What is the best way to download a 50-100MB CSV file from S3 and serve it to the web client on request?

Question

Ruby on Rails

It is planned to upload rather large CSV files (about 50-100MB) to AWS S3 hosting, to be subsequently downloaded on demand (one file per request) from the web admin panel running on Heroku hosting. The functionality for downloading a file must also be provided via the HTTP API.
Files on S3 are not accessible from the outside, which is why it is necessary to first download to a temporary tmp folder on the application instance on Heroku hosting, and then issue the result to a geographically remote web client.
Is this workflow normal when working with Heroku? What alternative to this scheme can you recommend?