DocMind AIEnterprise Document Analyzer
Powered by Google Gemini & pgvector

Ask Your Documents Anything

Upload a PDF, ask questions in plain English, and get grounded answers backed by your own documents — with source citations.

Document Ingestion

Upload a PDF to chunk, embed, and index its content into Supabase pgvector.

Drag & drop your PDF here

or click to browse files

PDF only · Max 20 MB

Ingestion Pipeline

  1. 1

    Parse PDF

    Extract raw text via pypdf

  2. 2

    Chunk Text

    512-word windows, 50-word overlap

  3. 3

    Embed

    gemini-embedding-001 (Vector)

  4. 4

    Store

    Supabase + pgvector HNSW index

AI Chat Interface

Ask questions about your uploaded documents. Answers are grounded in retrieved context.

Ask your documents anything

Upload a PDF first, then start a conversation.