Being one of the largest cloud providers in the world, AWS has multiple ways of storing data within its cloud. In this article we will look at two AI Accelerators that will allow you to source data from S3 or Athena, build and evaluate an ML model using DataRobot, and send predictions from that model back to S3.
Each AI Accelerator will perform the following steps to help you integrate DataRobot with your data in AWS.
Import data for training:
In the AI Accelerator integrating with S3, you will be able to take data in the parquet file format, assemble it and upload to DataRobot's AI Catalog.
In the AI Accelerator integration with Athena, you will create a jdbc datasource within DataRobot to connect to Athena and then pull data in via a sql query.
Building and evaluating a model: Using the DataRobot Python API, you will have DataRobot build close to 50 different machine learning models while also evaluating how those models perform on this dataset