Product · Document Q&A

Your documents, answered with page-level proof.

Upload the spec, the contract, the policy, the research report. Ask in plain language. Every answer cites the exact page and section it came from - and clicking the citation opens the document right there. Scanned PDFs included.

Document Q&A · 3 sourcesconfidence 0.94
Q

What are our recovery time commitments, and who owns them?

Tier 1 services commit to recovery within two hours1, backups are verified through quarterly restore drills2, and the operations team owns execution and verification3.

  • 1
    p. 6 · Recovery Policy · Backup & Security Standard

    Tier 1: patient-facing services restore within 2 hours.

  • 2
    p. 5 · 6.2 Database backup · Backup & Security Standard

    Quarterly restore drills validate every backup class.

  • 3
    p. 9 · Roles · Backup & Security Standard

    Operations owns backup execution and verification.

Why this design exists

“It's somewhere in the PDF” is not an answer.

Generic AI chat summarises documents confidently and cites nothing. That is unusable the moment someone asks “where does it say that?” - in a compliance review, a contract negotiation, or a UAT sign-off.

Document Q&A retrieves the relevant passages first and answers only from them. Every claim carries a page-and-section citation, and if the selected documents do not contain the answer, it says so plainly instead of improvising.

The pipeline

From upload to page-cited answer.

  1. 01

    Upload

    PDF, Word (.docx), or Markdown - up to ten files per batch, verified before processing starts.

  2. 02

    Read

    Every page is read, scans included: text, tables, and the position of each passage on the page.

  3. 03

    Structure

    Headings, sections, tables, and figures become one navigable outline of the original document.

  4. 04

    Index

    Every passage becomes searchable by meaning and by keyword, inside your workspace boundary.

  5. 05

    Ask

    Answers cite the exact page and section. Click a citation and the document opens right there.

Question shapes

What you can ask a document.

Factual
  • What are the recovery time commitments?
  • Which regulation grounds the data-residency clause?
  • What is the penalty for late delivery?
Synthetic
  • Summarise every obligation we accept in section 7.
  • List the controls that apply to patient data.
  • What changed between contract v2 and v3?
Cross-source
  • Does the spec match what the client agreed on the call?
  • Which requirements in the PRD were never discussed in review?
  • Compare the security posture across all three vendor documents.
Citations that open the page

Click the claim, land on the evidence.

Every citation carries the page, the section heading, and the position of the quoted passage. One click opens the original document at that exact spot - no scrolling through a 40-page PDF to verify a sentence. The original file is preserved untouched, images and all.

One question, whole project

Documents and recordings, one evidence layer.

Put the contract next to the negotiation call. Project Q&A searches every ready document and recording with a single question - document claims cite pages, recording claims cite timestamps, and new sources join the scope automatically as they finish processing.

The same proof discipline everywhere

One evidence layer. Four kinds of citation.

Citesvue applies one rule to everything it ingests: no claim without a source you can open. Documents are the newest surface that rule covers.

Documents
cite the page and section

Contracts, specs, policies, research - answered with page-level proof.

Recordings
cite the spoken moment

Quote, speaker, and timestamp - click to replay the exact second.

Explore →
Screens
cite the visible frame

On-screen text and error states extracted from every keyframe.

Explore →
Artifacts
carry their provenance

Bugs, decisions, and requirements that arrive with their evidence attached.

Explore →
Formats and fair limits

Every plan processes documents. Free included.

One document-page allowance covers every format. PDF (OCR) pages are the expensive subset within it: a PDF page consumes one of each, so PDFs are capped by the smaller number and Word or Markdown can spend the rest. The numbers below are the numbers the product enforces.

PlanDocument pages / mo (all formats)PDF (OCR) pages / mo (within that)Documents stored
Free1002510
Pro1000200100
Team2000 / seat500 / seat150 / seat
Business5000 / seat1500 / seat300 / seat
EnterpriseUnlimited / seatUnlimited / seatUnlimited / seat

Evidence Q&A questions are one monthly pool shared across recordings and documents. Full comparison on the pricing page.

Where it earns its keep

Built for documents that get audited.

Anywhere “where does it say that?” is a real question, page-cited answers pay for themselves.

Scenario · 01Legal

Contract review

Counsel asks for every obligation, penalty, and renewal trigger across a vendor agreement - and gets clause-level citations back.

Outcome · Review notes with page references, not highlights from memory.
Scenario · 02Compliance

Policy attestation

A compliance lead verifies that the backup and residency standard actually says what the auditor was told.

Outcome · Audit answers backed by the governing document.
Scenario · 03PM

Spec vs. reality

A PM puts the PRD next to the UAT recording and asks which requirements were never demonstrated.

Outcome · A cited gap list across document and video.
Scenario · 04Research

Report synthesis

A researcher asks one question across eight uploaded reports and gets a per-source citation trail.

Outcome · Cross-document synthesis with page-level sources.
Limits, on purpose

What Document Q&A will not do.

  • No answers

    without a page citation to open.

  • No guessing

    when the documents do not cover it - it says so.

  • No training

    on your documents. Ever.

  • No stale scope

    answers always run against the current ready sources.

Common questions

What teams ask about Document Q&A.

  • Yes. PDF pages are read with optical character recognition, so scans, exports, and image-heavy layouts all become searchable, citable text. Each PDF page consumes one document page and one PDF (OCR) page from your plan.
  • PDF up to 25 MB, Word (.docx) up to 5 MB, and Markdown up to 500 KB, in batches of up to ten files. Password-protected PDFs need the password removed before upload.
  • PDFs use their real page count. Word documents use content-based page-equivalents, and Markdown counts one page-equivalent per 3,000 characters. Your billing page shows exactly what was consumed.
  • Yes. Project Q&A searches every ready document and recording in a project with one question, and each citation stays attached to its original source - a page for documents, a timestamp for recordings.
  • No. Customer documents, recordings, and everything derived from them are never used for model training. Ever.
  • The original file, its preview, and its entire search index are removed together, and the stored-document slot frees up immediately.
Closing argument

Your next recording could be your most valuable asset.

Or it could sit in a Drive folder nobody opens again. The difference is whether it has citations attached.

  • SetupOne drag-and-drop upload, or send the notetaker. No plugins.
  • First insightCited Q&A on a 60-min recording in under 6 minutes.
  • Cancel anytimeFull data export, full right to erasure.