One input - your company name. What happens next is a search of your own public pages, a deterministic classification against the CRA's Annex III and IV, and a report that says what the evidence supports and stops where the evidence stops.
Type your company name - the only input. CRASPACE searches for your site, reads your product pages, and pulls out the products with digital elements. No questionnaire, no data entry.
Every product is matched against the CRA's Annex III & IV by a versioned, auditable rulebook. Grounded in public evidence - AI structures what it finds, it never sets the verdict.
A plain-English scope verdict with a confidence indicator, a 0–100 risk score, and which of the 11 CRA checkpoints your public evidence covers - plus the open questions that would change the answer.
The report identifies; it does not prescribe. It gives you scope, a class per product, a readiness picture and the questions still open - not a remediation roadmap, which depends on your architecture, your suppliers and evidence no public page carries. Where the eleven checkpoints have no public evidence at all it reports them as not scored, rather than printing a number built from silence.
Scope, class and score come from a deterministic, auditable rulebook (v0.5.1) that mirrors the CRA's annexes - not a chatbot's opinion. Same inputs, same result, every time.
The AI reads public sources and structures what it finds - it never sets the verdict. Every product is traced to a citation you can open.
We are not a notified body and we do not issue CRA certificates. Where a product's class requires a notified-body assessment or an EU cybersecurity certificate, we say so and hand off to an accredited body rather than imply we can complete it.
CRASPACE is an indicative scoping aid - not legal advice and not a formal conformity assessment. Confirm with qualified advisors before acting. How the verdict is decided →
Indicative assessment - not legal advice and not a conformity assessment.