all of my experience with McKinsey has been an utter waste. Worked with multiple contractors on multiple backend database projects. They work like a walking ad for AWS services but dont actually know shit. We have a 10+TB dataset (growing by gigs each day) that is spread across a few postgres servers. These servers are decent but foreseen as bottle neck in coming years.

This guys suggestion is to put our data in redis b/c it's fast and in memory. Looking up hosted REDIS solutions, (redislabs, we use for other smaller cache redis things). To put 1 TB of data in standard redis server would be ~50k/month :p