Challenge
A financial data company processed large files in batches: collecting them, merging and appending data, exporting the results and storing them for the next stage.
The work was done by hand, and the jobs were large enough to need serious computing power for a few hours and none for the rest of the week.
Approach
We automated the process in a cloud environment owned by the client, using computing power that is switched on for the job and off afterwards.
- Agree the requirements and data sources with the client's team.
- List every dependency and configuration the process needed.
- Automate collection, merging, export and storage.
- Run test files through the process and check the results.
- Train a technical user to start the jobs, and stay on to manage the environment.
Result
The company runs its batch processing on demand, in its own cloud account, without buying servers.
- Operating costs that follow the work.
- A repeatable process in place of manual steps.
- An environment the client owns.
Our advice
Work that needs a lot of computing power only occasionally is what the cloud is best at. Rent it by the hour and do not own it.
Related service: Cloud computing.