SI² — SuperIntelligence for Science
QualityAI
AI anomaly detection for procurement that flags items incorrectly classified as quality-insignificant before they slip through.
Research question
Quality experts cannot review every item passing through procurement. Items misclassified as quality-insignificant escape review, creating downstream safety and compliance risks.
Research approach
Applies AI-based anomaly detection to procurement item records, scoring each item against learned patterns of quality-significant behavior. Flagged items are routed to an expert review queue with confidence scores and explanations.
