Data science requires data sets to be cataloged and indexed. The data sets are versioned and might be in CSV files in S3, a database, or another data storage system. Splitgraph allows the user to query this data catalog like it is a Postgres database, routing queries to any data set across your catalog. Artjoms