Upload the questionnaire your client sent you: the AI extracts the questions one by one, proposes answers with what you already have, and marks the gaps instead of making anything up.
Your client's ESG questionnaire consists of 40 questions that you have already answered three times this year, with different wording and in another format.
aiDossier allows you to upload the questionnaire exactly as you received it —PDF, DOCX, or TXT, or by pasting the text if it was sent to you in Excel or on a portal— and breaks it down into questions. The AI reads the document (even if it's a scanned PDF, which it resolves through vision), extracts all the questions in the order of the original, respecting its wording and language, and provides you with a ready-to-work list. If the extraction fails, you can manually add questions and continue.
It then proposes an answer to each question by cross-referencing four sources you already have within the app: your validated answer bank, your company profile (business name, tax ID, industry code, employees, revenue, work centers, preventive modality, agreement), your supplier document file with the actual status of each certificate, and your carbon footprint from the last closed fiscal year. Each proposal indicates its basis, and when it relies on a specific document, it cites the supporting document from the file.
What it doesn't know, it doesn't fill in. If there isn't enough basis, the question is marked as a gap, with the reason: what is missing and which document or data would resolve it. And if the answer was based on an expired certificate, it automatically becomes a gap as well, because sending a response backed by an expired document is worse than not sending it at all.
The matching with the answer bank distinguishes certain from doubtful. When an entry answers exactly the same, it reuses it. When it only resembles, it does not use it: it shows you as “similar to…” and you decide whether it is the same question or not —“diversity policy” and “equality plan” are not the same, and the AI does not assume so—. Each validated answer enters the bank with its topic and tags, marked as expirable if it cites date-sensitive data, so the next questionnaire arrives partially filled.
The questionnaire, just as you received it
Upload the PDF, DOCX, or TXT, or paste the text directly if your client uses their own Excel or portal. Scanned PDFs are read through vision. The AI extracts all the questions in the order of the document and maintains its language: if the questionnaire comes in English, the answers will be in English.
Answers with your data, not generic data
Each proposal is built with your company profile, your supplier document file, your closed carbon footprint, and your previous answers. No filler paragraphs: your numbers, with their source indicated in 'what it's based on'.
Gaps are marked, not disguised
If there is no basis to answer, the question remains as a gap with the reason and a hint of what would resolve it. And if the supporting document is expired, it becomes a gap even if the AI had text to offer. The final judgment is always yours.
Answer bank that grows on its own
Validating an answer saves it in your bank with its topic, tags, and supporting document, marked as expirable if it cites dated information. You can search it, edit the canonical answer, see how many times it has been used, and filter those that need reviewing.
Secure or doubtful matching, never forced
If an entry from the bank answers exactly the same, it is reused. If it only resembles, it shows it separately and you decide: 'yes, use that' or 'no, it's different'. The false friends of ESG language do not slip into the answer by mistake.
The batch runs in the background while you advance
Launch the proposal for all pending questions and continue validating as they arrive: the work runs in the background and you can leave the screen. You will see the progress of validated answers and the gaps that remain to be filled.
Output ready for delivery
Export the questionnaire as a Q&A PDF with your data and the supporting document for each one, or copy all the text to paste it into the client's portal. If there are gaps, it will alert you before exporting.
In the same place as the documents that answer it
The assistant does not exist separately: it reads the supplier qualification file and the carbon footprint you already have in aiDossier, so the answers and the supporting documents never get out of sync.

Create your free account, upload the pending questionnaire, and see how many questions the AI can answer with what you already have. What’s missing will be indicated instead of being made up.