Live Interview#

This page describes how to use Power Interview AI during a live interview session, including transcription, AI suggestions, code assistance, stealth mode, and window management.

If the interview has not happened yet, start with the Mock Interview guide instead: the app runs a spoken practice session of its own, and rehearsing in it is the step before this one.

Starting a Session#

1. Launch Power Interview AI.
2. Review configuration and settings to ensure your profile is complete and devices are set up correctly.
3. Click the Start button to activate transcription and AI assistance.

Once started, the app begins capturing audio from your configured microphone (your voice) and the system audio loopback (the interviewer's voice). Both channels are streamed for real-time transcription and displayed in the transcript panel.

To stop the session at any time, click Stop in the UI or press Ctrl+Shift+Q from any window (this also exits stealth mode if active).

Main UI during active session	Main UI during active session - Stealth Mode

Transcription#

How It Works#

• Your voice is captured through your selected physical microphone.
• The interviewer's voice is captured automatically via system audio loopback (whatever is playing through your speakers from the video call). No additional device configuration is needed.
• Both streams are sent to the transcription service and returned as live transcription.
• Each line is labeled by speaker name: your configured name or "Interviewer", with a timestamp.
• Press Ctrl+Shift+F8 to show or hide the transcription dock. This works in stealth mode too, where the control panel carrying the toggle button is hidden.

Interview Language#

Click the language button in the control panel to choose the language of your interview. The button shows the current code (EN, ES, JA) so you can read the setting without opening the menu.

One setting decides three things:

• Which speech model transcribes the call
• The language your reply suggestions come back in
• The language of your exported report, including the headings the summary writes for itself

Available Languages#

English, Spanish, German, French, Portuguese, Italian, Dutch, Polish, Russian, Ukrainian, Czech, Romanian, Greek, Hungarian, Swedish, Danish, Norwegian, Finnish, Turkish, Hindi, Japanese, Korean, Chinese, Vietnamese, Thai, Indonesian, Arabic, and Hebrew.

Each entry shows its own name alongside the English one, so you can find your language whichever of the two you are looking for. Typing in the open menu jumps to a language by its English name.

Changing Language Mid-Interview#

The picker stays live while a session is running, because an interview that switches language is exactly the case you cannot prepare for by restarting.

The two halves move at different speeds:

• Suggestions follow immediately - the next one is generated in the new language.
• Transcription carries its language as a connection setting, so both audio channels reconnect. Expect a second or two of gap, and whatever was being spoken at that moment may be lost.

The button shows a spinner while the reconnect is in progress. If transcription cannot reconnect, the button turns red and reads "Suggestions only" - your answers are in the new language but the transcript is still arriving in the old one. Stopping and starting the assistant clears it, since the next session opens both channels on the stored language.

Right-to-Left Languages#

Arabic and Hebrew render right-to-left throughout the transcript, suggestions, and exported report. Punctuation lands at the correct end of the sentence, and answers that mix scripts - a Hebrew explanation around a Latin product name, for example - keep their order. Code blocks stay left-to-right, since code is left-to-right in every language.

The App's Own Interface#

Buttons, menus, and dialogs stay in English regardless of the interview language. This is deliberate: an English button on a Spanish interview is a small inconvenience, while an English transcript of Spanish speech would be a wrong answer read out loud.

AI Reply Suggestions#

Reply suggestions are generated based on the current transcript, your profile (name, CV), and the job context you configured.

Getting a Suggestion#

Suggestions are triggered automatically as the AI determines a response is useful, or you can request one manually through the UI. The suggestion streams into the reply panel word by word in real time.

Scrolling Suggestions#

You can use the keyboard to scroll through reply suggestion content without leaving your video call window (use J/K for line‑by‑line movement or L to jump to the end):

AI Code Suggestions#

For technical interviews with coding problems, Power Interview AI can analyze your screen and suggest solutions.

Workflow#

1. When a coding problem appears on screen (in your browser, IDE, or shared screen), press Ctrl+Shift+F9 to take a screenshot. You can also skip the queue and automatically submit in one step by using Ctrl+Shift+F12 (capture + submit).
2. You can capture up to 4 screenshots to provide more context (e.g., multiple parts of the problem due overflowed content or different tabs).
3. Press Ctrl+Shift+F11 to submit the screenshots for analysis (not needed if you used F12).
4. The AI processes the images and streams a suggested solution with syntax-highlighted code into the code suggestion panel.
5. To clear the captured screenshots (e.g., mistakes), press Ctrl+Shift+F10 to clear all captured screenshots.

Scrolling Code Suggestions#

You can also jump to the end of the code panel using the third shortcut (O).

Stealth Mode#

The Power Interview AI window is always hidden from screen capture and screen share - interviewers cannot see it at any time, regardless of whether stealth mode is on or off.

Stealth mode is about focus control. When active, the window will not steal focus from your coding challenge, IDE, or video call. You stay in full control of your keyboard and mouse at all times.

Activating Stealth Mode#

Press Ctrl+Shift+M to toggle stealth mode on or off.

When active, the window collapses to a minimal status bar showing running state, credit balance, and active hotkeys.

Opacity Toggle#

Press Ctrl+Shift+N to toggle a low-opacity overlay, letting you glance at suggestions through a semi-transparent window without switching focus away from your active window.

Professional Mode#

Press Ctrl+Shift+F7 (or click the toggle in the control panel toolbar, next to the transcription/Clear/Export buttons) to switch AI reply suggestions between full sentences and short hints - a headline plus keyword bullets you can read at a glance. This works in stealth mode too.

Window Management#

Because you cannot use the mouse to interact with Power Interview AI while focused on your interview, all window controls are available via keyboard shortcuts.

Positioning the Window#

Use a numpad-style layout to snap the window to any screen position:

Fine-Tuning Position and Size#

Profile & Settings#

Access your settings from the menu icon (☰) in the top-right of the titlebar.

Profile Settings#

This is your interview configuration, not general account info (like your email or credits) - it's saved under your account so signing in on another device restores the same full name, profile, and context automatically.

Device Settings#

Both can be changed while a session is running. The microphone swap is instant and leaves the transcript unbroken; the language switch reconnects transcription and costs a second or two. See Interview Language for what happens on each half.

If a microphone you pick cannot be opened - unplugged, held by another app, or refused by permissions - the session stays on the one it was already using, and the control bar flags the device that failed rather than leaving you to work it out from silence.

Menu#

Click the menu icon (☰) in the titlebar to access:

• Your account email
• Configuration - open the profile/settings dialog
• Change password - update your account password
• Buy Credits - go to the payments page
• Stealth mode - toggle stealth mode on or off
• Light mode / Dark mode - toggle the app theme
• Documentation - open this documentation in-app
• Sign out - log out of your account

Session Tools#

The control panel toolbar includes three utility buttons (at the right edge): toggle transcription, Clear, and Export.

Export Interview#

The Export button offers a choice of two formats - Word Document (.docx) or Markdown (.md) - each containing:

• An AI-generated summary of the interview
• Full timestamped transcript with speaker labels
• All reply suggestions with the questions that triggered them

A save dialog appears so you can choose where to save the file.

Export Example - generated interview report in Word

Clear Session#

The Clear button removes all transcripts, reply suggestions, and code suggestions from the current session. This does not affect your saved profile, credentials, or settings.

Credits & Payments#

Credits are consumed while the assistant is running - covering AI reply suggestions and transcription. Your remaining credit balance is shown live in the titlebar, and in the status bar while in stealth mode.

A mock interview is charged differently: per question, follow-up and report rather than by the minute, with its transcription unmetered. The price of a session is quoted in its setup dialog before it starts.

To buy credits:

1. Click the menu icon (☰) in the titlebar.
2. Select Buy Credits.
3. Choose a credit package, complete payment, and your balance updates automatically.

You can also view your payment history and check pending payment status from the same page.

Complete Hotkey Reference#

On macOS the base modifier is Control+Option (⌃⌥) rather than Ctrl+Shift, because Ctrl+Shift combinations collide with system shortcuts there. ⇧ is Shift and ⌘ is Command. Every Ctrl+Shift+… shortcut mentioned elsewhere on this page follows the same mapping.

The canonical list of hotkeys is maintained in src/main/hotkeys.ts in the desktop app; update both the code and documentation when making changes.