Meridian
Supply Chain Risk Intelligence Platform
Problem
Supply chain disruption signals arrive as multi-source noise. Operators need scored risk, graph context, and explanations—not another offline notebook forecast.
Approach
Combined GDELT and ACLED geopolitical signals, AIS vessel tracking, and weather into a live Neo4j knowledge graph with 250+ risk relationships and 10K+ supplier nodes.
Scored risk with XGBoost and SHAP explanations, plus Monte Carlo simulation for disruption scenarios.
Served 14 authenticated FastAPI endpoints at 500–700 req/s and validated the model against 50+ historical disruption events including the 2021 Suez blockage and Red Sea shipping crisis.
Results
10K+ suppliers. 250+ risk edges. 500–700 req/s
- Operational risk scoring over live multi-source supplier and disruption data
- Explainable XGBoost outputs with SHAP and Monte Carlo simulation support
- Validated against 50+ real historical disruption events
Stack
Learnings
Risk products earn trust when scores are explainable and the data path is observable. Graph context plus SHAP beats a black-box probability every time.