Let diligence ask the documents — and get answers from the room.

AI document Q&A lets the people reviewing your deal — buyers, investors, lenders, LPs — ask questions in plain language and get answers generated from the documents already in your data room. A reviewer types "what's the ARR growth quarter over quarter?" or "which agreements have change-of-control clauses?", and the AI reads the files, answers, and cites the source so the reviewer can open the underlying document and read it for themselves. It answers per-document (against one file) and cross-document (across the whole room), and it is grounded: it answers only from the room's contents, not the open web, and will not invent a number that is not in your files.

This solves the specific way diligence turns into email. A good partner meeting or first bidder call is followed by a drip of "can you also send" requests — the model, then the customer detail, then the cap table, one at a time, each round trip costing days while momentum bleeds out. A room where every reviewer can ask and get an answer from the documents collapses that serial back-and-forth into parallel self-serve access. In practice, investors and buyers self-serve 80-90% of diligence questions this way — so the founder or advisor stops answering the same working-capital question six times and spends their time on the handful of judgment questions that actually need a human.

AI document Q&A is the machine answering from documents. It is distinct from structured, human-to-human deal Q&A — the moderated Q&A workflow where bidders submit questions, the AI drafts a cited answer, and your team reviews, routes, and approves before anything reaches a specific counterparty. That human-workflow layer — with role-based routing and a multi-role approval chain (Advanced Q&A) — sits on the higher tiers. This page is about the AI kind: reviewers ask the documents, and the documents answer. AI document Q&A is on Peony Business at $30 per admin per month; the Data Room plan at $52/admin/month adds 5x AI Q&A usage, dynamic watermarking, and granular permissions on top.

For a moderated, human-approved Q&A process — where bidders submit questions, your team reviews and edits each answer, and nothing reaches a counterparty until you approve it — use the Smart Q&A workflow on the Data Room plan, or the structured Advanced Q&A module (role-based routing, multi-role approval) on Deal Team. AI document Q&A answers the reviewer directly; the workflow tiers put a human in the loop before release.

For representations you are legally standing behind — a warranty in a purchase agreement, a figure in a disclosure schedule — the answer belongs in the documents and gets confirmed by you and counsel. Treat AI document Q&A as a first-pass diligence assistant that points reviewers to the source, not as the authoritative statement itself.

For questions the documents can't answer — anything not in the room. The AI answers only from what you have uploaded; it will not guess from the open web. If the answer isn't in a file, upload the file, or answer it directly.

For connecting your own external LLM — routing ChatGPT, Claude, or Gemini directly against your room with every AI query logged in the same audit trail as a human viewer — that is a Peony Enterprise capability. For the default deal case, the built-in Q&A is the one that stays inside the trust boundary without extra setup.