All documentation

AI & Analytics

Wyverix is an MCP-native AI compliance platform. The AI answers questions, explains risk, catches anomalies and — with your approval — takes action.

Package: Business (T4) and above; advanced AI Premium (T5) admin · dpo · analyst

Assistants

  • DPO Copilot — chat grounded (RAG) on KVKK text + Kurul decisions; protected against prompt injection.
  • DPO Cockpit — a unified command console for the DPO.
  • Multi-Agent Consultation — Legal / Technical / Precedent agents debate via a live (SSE) stream and produce a joint report.
  • Knowledge Vault & Email/Voice Briefing assistants.

Trust & explainability

  • LLM Guard — injection, hallucination and PII redaction on LLM input/output (dashboard + playground).
  • Explainable Risk (SHAP) — shows how much each factor contributes to a score.
  • UBA — user behavior anomaly; insider threat detection.
  • Anomaly & Predictive Breach — proactive alerts via ML.

Advanced capabilities

  • Multimodal PII — OCR and face detection in images/audio/handwriting.
  • Autonomous Compliance Agent — plan → approve → execute (human-approved).
  • Privacy Budget — differential privacy budget tracking.
  • MCP Server — securely exposes Wyverix tools, with tokens, to tools like Claude Desktop / Cursor.

This document is informational. Feature scope may vary by license package, and screen flows may change between releases.

Try this feature live

Test Wyverix with your own data, from a single panel.

Request a Free Demo