SmartMeterResearch
Smart Meter Disaggregation + Energy Savings Recommender application using Postgres, Redshift, Graphlab (Online Bayesian Checkpoint Model), Kinesis, and Spark. Started as part of GalvanizeU Data Engineering course final project and extended on personal time to experiment with scalable machine learning on sensor data.
Update (5-11-2016): Adding in experiments with demand forecasting using Recurrent Neural Networks via Keras