How Israel’s Ministry of Finance transformed tax revenue forecasting with AI

This article describes how Israel's Ministry of Finance transformed tax revenue forecasting using AI and AWS services, achieving significant accuracy improvements and cost savings.

• Deployed automated solution evaluating multiple AI models alongside traditional statistical approaches for tax forecasting
• Ingests 20+ years of historical tax data and dozens of economic indicators to improve predictions
• Integrated Amazon QuickSight with natural language capabilities in Hebrew for self-service analytics
• Reduced forecast errors by up to 40%, saving approximately $10 million monthly in optimized debt issuance
• Improved organizational memory by making forecasting processes more structured, reproducible, and transparent
• Reduced analysis time from days to hours and economic shifts now reflected in forecasts within weeks instead of 8 months

The Ministry's approach demonstrates how combining AI tools with professional judgment, broader data sources, and clear business objectives strengthens government financial forecasting capabilities.