

• Upload individual files directly through the browser
• Specify a directory path on your machine to ingest recursively

Documents are placed in data/documents/ and code files in data/code/ within the RagGo directory.

MCP tools

You can also ingest documents via MCP tools from any connected client:

// Ingest a single file
{
  "tool": "ingest_file",
  "input": {
    "file_path": "/path/to/document.pdf"
  }
}

// Ingest an entire directory
{
  "tool": "ingest_documents_directory",
  "input": {
    "directory_path": "/path/to/my-documents/"
  }
}

Supported file types

Chunking

RagGo splits documents into chunks of up to 1000 tokens. This provides a good balance between precision and context for most document types.

Re-ingestion

Ingesting the same path again updates changed files and skips unchanged ones (by content hash).