RAG & Documents

Upload your documents and chat with them using RAG (Retrieval-Augmented Generation). SafeRag finds relevant content from your documents to provide context-aware responses.

What is RAG?

RAG (Retrieval-Augmented Generation) is a technique that enhances AI responses by retrieving relevant information from your documents before generating an answer.

How It Works

Upload Documents

Add your documents to SafeRag. They're processed and stored locally as vector embeddings.

Ask Questions

In RAG mode, ask questions related to your documents.

Retrieval

SafeRag searches your documents for relevant sections using semantic similarity.

Enhanced Response

The AI uses the retrieved content as context to provide accurate, document-based answers.

💡 All Local
All document processing happens on your Mac. Your documents are never uploaded to external servers.

Supported Document Types

SafeRag can process the following file types:

Type Extensions Notes
PDF .pdf Text-based PDFs. Scanned PDFs require OCR.
Plain Text .txt Any plain text file
Markdown .md Markdown documents
Images .png, .jpg, .jpeg For visual context (with compatible models)

Uploading Documents

Access the Documents section from the sidebar or press Cmd + D.

Method 1: Click to Upload

Click the Upload Documents button and select files from the file picker.

Method 2: Drag and Drop

Drag files directly onto the Documents area to upload them.

📸
Screenshot: Documents View
Capture the Documents section showing the upload area (drag-drop zone), upload button, and a list of uploaded documents with their details.
documents-view-upload.png

Upload Process

When you upload a document:

  1. The file is copied to SafeRag's local storage
  2. Text is extracted from the document
  3. The text is split into chunks
  4. Each chunk is converted to a vector embedding
  5. The document appears in your list
💡 Processing Time
Large documents may take a few seconds to process. You'll see a progress indicator during upload.

Document Limits

Plan Document Limit
Free 5 documents
Pro Pro Unlimited documents

The free plan allows up to 5 documents. To upload more, upgrade to SafeRag Pro.

Managing Documents

Viewing Document Details

Click a document to see its details:

  • Document ID - Unique identifier
  • Chunk Count - Number of text chunks created
  • Upload Date - When the document was added
  • User - Who uploaded it (Multi-User mode)
📸
Screenshot: Document Details
Capture a document detail view showing the document name, ID, chunk count, upload date, and any other metadata.
document-details-view.png

Deleting Documents

To remove a document:

  1. Select the document in the list
  2. Click the Delete button (trash icon)
  3. Confirm the deletion

Bulk Selection

Use checkboxes to select multiple documents for bulk deletion.

Watch Folders Pro

Watch Folders automatically index documents from specified directories. When you add or modify files in a watched folder, SafeRag updates your knowledge base automatically.

Adding a Watch Folder

Open Watch Folders Tab

In the Documents section, switch to the Watch Folders tab.

Add Folder

Click Add Folder and select a directory from your Mac.

Automatic Indexing

SafeRag begins indexing all supported documents in the folder.

📸
Screenshot: Watch Folders
Capture the Watch Folders tab showing a list of watched folders with their status (Active/Inactive), document count, and path.
watch-folders-list.png

Watch Folder Features

  • Status Toggle - Enable/disable watching without removing the folder
  • Document Count - See how many documents are indexed from each folder
  • Recent Documents - View recently added documents
💡 Pro Feature
Watch Folders requires SafeRag Pro. During the free trial, you can test this feature for 24 hours.

Using RAG Mode

To chat with your documents:

Create a RAG Session

Create a new session and select RAG Mode from the mode picker.

Ask Document-Related Questions

Ask questions about your documents. For example: "What are the key points from the quarterly report?"

Review Responses

The AI will reference relevant sections from your documents in its response.

📸
Screenshot: RAG Chat
Capture a RAG mode conversation showing a question about documents and the AI's response that includes information retrieved from the documents.
rag-chat-example.png

Tips for Better Results

💡 Use Specific Questions
Ask specific questions rather than general ones. "What was Q3 revenue?" works better than "Tell me about the report."
💡 Organize Documents
Keep related documents together. RAG works best when your knowledge base is focused on related topics.
💡 Use Embedding Models
For better retrieval accuracy, ensure you have a dedicated embedding model installed. SafeRag uses this for document similarity searches.

Next Steps

🤖 Manage AI Models 💬 Chat Interface