Any Tips for Downloading Petabytes from Cloud to Local

One of my colleagues request to download a petabytes from multiple cloud locations to in-house server room. (what?)

At first, we consider to use the AWS Snowball. but requester's team are using multiple cloud providers and regions. Because of growing data every day, we can't afford to call the AWS Snowball/Snowmobile every months/quarters.

Next idea is using AWS Direct Connect (and other providers') and syncing to local object storage. (maybe ceph?)

Is there any suggestion/idea/solution for solving this wisely?