I am a noob developing a native app using electronjs and azure real-time speech translation sdk. The azure speech sdk requires that connection to this service be configured on each app via the Translation Config object. To set up a connection requires entering the azure speech key and region for each new session.

The strategy I have found in SO and code LLMs–if I understand correctly–is for the app to request my encrypted credentials from an express js server for every new session. The user app will decrypt using a private key and configure the translation object.

Is this a solid strategy? If not, which techniques are industry/Apple dev standards for providing my users with the speech translation using my credentials?