Descartes Labs, a well-funded startup based in New Mexico, provides companies with geospatial data and the tools to analyze them to help them make business decisions. Today, the company announced the launch of its Descartes Labs platform, which promises to merge their data with all the tools that data scientists – even with no background in analyzing this type of information – need to analyze these images and build machine learning models based on them on the data contained therein.
Descartes Labs CEO Phil Fraher, who held this position just a few months ago, told me that the company's current business often involves a lot of consulting work to get its customers up and running. These customers range from energy and mining companies to government agencies to financial service providers and agricultural companies. However, many do not have the internal know-how to use the data provided by Descartes Labs immediately.
“For the most part, we still need to evangelize how geospatial data can be used to solve business problems. That is why many of our customers rely on us for advice, ”said Fraher. "But what's really interesting is that even some of our existing customers are now hiring more early adopters, more business and analytics teams, and data scientists to focus on geospatial data. So what's really exciting about this launch is that we are put our platform tool in the hands of those who can now do their own work. "
This new platform gives these customers access to the tools and data that the Descartes Labs team uses in many ways, allowing them to work with the company to solve their problems and use the new modeling tools to find solutions for their individual companies create.
"Previously, a data science team from a company interested in this type of analysis also needed to know how to handle very large or petabyte-sized Earth observation records," said Fraher. "These are very unique and specific skills, and because of this type of entry barrier, adoption of some of these technologies and data sources has been slow."
To enable more companies to get started with this data (and become Descartes Labs customers), the company relies on the industry-standard tools with hosted Jupyter notebooks, Python support and a number of APIs. It also includes tools to transform and clean up incoming data from Descartes partners so that it can be used by data scientists.
"It's not just a simple ETL-like data processing pipeline," said Descartes Labs chief engineer Sam Skillman. "It is something where we have to combine very detailed data science, remote sensing, and extensive computational functions to bring all this data in a way that normalizes it and makes it ready for analysis."
All of these analyzes are of course carried out in the cloud.
The new platform is now available for companies who want to try it out.