Sense Labs — Home Energy Monitoring

Overview

I completed two internships at Sense Labs, a home energy monitoring startup. In the first, I built a data augmentation tool to improve the ML model that detects air conditioner energy signatures — generating synthetic training samples to address class imbalance and improve detection accuracy. In the second, I conducted a user segmentation analysis across the Sense user base, identifying behavioral clusters and presenting findings directly to the CMO and VP of Technology.

Key Features

• •Data augmentation pipeline generating synthetic AC energy signature training samples
• •Improved ML model accuracy for air conditioner detection via augmented dataset
• •User segmentation analysis across the full Sense user base
• •Findings presented directly to the CMO and VP of Technology

Outcomes

The augmented dataset measurably improved AC detection accuracy in the production ML model. The segmentation analysis was incorporated into Sense's product roadmap discussions.