## TL;DR
AI predicts supply chain disruptions before they happen -- monitoring millions of signals for factory fires, port strikes, and supplier bankruptcies across multi-tier global supply chains. From Resilinc to o9 Solutions, AI supply chain risk management transforms reactive firefighting to proactive resilience.

## Core Explanation
Supply chain risk AI: (1) Event monitoring -- NLP + event extraction from global news in 50+ languages, weather forecasts, government databases. ML classifies events by severity, location, and supply chain impact; (2) Impact analysis -- given a disruption location, AI maps affected suppliers, factories, logistics routes, and downstream customers. Graph traversal on supply chain network; (3) Risk scoring -- ML assigns risk scores to suppliers based on financial health, operational performance, geographic concentration, and political stability; (4) Mitigation -- AI recommends inventory buffer levels, alternate suppliers, and rerouting options.

## Detailed Analysis
Resilinc: monitors 400+ risk event types across 100M+ data sources. Maps client supply chains to create "what-if" impact scenarios. During COVID-19, Resilinc mapped supplier dependencies in Wuhan within 24 hours. Everstream Analytics: NLP on multilingual news + weather + IoT (ship tracking). Predictive risk scoring: probability of disruption at each node. Supplier risk: Dun & Bradstreet provides financial health scores. AI combines financial + quality + delivery + geopolitical into composite supplier risk score. China+1 risk: Taiwan semiconductor supply chain concentration. AI maps global semiconductor supply chain dependencies from TSMC through component makers to end products. Multi-echelon: Tier 1 supplier visibility is common; Tier 2-3 visibility requires mapping efforts. AI can infer Tier 2-3 from public data (trade data, factory registration). Key challenge: supply chain data is proprietary -- companies reluctant to share supplier lists with AI platforms.