Question 1

Is this for practice or for use in a live interview?

Accepted Answer

CuePreem is built for preparation: mock interviews, resume work, and rehearsing your own stories. If you use it in a real interview, you’re responsible for following that employer’s rules on AI tools. The default Assistive overlay stays visible in any screen-share.

Question 2

Where does my audio go?

Accepted Answer

By default, nowhere: speech is transcribed on your device (Apple’s on-device recognizer or the built-in Whisper model) and raw audio is never saved. If on-device transcription isn’t available, audio goes only to the provider you configured.

Question 3

Can the interviewer see the overlay?

Accepted Answer

In Assistive mode (the default), yes — the overlay is visible in your screen-share or recording like any other window. In Focus mode (opt-in), it stays on your own display and doesn’t show up in most screen-shares or recordings, though some capture methods can still include it. How you use it is your responsibility.

Question 4

Do my API keys or transcripts leave my machine?

Accepted Answer

BYOK keys stay in your OS keychain and go only to your provider. With Local Ollama, nothing reaches the cloud at all. Transcripts are AES-256-GCM encrypted on your device with a key kept in your OS keychain, and in Local or BYOK mode they never reach a CuePreem server. If you choose Cloud mode instead, a transcript-derived prompt transits our own gateway to reach the model — we never store it or train on it, and the gateway’s log format has no field for it.

Question 5

How fast is it?

Accepted Answer

We design for the first note in under two seconds at the median: on-device transcription, fast intent classification, and streaming the first token before the full answer is done. Real timing depends on your mode — your hardware on Local, your provider on BYOK.

Question 6

Which platforms does it run on?

Accepted Answer

CuePreem runs natively on macOS 13 or later on Apple Silicon (M1 or newer) and on 64-bit Windows 10 and 11, built on Tauri 2 and tuned for low-latency on-device audio on both.

Question 7

Does CuePreem join my meetings as a bot?

Accepted Answer

No. CuePreem never joins as a participant or posts a bot into your call — it captures your own device’s system and microphone audio locally, the same way a personal note-taking app would. A recording-consent reminder opens every live session, so you can let participants know you’re taking notes.

Question 8

What happens to the résumé and job description I upload for prep?

Accepted Answer

The files themselves never leave your computer — they’re parsed locally. Only the extracted text goes to whichever mode you’ve configured: your own API key or a fully local Ollama model, neither of which ever reaches CuePreem’s servers, or CuePreem’s own Cloud tier, which does but never stores or trains on it. The fit score assists your judgment; there’s no auto-reject anywhere.

CuePreem — interview prep for candidates, interview notes for recruiters