Today
Customers and OEMs send ever longer information security questionnaires: 300 rows in their own spreadsheet on access rights, encryption, backups, suppliers and emergency plans. The answers already exist, spread across policies, audits and earlier questionnaires, but a small IT team gathers them again every time. It takes weeks, and the same point is answered a little differently for every customer.
An example
Before awarding a framework order for AB brackets in Ti6Al4V, a satellite maker asks Rennhardt Aerospace to complete a questionnaire with 284 questions, 40 of them on protecting export-controlled drawings. The agent answers 231 from the IT security policy, the statement of applicability, the last internal audit and two earlier questionnaires. It flags 38 because the evidence is older than a year, such as the last restore test, and 15 as gaps, including missing multi-factor sign-in on two computers in production. The head of IT and the information security officer review the 53 questions in two days instead of three weeks and return the questionnaire in the customer's spreadsheet.
Your data, wherever it lives
Your answers sit in IT policies and concepts in the wiki and the DMS, in audit and test reports as PDFs, in tickets and in the spreadsheets of earlier questionnaires that went back and forth by e-mail. demi reads, understands and links them, with your rights. The AI fills in the next questionnaire in hours instead of weeks, every answer with its source, and your information security team releases what goes out.
Where the data lives today
- Office files
- DMS & archive
- Wiki & intranet
- SharePoint
- E-mails
- PDFs
- Tickets
How demi solves it
Three layers in this order: connect the data first, then the rights, then the AI.
- 1 · ConnectData Layer
The case needs data from the DMS, the wiki, spreadsheets, e-mail and the ticket system. demi replaces nothing and links your IT policies, the statement of applicability, audit and test reports, evidence from tickets and every answer released before into one knowledge base in which each statement has its proof and its date.
- 2 · PermitPermission Layer
Security documents and answers are confidential in themselves: only the IT, information security and quality team sees them. What goes to the customer is released by the information security officer, and internal documents such as network diagrams leave the company only with their own release. Every release is logged.
- 3 · ActIntelligence Layer
The agent assigns each question to a topic, finds the matching rule and the current evidence and writes the answer in the customer's format, each with its source. If current evidence is missing or two earlier answers contradict each other, it flags the question instead of guessing. What you do not meet appears as a gap, never as a yes.
How it runs
- 1
The questionnaire arrives as a spreadsheet, a PDF form or through a customer portal.
- 2
The agent assigns each question to a topic such as access rights, encryption, backup or emergency plan.
- 3
It answers from policies, evidence and earlier released answers, each with its source.
- 4
Questions without current evidence and real gaps go, flagged, to IT and information security.
- 5
Information security checks, completes and releases, and the questionnaire goes back in the customer's format.
What changes
Questionnaires back in days instead of weeks
The same question answered the same way for every customer
Real gaps visible before a customer finds them
What does it bring you?
Check it with your own numbers
Every starting value is an assumption you can change. The calculation sits right below, step by step.
Your numbers
Result
+216 hfreed per year
Assumptions and calculation
demi answers each questionnaire from policies and evidence; your IT checks only the flagged questions.
15 h → 9 h per case
How we calculate
3 security questionnaires × 15 h × 40% × 1.0 (manual) × 12 months = 216 h
Value per year, before demi costs
Freed per year+216 h
Assumption: labor cost of $48.62 per hour in US manufacturing in June 2026 (BLS), plus workplace, IT and overhead.
= Value per year, before demi costs
+$14,040
gained per year
216 h ÷ 1,760 h per role = 0.1 full-time roles
An estimate, not a promise. The starting values are assumptions for an industrial company with 150 to 500 people.
Combine with more cases and get the packageQuestions about this use case
Which security questionnaires can demi answer?
Any of them, whether based on ISO 27001, TISAX, NIS2 or a customer's own catalog. It can arrive as a spreadsheet, a Word file, a PDF form or in the customer portal; the answers go back in the same format. The agent also prepares follow-up questions from the customer on individual answers, with sources.
Will demi write yes when we do not meet a requirement?
No. The agent answers only from documented rules and evidence. If evidence is missing or outdated, it flags the question, and what you do not meet appears as a gap with a note on what is missing. How you present the gap to the customer is your decision.
How do our security documents stay protected?
Policies, network diagrams and answers are visible only to the team allowed to see them, and the agent works with exactly those rights. The customer receives only what you release. The agent never attaches an internal document unasked; at most it suggests one.
Does demi learn from earlier questionnaires?
Yes. Every released answer is stored with question, date, customer and evidence. For the next questionnaire the agent proposes it again and checks whether the evidence is still current. If a policy changes, it shows which stored answers need updating.
