Book a demo
AI, grounded in your own data

AI drafts. People decide.

Questionnaire answers drafted from your own controls and policies, and an assistant that reads your live program. Everything it says is cited, and a person signs it off. This page is a conversation: scroll to ask what buyers ask us.

Book a demo Watch the tour
Ask CompliLensRead-only
Answers cite your own records. Restricted data is excluded.
Can it fill in a customer's security questionnaire?
CompliLens

A helper that drafts answers to the long security questionnaires customers send.

Questionnaires can take days. Auto-fill turns hours of searching into minutes of reviewing.

How it works
  1. Click New Session and name it after the customer.
  2. Paste the questions (one per line).
  3. Click Generate. For each question, CompliLens searches your own controls, evidence, published policies, context, objectives, drills, access reviews and certifications, then drafts an answer using only what it found.
  4. Each answer shows citations: the exact sources used. If nothing relevant exists, it says so instead of inventing an answer.
  5. Material related to Restricted data is excluded automatically.
  6. A person edits the answer, then clicks Mark Reviewed.
  7. Download the finished questionnaire as a PDF.
Show me.
CompliLens

Pick a question. Each draft cites the records it came from, and anything restricted is excluded.

● Draft · needs review
Can I just ask about our own program?
CompliLens

A chat box where you ask questions about your own compliance data in everyday language.

Instead of clicking through many pages, just ask.

How it works
  1. Open AI Assistant and type a question, for example "How many risks are overdue for review?" or "Which controls have no evidence attached?"
  2. The assistant picks the right data lookups, reads your live data, and writes an answer that streams onto the screen.
  3. Small chips above the answer show which lookups it used, so the answer is transparent.
  4. It can answer questions about risks, NCRs, vendors, assets, incidents, training, certifications, framework gaps, overdue CAPAs, policy acknowledgement gaps, and more.
Try it on something real.
CompliLens

Choose a question. I show which lookups I used before I answer.

What stops it from going wrong?
CompliLens

Six guardrails, built in rather than bolted on.

Read-only assistantIt looks things up. It cannot create, edit or delete a record.
Citations on every draftSee exactly which control, policy or evidence was used.
Sensitivity guardRecords matching restricted data categories are excluded first.
Human reviewDrafts become final only when a person marks them reviewed.
Kill switch and spend capAdmins can turn AI off, and a monthly cap stops runaway usage.
Usage logEvery AI call is logged per organisation.
Is CompliLens itself on your AI register?
CompliLens

Yes. The assistant is listed in our own AI system inventory as limited risk, with notes on human oversight. Any future version that can take actions would need its own review.