Kevala remains on an urgent mission to decarbonize the global energy economy using the most comprehensive data sets available on earth. I contributed to this process by managing Kevala's development of our software tools: our flagship products (Assessor suite), new product prototypes, internal tools, as well as bespoke application development work.
As a new manager, I worked to build structure around the API, UI, and data engineering teams while continuing to write and review code - JavaScript (Svelte) + Python (Django) every day.
In this time, we began ramping up data acquisition work by building out our data pipeline and teams. We also built out structure for how different types of electric analyses: electricity demand, behind the meter and front of the meter storage, photovoltaic production, electric vehicle adoption and charging, rates and billing, hosting capacity, bulk energy market prediction, energy efficiency and demand response curtailment.
I spoke about this work on the Heroku podcast Code[ish]
I joined Kevala when there were just 8 other employees. In the first year, we worked together to release a novel interface for surfacing grid data. We built automated data pipelines that fed information on tens of millions of homes, rooftops, power lines, and garages to heuristic models in Pandas and machine learning models in TensorFlow and GCP. These predictions and analyses are served through Django microservices to various front ends: auth services, mapping software, data visualization platforms, varyingly implemented in Svelte, React, and Polymer.
