This is pretty amazing.
If you look at old tech decks of mine I always tell startups to start their data collection journey on day 1 or as soon as possible. I tell them to collect and shove it in S3 and keep it for later. The idea was a data person could extract and get it into a queryable format at some point but if you don’t collect the data you are missing out on your own history.
Meaning you may not know how to derive insights but best to keep the data.
Now look what AWS has done – you can literally just query S3.
Impressive.
Amazon Redshift Spectrum – Run SQL queries directly against exabytes of data in Amazonn S3.
Dreampipe's Blog 