TATA POWER

Causal AI & Load Platform

⚡ Local Model: gemma-4-e2b
Database Online
Daily Mean Load: -- MW
Peak Load: -- MW
Mean Temp: -- °C
🌡️
Temperature Effect
+7.60 MW / °C
95% CI: [+6.46, +8.74]
Marginal demand shift per 1 degree change in temperature (°C)
💧
Humidity Effect
+0.85 MW / %
95% CI: [+0.72, +0.97]
Marginal demand shift per 1% change in relative humidity
🪔
Festival Impact
+7.28 MW
95% CI: [+6.19, +8.38]

24-Hour Heterogeneous Treatment Effect (HTE)

Hourly causal sensitivity curves for selected date (2024-01-30)

2024-01-30

Load Curve & DTW Nearest Neighbors

Complete 15-minute load profile compared against historical days using normalized Dynamic Time Warping

Top Similar Historical Days (DTW)

Searching nearest historical days...

Machine Learning Feature Attributions (SHAP)

Global predictive influence

Predictive XAI

Safe Parameter-to-SQL Console

Quick Samples:
Generated Parameterized SQL: SELECT ROUND(AVG(wdLoad), 2) as aggregate_value FROM load_data WHERE season = 'Summer' AND festival = 1
Calculated Result: 384.25 MW
Matched Intervals: 96 intervals