At Blue Sky Analytics, we wanted to empower decision-makers across sectors with climate intelligence by leveraging satellite data, AI and the cloud. The idea was to harness the true potential of satellite data as a result of the booming upstream space market to draw deeper and decision-ready environmental insights.

However, from an implementation standpoint, we realized early on in the company’s journey that we were solving an engineering problem as much as we were solving a data problem. To process and deliver insights from satellite data, substantial work was required to build a technology infrastructure that could efficiently handle such large amounts of data, and deploy and maintain complex machine learning algorithms, while offering it to enterprises in an accessible way through APIs.

To this effect, we set our minds to building a revolutionary new technology infrastructure that would significantly cut down the time of deploying each new dataset while continuing to evolve the underlying models as new sources come online & improved research is available.

Instead of treating each dataset like an individual product, we decided to build a completely automated, scalable, and reusable infrastructure that could ingest data from various sources, including satellites, run them through various machine learning algorithms, and deliver the resulting datasets via APIs. Hence, making our tech infrastructure our core value proposition and differentiating us from other data companies.

This infrastructure is what we call our “Geospatial Data Refinery“. Similar to an oil refinery, it takes in and analyses terabytes of raw data from satellites, on-ground monitoring devices, and any other credible ancillary public sources. As a first step, the data is cleaned, standardized, pre-processed, and stored on the cloud. After this, we run it on our proprietary algorithms and disseminate intelligent data to various stakeholders via APIs and our visualization platform, SpaceTime™.