EnerGaze
Advanced Energy Program
Evaluation Toolkit
Powered by
EcoMetricx
Comprehensive Python package for evaluating energy efficiency programs using advanced statistical models. Implements methodologies from the Uniform Methods Project (UMP) with robust tools for energy savings estimation, treatment effect analysis, and program impact assessment.
Key Capabilities
For Utilities
Drive program insights, customer segmentation, and load impact modeling with minimal overhead.
Automated UMP Behavioral Model Execution
Run standardized load impact models with configurable baselines, control groups, and event filters.
Peak Load Reduction & TOU Impact Analysis
Quantify demand response performance, rate design effects, and behavioral savings across time slices.
Customer Segmentation Engine
Cluster customers by usage patterns, responsiveness, or behavioral archetypes using unsupervised ML.
Data Pipeline Integration
Connect directly to utility AMI databases or S3 buckets; support for hourly, 15-min, or 5-min intervals. Can down sample when needed.
Interactive Dashboards & API Access
Visualize savings, load shapes, and confidence intervals; export results.
Run standardized load impact models with configurable baselines, control groups, and event filters.
Peak Load Reduction & TOU Impact Analysis
Quantify demand response performance, rate design effects, and behavioral savings across time slices.
Customer Segmentation Engine
Cluster customers by usage patterns, responsiveness, or behavioral archetypes using unsupervised ML.
Data Pipeline Integration
Connect directly to utility AMI databases or S3 buckets; support for hourly, 15-min, or 5-min intervals. Can down sample when needed.
Interactive Dashboards & API Access
Visualize savings, load shapes, and confidence intervals; export results.
For HERs Implementors
Validate home energy ratings with real-world usage and behavioral overlays.
Pre/Post Retrofit Analysis
Compare AMI-derived usage before and after upgrades, controlling for weather and occupancy.
Behavioral Attribution Layer
Separate technical savings from behavioral changes using UMP-compliant methods.
Batch Processing for Portfolio-Level Insights
Analyze hundreds of homes simultaneously; flag outliers or underperformers.
Compare AMI-derived usage before and after upgrades, controlling for weather and occupancy.
Behavioral Attribution Layer
Separate technical savings from behavioral changes using UMP-compliant methods.
Batch Processing for Portfolio-Level Insights
Analyze hundreds of homes simultaneously; flag outliers or underperformers.
For Evaluators
Deliver defensible, reproducible, and transparent impact assessments.
Statistical Toolkit for Confidence & Attribution
Includes regression-based models, matched pairs, synthetic controls, and bootstrapping.
Custom Model Plug-In Architecture
Easily integrate proprietary or experimental models alongside UMP standards.
Scenario Testing & Sensitivity Analysis
Run counterfactuals, alternate baselines, or weather-normalized comparisons.
Model Audit Trail & Reproducibility Engine
Every run is logged with parameters, assumptions, and versioning for defensibility.
Includes regression-based models, matched pairs, synthetic controls, and bootstrapping.
Custom Model Plug-In Architecture
Easily integrate proprietary or experimental models alongside UMP standards.
Scenario Testing & Sensitivity Analysis
Run counterfactuals, alternate baselines, or weather-normalized comparisons.
Model Audit Trail & Reproducibility Engine
Every run is logged with parameters, assumptions, and versioning for defensibility.