How Quantum-ArchiveMaster works

This public site is a read-only demo (Search, Q&A, Analytics on sample documents). The steps below describe the full private deployment you would run on your own servers or VPC—where uploads, folder import, wiki, and local models are available.

End-to-end flow

1. 1. Upload files

   Authenticated users upload PDFs and common office documents one at a time. Each file is stored in the private archive and queued for text extraction.

2. 2. Import a folder

   For large libraries, connect a documents folder, preview counts, and run a bulk import job. Long jobs support progress tracking, pause, and resume—so overnight imports are practical.

3. 3. Extract text (with OCR when needed)

   Native text is extracted from digital files. Scanned or image-heavy PDFs go through OCR so content becomes searchable and usable for Q&A.

4. 4. Index into the archive

   Documents appear in the Archive view with links to originals. Search indexes titles and extracted text so teams find knowledge by content, not filenames alone.

5. 5. Document wiki summaries

   After a successful import or upload, a background job generates a wiki summary page per document. Status badges show ready, pending, or failed. Overview-style Q&A can prefer ready wiki text.

6. 6. Search

   Users search the archive and open matching documents. On-prem search uses your deployment’s search stack; this cloud demo uses a lightweight keyword index over sample PDFs.

7. 7. Archive Q&A

   Ask natural-language questions. Answers are grounded in retrieved passages with citations and source links. On-prem typically uses a local LLM (e.g. via Ollama); this demo uses OpenAI gpt-4o-mini.

8. 8. Analytics dashboard

   Operators track uploads, searches, and LLM queries with today/yesterday snapshots and yearly / monthly / weekly views—so usage is visible without exporting logs.