Skip to main content

    Free tool

    RAG Readiness Checker

    What makes documents ready for retrieval-augmented generation (RAG)?

    RAG works best when documents are machine-readable, access rights are simple, and answers can point to a source. Scanned files, important tables, several languages, per-document permissions, sensitive data and frequent changes each add preparation work. This checker flags which of those apply to you and suggests an architecture.

    Describe your documents

    1. 1. Which document types will the AI read? (select all that apply)
    2. 2. How many documents?
    3. 3. Which languages are the documents in? (select all that apply)
    4. 4. How often do the documents change?
    5. 5. Do different users have different access rights?
    6. 6. Do the documents contain personal or sensitive data?
    7. 7. How exact must the answers be?
    8. 8. Do tables, forms or diagrams in the documents matter?

    Answer all eight questions to see your result.

    Frequently asked questions

    How is the readiness level worked out?

    Each answer adds preparation points — for example scanned files, per-document permissions or answers that must be exact. 0–3 points is Ready, 4–7 is Ready with preparation, and 8 or more Needs groundwork.

    Does RAG work with Indian languages?

    Yes, but check that the embedding model and tokenizer handle each language well, and test retrieval quality in every language separately rather than assuming English results carry over.

    Is my answer stored?

    No. The checker runs in your browser and nothing you enter is sent to us.

    More free tools

    Assess AI ReadinessContact