AI Transparency Notice

Last updated: April 18, 2026

This notice explains how DocMind AI uses AI and retrieval systems to answer questions on your uploaded documents.

1. Processing Flow (High-Level)

You upload a PDF to private storage.
Text is extracted, chunked, and converted into embeddings.
Embeddings are stored in a vector database for semantic retrieval.
At query time, retrieval returns relevant text chunks.
Reranking and LLM generation produce an answer with citations.

2. What Is Sent to External AI Providers

Your question text.
Relevant extracted document excerpts/chunks required for ranking and answer generation.
Embedding requests for chunk text and query text.

3. What Is Not Shared as Public Content

We do not publicly publish your files. We do not market, sell, or openly expose your uploaded documents. Processing is limited to service operation and configured subprocessors.

4. Important Accuracy Notice

AI responses may be incorrect, incomplete, outdated, or contextually wrong. Always verify important outputs, especially for legal, financial, compliance, or contractual decisions.

5. Language and OCR Features

For multilingual and Hindi workflows, DocMind AI may use OCR fallback and language routing logic. OCR and retrieval quality may vary by input PDF quality.

6. User Controls

Delete documents anytime in-app.
Document deletion is intended to remove related vectors, extracted chunks, and document-scoped history, in addition to the stored file.
Contact support for full account-level deletion requests.

7. Related Documents

Questions: adityagupta20042003@gmail.com