Primary
GatewayOverviewDecision RoomExecutive OverviewExecutive AwarenessCurrent SituationAI AnalystLoading governed environmental evidence…
WP-AI-009 · Governed Open-Data Intelligence
Natural-language analysis over controlled open-data records, with mandatory citations, explicit data cutoff, geographic scope, confidence, limitations, source-state visibility and a complete audit trail.
Provider state
AI Assistance Available
GROQ CHAT COMPLETIONS · openai/gpt-oss-120b
Governed query
Retrieval is limited to governed UNIOREVIS records. Retrieved text is treated as untrusted data, citations are mandatory, and the AI layer has no authority to modify platform state.
Required operational questions
AI-Assisted Analysis
No analytical output yet
Each answer will show the source records, cutoff, geographic scope, confidence, limitations and source state.
Authority boundary
Answers are built from persisted Canonical Observations, Raw Evidence references, ReliefWeb context, source-health records and existing protected analytical outputs within the active platform scope.
Each execution stores the provider or fallback service, model version, prompt version, input references, source cutoff, scope, generated text, confidence, limitations, citations and review status.
The AI route cannot create observations, modify evidence, activate Signals, set Severity or Velocity, change the Risk Clock, create authoritative Impact Windows, publish Live Outlooks or approve reports.