A pharmaceutical client with a data-intensive workload, including a large data lake, faced annual cloud costs in the millions.
The major expenses were associated with S3 storage fees and Redshift’s dense compute and storage costs, billed by the second.
With rapid data growth spurred by AI training investments, the financial sustainability of their current path was in jeopardy.
Task
The company required significant reductions in data management costs—aiming for more than 50% savings.
They sought a solution that would offer greater control over their data and cost structure.
Action
Co-location emerged as the best strategic option, allowing full data control and significantly lower storage costs.
After creating a budget for necessary hardware and co-location expenses, we proceeded to set up and migrate their workload from Redshift and S3 to a private hardware setup.
Technologies selected for this transition included Hadoop and Apache Drill to replace Redshift and S3 functionalities.
As part of our service, we took care of management and upkeep of all the servers and storage appliances.
Result
The migration strategy involved:
Transferring data from S3 using the AWS Snowball appliance.
Configuring S3 data for cold storage in S3 Glacier to further reduce costs.
Establishing a proficient data management system at the co-located data center.
Optimizing the new Hadoop and Apache Drill setup for performance.
Implementing nightly backups to Amazon Glacier.
We achieved more than 70% in cost savings over three years while tripling the client’s data processing capabilities.
Early Access Waitlist
Fifteen minutes could save you 50% or more on cloud costs.