TL;DR
🧮 Deterministic = same inputs → same outputs, always, traceable to a rule version. NexCyber doesn't use an LLM to decide your compliance — we use a rule engine + your evidence.
🧠 Why deterministic matters
- 🔍 Auditable: every output is traceable to a rule version
- 🔒 Reproducible: re-running the same assessment gives the same result
- 🛡️ Trustworthy: no hallucinations, no probabilistic answers
- 📋 Defensible: in front of an auditor, you can show exactly why an obligation was flagged
🤖 Where AI does sit in NexCyber
- ✅ Captain Assistant (support chatbot) — answers your questions about the platform, never decides your compliance
- ✅ Suggestion layer — surfaces likely-applicable obligations for review, but a rule decides
- ❌ No LLM in the compliance decision path
📜 Rules versioning
Every assessment output includes:
- Rule pack version (e.g.
rules-2026.05.1) - Evidence hash
- Issuance timestamp
- Signature
If a rule pack updates, you re-run and see the diff. No silent changes.
➡️ Next
💬 Need help?
Reach out via our live chat (bottom-right) — Captain AI replies instantly, human experts within business hours.
Email support@nexcyber.eu with [P1] for Command/Strategic priority issues.
ℹ️ Disclaimer — NexCyber provides a readiness analysis, not legal advice. Final compliance may require legal review or notified body certification.
Last reviewed: 2026-06-02 · NexCyber Help Center