Create user-based evaluation screen with server integration

[qkb2]

Our app needs to make a request to the server for an evaluation while providing necessary data (amount of data to send is to be determined). The app must necessarily anticipate the response after a long time. It may for example listen and wait for any response after initial ACK from the server. When the info. arrives, evaluation metrics need to be displayed to a user on the eval. screen (maybe as long a the server is processing, show some "Waiting for evaluation..." to a user).

[plebcio]

The waiting part shouldn't stop the user from writing more test. Ideally, the newly written data should be aggregated together with the previously sent data. Then, we can send the new data and do inference again.

[qkb2]

We can theoretically just keep sending longer and longer data to a server, but then I suppose the server arch. must be somewhat more sophisticated.

[plebcio]

Unless we do data processing on device. The server would just get another request with more data. This way the "state" is stored on the device - we keep a simple stateless server api.

[qkb2]

Ideas for server inner workings can go here: wylupek/Keystroke-GNN#7 (comment).

[qkb2]

IMO server implementation should be done first - app client should be created accordingly after.