Verdigris is an artificial intelligence (AI)-enabled energy management platform for commercial facilities. My role at Verdigris is to work at the intersection of AI and Insights team to help bring Machine Learning algorithms into production.
o Reduced time and effort taken for development by creating a cross-team framework using Spark and Docker to access data.
o Unblocked the creation of auto labeling (critical for the company) by creating interactive device labeling dashboard.
o Pushed towards test-driven development by creating and improving existing tests, catching more bugs before release.
o Created accuracy metrics and automated plotting to help evaluate Verdigris’s energy forecasting algorithm.
o Converted ML algorithms like Classification, Forecasting, Disaggregation into products by developing backend pipelines.
o Developed CI/CD pipelines using CircleCI and Docker to build, test and deploy code to AWS Batch, ECS and Lambda.
o Customized an existing BI tool using Javascript to enable faster development, serving the needs of the fast-paced start-up.
o Developed interactive charts using Highcharts, to showcase various Machine Learning products increasing their value.