No hotkey, no clicks
The moment the interviewer finishes a question, the answer is already on your screen — automatically. You never touch a key, so nothing looks suspicious on camera.
Answers appear on their own, no hotkey
A real-time AI interview copilot that hears the question and puts a structured answer on your screen while the interviewer is still talking. Screenshot a coding task and get a worked solution.
Interview Copilot is a desktop app
It's a desktop app for macOS, Windows and Linux — a phone cannot capture the interviewer's audio or hide a window from screen share. Open this page on your computer to install it.
And during the interview this phone can be your screen: scan the QR code in the laptop app and the answers appear right here. How phone mode works →
Free: 5 minutes of live answers + 3 screenshots, no card. Pro is $25 for 30 days, no auto-renewal.
Works in every round: Recruiter screen·Behavioural·Technical·Live coding
The recruiter asks, and the answer is already on top of your call, where only you can see it. The people are staged; the app and its answers are real.
No sound needed.
The overlay never appears in screen-share or recordings on Zoom, Google Meet, Microsoft Teams, or any call app.
Auto-Answer
Auto-Answer is what makes Interview Copilot different. Most tools need you to press a hotkey for every question. Ours doesn’t.
The moment the interviewer finishes a question, the answer is already on your screen — automatically. You never touch a key, so nothing looks suspicious on camera.
Live transcription plus streaming AI means the answer appears as the interviewer finishes speaking.
It listens continuously, detects when a real question ends, and triggers the answer itself — no manual trigger, no missed questions during a fast conversation.
Because it’s fully automatic, you keep eye contact and focus on delivery instead of reaching for shortcuts — the interview feels natural.
Prefer control? A single hotkey answers on demand too. But Auto-Answer is why people pick Interview Copilot — set it once and forget the keyboard.
Under the hood it’s a simple but reliable pipeline. Here’s what happens every time the interviewer asks something.
The app captures the other side’s audio (the call’s system sound) and transcribes speech in real time — nothing to switch on by hand.
It detects when an utterance is genuinely finished (a pause or a question intonation) instead of reacting to half-sentences. Fillers like “umm, so” are ignored.
The instant a question ends, the model generates an answer to exactly that question. If the interviewer clarified or rephrased, Copilot takes the latest wording.
If another question arrives while it’s still answering, it goes into a queue and is handled next. Duplicates and overlaps are filtered out: one answer in flight at a time.
Coding tasks
A sample coding task is open in the browser. Interview Copilot appears in a separate floating window: capture the screen, review the solution, paste it into the editor, and run the tests.
[2, 7, 11, 15], target = 9 → [0, 1][3, 2, 4], target = 6 → [1, 2]def two_sum(nums, target):
On Windows it is Ctrl+Shift instead of ⌘⇧.
A coding task gets a working solution. Finished code gets checked for bugs, race conditions and edge cases. A multiple-choice test gets an answer to each question. SQL and system design prompts too — it works out what is on screen by itself.
The statement doesn't fit on one screen? Press ⌘⇧K to capture a frame, scroll, capture again — up to three — then ⌘⇧⏎. The current screen goes in last, and the model reads all the frames as one task.
Ran the solution and got an error or red tests? ⌘⇧B sends the screen together with the previous solution. You get what went wrong, the corrected code and what changed — in words you can say to the interviewer.
Every answer comes with short reasoning and the complexity, so you walk the interviewer through the approach instead of reading code cold.
Resume personalization
Add your resume and the job description, and answers come in the first person: your projects, stack and numbers, tuned to the role.
Frontend Engineer · 5 years
Payments in React, TypeScript · LCP 3.1 → 1.4 s
Senior Frontend · fintech
Requirements: React + TypeScript, payment systems
PDF, DOCX or TXT on the “Resume” tab — just once.
Paste the text into the “Job description” field.
A specific project, stack and result — instead of generic phrases.
Your resume and job description are stored on your device. On Pro they go to the AI as context for your answers; on Free they are not sent. Privacy →
System design
“Design WhatsApp” is not a two-line question. The interviewer grades how you drive the round: what you clarify, how you size the load, where the bottleneck is and what you trade off. Interview Copilot lays out exactly that path while they are still listening.
AUTO keeps answering briefly; press ⌘⇧D and the last question becomes a full design plan.
How would you design a URL shortener?
I'd start from requirements and optimise for reads — redirects outnumber new links a hundred to one.
Phone as a second screen
The laptop listens to the call and reads the screen. The answers stream to your phone — scan a QR code, nothing to install. Hide the overlay, and the laptop screen stays clean on the share and in the recording.
macOS, Windows or Linux. It hears the interviewer and reads the task on screen.
The phone opens a page in its browser and connects in seconds. No app to install.
Voice answers, coding solutions and design plans arrive as they are written. The overlay can be hidden.
Any phone with a browser · Zoom, Teams, Meet, Tencent Meeting · included in Free and Pro
Interviews in another language
The interviewer can speak English, German, Spanish or any of 30 languages. Every question appears translated into your language right under it, and the answer arrives in your language too. Interview in English? Hover over the answer to see how to say it in English.
The interviewer’s words appear live — and the translation shows up in the same block, right below.
Pick any of 30 languages in Settings. The answer is written in it, so you get it instantly.
Hover over the answer to see how to say it in English — a ready-to-speak version you can read out loud.
Choose the interviewer’s language and yours in Settings · included in Free and Pro
“How to say it in English” appears in English-language interviews
200,000+ answersgiven during real interviews
Honestly didn't expect it to be this quick — the answer was already on screen before I'd finished listening to the question.
Had a screen-share coding round and it read the task straight off the screen, then gave me an approach to start from.
I'd prepped in English, but they asked in my own language. Didn't have to change a thing.
Trusted by candidates interviewing at world-class engineering teams
Two modes
Technical mode is on by default. If your interview is not about code, switch it once in Settings — the mode changes how the AI answers, everything else works the same.
For any profession: lawyer, manager, HR, sales, finance, marketing, healthcare, education. Answers follow STAR, sound like you, and never invent statutes, standards or figures.
Everything the app did before: live answers to technical questions, coding tasks read straight off your screen and reviews of someone else’s code.
Comparison
Every other interview copilot needs you to trigger each answer — a hotkey, a click, a tab-switch the interviewer can notice. Auto-Answer replies the instant the question ends. You just talk.
How it works
Set it up once. On the call, it does everything in real time.
Add your CV and the job description. The AI answers in first person, grounded in your real experience — never generic.
One-time permission lets the app hear the system audio of the call, not your microphone. On macOS, quit and reopen the app after you turn it on.
A transparent window sits in the corner of your screen — invisible to screen sharing and recording.
It transcribes each question and instantly shows a concise, confident answer you can read out naturally.
Enable Auto mode and stop touching your keyboard — the AI detects each question and answers automatically, in real time, while you just read and speak.
No answers? Play the call through your computer’s default output device. How to install →
Pricing
The app is free and runs on Groq. Want dramatically better, more accurate answers? Pro runs on the latest Claude, personalized to your resume — $25 for a month. No subscription: when the month ends, you just pay for another one.
No auto-renewal — we never charge you again
Pay once by card or crypto. You’ll sign in first.
No auto-renewal — we never charge you again
Everything in Pro, plus:
ULTRA
Ultra records both voices, turns the call into a transcript and reviews every answer: what was wrong, what was missing and how to say it.
Hints work as always. Ultra records you and the interviewer in the background.
Every interview stays there for 30 days: download the transcript or press “Analyze”.
Each question gets a verdict, what was missing, a model answer and what to study.
Ultra also records your microphone during interviews, and you can switch that off. Transcripts and reviews are deleted after 30 days. Recording laws differ by country, and some require the other side’s consent. Privacy →
What is the difference between a liveness and a readiness probe?
Liveness checks if the pod is ready to receive traffic, and readiness restarts the container.
How do you handle secrets in your scripts?
I put them in a .env file next to the script. The repository is private.
A job fails halfway. What can happen if you retry it?
Hmm, I am not sure. Can we move on?
Strong on infrastructure, weak on reliability: secrets, retries and probes.
Liveness checks if the pod is ready to receive traffic, and readiness restarts the container.
Liveness and readiness are swapped. Readiness decides whether the pod gets traffic, liveness restarts a stuck container.
Readiness keeps the pod out of the service until it is ready. Liveness restarts the container if it hangs. For a slow start, add a startup probe.
I put them in a .env file next to the script. The repository is private.
We always run terraform plan and somebody reviews the pull request.
Hmm, I am not sure. Can we move on?
Beyond the interview
A real IT recruiter tells you what to change in your resume so your applications get answered more often.
Right in the app — the same resume it already uses to answer for you.
For example, Senior Go Backend Engineer. The review is done for that specific role.
What to replace, what to add and what's missing for this role. The answer arrives in the app.
Use cases
From your first screening call to the final onsite — Interview Copilot adapts to the interview in front of you.
Answer system-design and algorithm questions with confidence, and screenshot a coding task to get a working solution in seconds.
Turn "tell me about a time..." into a crisp STAR answer grounded in your real projects — no rambling, no blanking out.
Handle recruiter phone screens smoothly: expected salary, notice period, "why this company" — answered in your voice.
Works invisibly over Zoom, Google Meet and Microsoft Teams. The overlay never shows up in screen-share or recordings.
Get a structured approach to scalability, databases and trade-offs, tailored to the role and your stack.
Understand fast questions and reply with clear, natural phrasing — with automatic correction of technical terms.
FAQ
An interview copilot, or interview co-pilot, is an AI interview assistant that helps you live during the real call. Interview Copilot is a desktop app that hears the interviewer's questions in real time and shows an answer on your screen, invisible in screen sharing. Unlike prep trainers, it works during technical interviews and coding rounds.
No. The overlay is a special invisible window that does not appear in screen-share or recordings on Zoom, Google Meet, Microsoft Teams, or any other call software. The invisible interview assistant that stays hidden on screen share →
The app is free and runs on Groq — 5 minutes of live recognition plus 3 coding-screenshot solutions, no resume personalization. For dramatically better answers on the latest Claude, personalized to your resume, get Pro for $25 a month: open the payment page, pay by card, and access turns on automatically within a minute. There is no subscription — when the month ends, you simply pay for another one. All sales are final, so try the free version first.
The app listens to your computer’s system audio, not your microphone, and streams it for speech-to-text; we don’t store audio recordings. Your resume and job description stay on your device: on Pro they go to the AI as context so answers fit your experience, on Free they aren’t sent. A screenshot is sent only when you press the coding hotkey. To diagnose failures we keep a technical log (including question and answer texts and screenshot images) for 30 days, and you can delete your account in Settings at any time. Ultra also records your microphone, which you can switch off. Full details: Privacy →
macOS (Apple Silicon and Intel), Windows and Linux (beta). Setup is a few steps; the install guide covers the permission macOS asks for on the first run and what to do if Windows shows a warning. How to install →
Auto-Answer is our signature feature. With it on, the AI automatically answers every question the interviewer asks — the instant they finish speaking — without you pressing any key. It detects questions in real time and shows the answer hands-free, so you can focus on speaking naturally instead of reaching for a hotkey.
Yes. It answers whatever the recruiter asks — “Tell me about yourself”, why this company, motivation, salary expectations — with a short first-person answer you can say naturally. Add your CV and the job description (Pro) and the answers draw on your real experience. The same app then covers the next rounds: technical questions, live coding and code review.
Yes. Technical mode covers any IT role: SQL and statistics for analysts, test design for QA, product cases and metrics for PMs, plus behavioural questions for everyone. If your interview is not about tech at all — sales, HR, finance, law — switch to General mode once in Settings. Not sure? Try it free on a practice call first.
It captures the system audio of the call — not your microphone — and transcribes the interviewer's questions in real time.
Yes. Upload your CV and the job description, and the AI answers in first person, grounded in your real experience instead of generic textbook responses.
The app interface is available in 30 languages, including English, Ukrainian, Chinese, Japanese and Spanish. The interviewer’s question and the answer don’t have to be in the same language, and technical terms are corrected automatically for software engineering interviews.
Yes. Press ⌘⇧⏎ (Ctrl+Shift+Enter on Windows) and it works out what is on screen: a coding task gets a working solution with an explanation, finished code gets checked for bugs and edge cases, a multiple-choice test gets an answer to every question. A long statement can go in up to three frames with ⌘⇧K, and if your solution fails, ⌘⇧B returns the fix.
Yes. Interview Copilot works on any video call — Zoom, Google Meet, Microsoft Teams, Webex, Tencent Meeting and others. It captures the call audio and stays invisible during screen share, so the interviewer only ever sees your normal screen.
Yes. For live coding and technical interviews, screenshot the task and get a working solution with a step-by-step explanation and time complexity in seconds — so you can talk through the approach naturally.
Most AI interview assistants are prep tools — mock interviews and practice questions before the call. Interview Copilot is a live interview copilot: it works during the real interview, hears the interviewer through system audio, and answers every question automatically in under a second. And unlike browser-based tools, it is a native overlay that stays invisible on screen share.
Yes. Press 📱 in the app and scan the QR code: the answers stream to your phone's browser in real time, with nothing to install on the phone. The laptop keeps listening to the call and reading the screen, and you can hide the overlay so the laptop shows nothing at all. Your AI interview assistant, on your phone →
Download Interview Copilot and try the demo free — set it up in under two minutes.
Choose your platform
Interview Copilot is a desktop app
Send link to my emailOn Linux the overlay isn't hidden from screen sharing — connect your phone as a second screen: the overlay hides and answers go to the phone
After downloading, open the file. On macOS the app asks for Screen Recording permission once: allow it, then quit and reopen the app.
Ask anything about the app — the Interview Copilot team reads every message.
So we can reach you if you lose this chat.