Talk to Your Data (RAG)

What is RAG?

Retrieval-Augmented Generation (RAG) lets you ask questions in plain English and get answers that come directly from your own data — not from general AI knowledge.

Instead of an AI that guesses, RAG works in two steps:

1. Retrieve — find the most relevant chunks of your documents, files, or database entries
2. Generate — use a language model to compose a precise answer, grounded in what was retrieved

The result is an AI that knows your data as well as your best analyst — and can answer questions about it instantly.

How to Access

Open the File Manager and click the Sparkle icon (✦) next to any file, folder, or dataset. This opens the RAG chat interface for that data source.

You can ask questions directly, or select multiple files and folders to query across a broader knowledge base.

What You Can Do

• Ask questions about documents — contracts, reports, research papers, manuals, meeting notes
• Query your database — get natural language answers from structured records without writing SQL
• Build a knowledge base — index your internal wiki, product docs, or support history and make it searchable by conversation
• Cross-document analysis — compare information across multiple files in a single question

Example Prompts

"What are the key risk factors mentioned in the last three due diligence reports?"

"Summarize the renewal terms across all client contracts uploaded this quarter."

"Which studies in my research folder show a statistically significant result?"

"What does our internal policy say about data retention for EU customers?"

Tips

• The more structured and clean your data, the better the answers. Use the Data Transformation node to preprocess before indexing.
• For large document sets, organize files into folders by topic so retrieval stays focused.
• RAG works best for factual, document-grounded questions. For building and training custom AI models, see Train AI Model.

Responsible Developers: Aman, Julia.