Fellow head of data here. For a while, my team ran on MySQL, a shared file server we mapped as a local drive, a team server in the back closet for tasks too lar

Fellow head of data here. For a while, my team ran on MySQL, a shared file server we mapped as a local drive, a team server in the back closet for tasks too large for our laptops, a MySQL database, an SVN server, and a Wikimedia server for documentation. We're upgrading a few of those items, but it's amazing how cheaply you can run a machine learning team if you only have a few TB of structured data to analyze.