Document Q&A for analysts

Ask your documents.
Get the page number back.

Upload PDFs into a chat. Every answer is drawn from those files and cites the page it came from, so you can check it in a second rather than trust it.

182pages per report,
indexed in a minute
5reranked passages
behind each answer
0answers from outside
your own files
Greggs FY25 review · 1 document ready
What was underlying operating profit, and how did it move?

Underlying operating profit was £187.5 million, down 4.0% on 2024 when it was £195.3 million. The margin was 8.7%, against 9.7% the year before.p.59

greggs-annual-report-2025.pdf · page 59
Underlying operating profit 187.5 195.3 -4.0% Underlying operating profit margin 8.7% 9.7%
Searching documents
SEARCH RERANK ×5 CITE p.59
How it works

Three steps, and nothing hidden

01

Start a chat, add PDFs

Each chat holds its own documents: annual reports, board packs, contracts. A chat only ever searches the files you put in it.

02

Ask in plain language

The agent decides whether it needs to search. When it does, it queries by meaning and by keyword at once, then a local reranker keeps the best five passages.

03

Check the source

Every claim carries a page reference. Hover it and the original text appears, exactly as it sits in the PDF.

It will tell you what the filing says, not what to do about it

Ask for a recommendation and Vouch declines, then shows the figures the documents actually report. The judgement stays with you.

No answer without a citation
Scans that cannot be read are flagged, not silently skipped
Retrieval steps stay visible while the agent works
Should I buy the stock?

I can tell you what your documents say, but not whether to buy or sell.

Here is what they report instead: sales £2,151.2m, up 6.8%. Underlying profit before tax £171.9m, down 9.4%.p.59

Security

Scoped to your tenant, and to one chat

Every query is filtered

Retrieval carries your tenant and chat identifiers, so one chat cannot reach another's files.

Reranking runs locally

Candidate passages are scored in your own environment before anything is sent to a model.

Documents stay put

Files are indexed once into your own vector store and are never used for training.

Auditable answers

Each response records the passages it used, with file name and page.

Put a report in and ask it something

First chat is free. No configuration, no connectors to set up.