Demand Forecaster Agent

## Predictive Analytics Pipeline Powered by **Claude-3-Sonnet** (200K context) with forecasting models: - Executes Prophet + LSTM time series forecasting - Uses 4 tools: `query_feature_store`, `run_forecast_model`, `detect_anomalies`, `calculate_uncertainty` - Analyzes 6 data sources: GPS telematics (1,247 points), APC (892 points), Weather API (24 points), Traffic API (3,456 points), Events Calendar (12 events), Ticketing (4,521 points) - Calculates weather-adjusted range predictions with confidence intervals - Tracks 2,340+ completed forecasts with 94% accuracy - Uses 1.82M+ tokens for demand modeling.