System requirements for AI interview assistants

AI interview assistants that capture both sides of a call generally run as native desktop applications, not browser extensions. That means the first requirement is an operating system the app actually supports. SubcueAI ships native builds for macOS and Windows.

• macOS: a recent version that supports system audio capture permissions and screen/recording entitlements used by modern apps. Apple Silicon (M-series) and recent Intel Macs both work in practice.
• Windows: a current 64-bit version of Windows 10 or Windows 11 on standard consumer hardware.
• Not supported: ChromeBooks, iPads, iPhones, Android phones, and Linux are out of scope for a desktop interview assistant that needs OS-level audio access.

You also need to be able to install software on the machine. If your laptop is locked down by an employer with MDM, you may not have permission to install anything — in that case, see detectability and privacy for honest limits.

An AI interview assistant is only as good as the audio it can hear. Two streams matter:

• Your microphone — built-in laptop mics are usually fine; a USB or headset mic is better in noisy rooms.
• The interviewer's voice — captured from the system audio of your meeting app (Zoom, Google Meet, Microsoft Teams, or a browser tab).

SubcueAI is designed around dual audio capture so it can transcribe both sides locally without joining the call as a bot. You'll need to grant microphone and system-audio permissions the first time you run it. The tutorial walks through the exact permission prompts on macOS and Windows.

Important honest limit: if you wear Bluetooth earbuds that route the interviewer's audio off-device, system audio capture can become unreliable. Wired headphones or laptop speakers are the safest setup.

Real-time transcription and answer generation rely on cloud models, so a stable internet connection matters more than raw CPU power. A typical home broadband or tethered 5G connection is enough; flaky hotel Wi-Fi during an interview is the most common cause of lag.

• CPU/RAM: any laptop sold in the last few years that comfortably runs Zoom or Google Meet will run a lightweight overlay app alongside it.
• Display: one screen is enough. The SubcueAI overlay is a small floating window you can position anywhere — many users park it just above or beside the meeting window.
• Bandwidth: low-Mbps connections work because only audio transcripts and short text prompts are exchanged, not video.

Beyond hardware, you need an account and (for paid features) credits or a subscription. See pricing for current plans and the free tier. To go from zero to a working setup most people need about 5–10 minutes:

• Download the native app for your OS.
• Grant microphone and system-audio permissions.
• Do a quick test call with a friend or a recorded video to confirm both voices are transcribed.
• Open your meeting app (Zoom, Google Meet, or Microsoft Teams) and start the real interview with the overlay already running.

For a guided walkthrough, see the tutorial, and for the broader getting-started topic see getting started.