✨ All Features

Built for the
modern job interview

Every feature designed to give you an edge — without looking like you have one.

⬇ Download Free 📖 Read the Guide
⚡ Core Feature
⚡

Streaming Answers

Answers stream word-by-word like someone typing live next to you. The first word appears in under 2 seconds — no waiting for a full response to generate before you can start reading.

✓Powered by Gemini 2.0 Flash SSE streaming endpoint
✓Works in both the main window and the invisible overlay simultaneously
✓Automatic fallback to standard response if streaming fails
✓Blinking cursor shows the AI is still generating
Meeting Notes — AI Answers
💬 "Walk me through your experience with React performance optimization."
I've worked extensively on React performance. At my previous company, I reduced initial render time by 60% using React.memo and useMemo for expensive computations. Key techniques: code splitting with React.lazy, virtualized lists for large datasets using react-window, and moving state colocation closer to consumers to avoid prop-drilling re-renders...
⚡ First word appeared 1.4s after question detected
Your Interview Screen (what Zoom sees)
👤
Your face — visible to interviewer
Zoom is sharing your screen right now
Meeting Notes Overlay
In my previous role I led a team of 5 engineers building a distributed payment system processing $2M daily...
👁️ Invisible to Zoom · Teams · OBS
🔒 Exclusive
👁️

Screen-protected overlay

The floating answer overlay sits on top of all your windows but is completely invisible to screen capture software — including Zoom, Microsoft Teams, Google Meet, OBS, Loom, and all recording tools.

✓OS-level setContentProtection(true) — not a CSS trick
✓Drag anywhere on screen; adjustable opacity (20–100%)
✓Toggle with ⌘+Shift+H from any app
✓Panic hide with ⌘+Shift+Esc — all windows vanish instantly
📸 Screenshot Mode
📸

Screenshot → AI Solution

See a coding problem, system design question, or diagram on screen? Press ⌘+Shift+S (or the Screenshot button). Meeting Notes captures your entire screen, sends it to Gemini Vision, and returns a full solution with working code.

✓Works on HackerRank, LeetCode, CoderPad, shared PDFs, whiteboards
✓Returns complete, working code in your preferred language
✓Explains time/space complexity and trade-offs
✓Android: pick any photo from camera roll for image analysis
📸 Screenshot Analysis Result
Problem detected: "Given an array of integers,
find the two numbers that add up to target."
def two_sum(nums, target):
    seen = {}
    for i, n in enumerate(nums):
        complement = target - n
        if complement in seen:
            return [seen[complement], i]
        seen[n] = i
# O(n) time, O(n) space ✓
Full solution with explanation in 1.8s
Answer Style Selector
💬 Natural
⚡ Concise
📚 Detailed
⭐ STAR
Natural: Balanced, conversational answers that sound human and authentic. Best for general use.

Concise: Max 80 words. Bullet points where natural. Ideal for rapid-fire phone screens.

Detailed: Up to 300 words with trade-offs and reasoning. Perfect for system design rounds.

STAR: Strict Situation · Task · Action · Result format. Required for behavioral questions.
🎭 Exclusive
🎭

4 Answer Styles

No other interview AI lets you control how answers are written. Switch styles per-session based on the interview type — concise for initial screens, detailed for system design, STAR for behavioral rounds.

✓Natural — balanced, human-sounding, great for general use
✓Concise — max 80 words, bullet points, for phone screens
✓Detailed — up to 300 words with trade-offs, for system design
✓STAR — strict Situation·Task·Action·Result for behavioral
🏆 Exclusive
🏆

Post-Interview Debrief

After every session, Meeting Notes automatically generates a structured debrief — summary, key topics covered, action items to study, and AI performance feedback. No other interview tool does this.

✓AI summary of the full conversation
✓Key topics and technology areas covered
✓Action items: what to study before the next round
✓Performance feedback and improvement suggestions
✓Saved to History for round-2 preparation
🏆 Session Complete — Google SWE L5
Duration47 min
Questions answered8
Key topics covered
Distributed systems System design Java concurrency Leadership
AI Feedback: Strong technical depth on distributed systems. Consider adding more quantitative metrics to behavioral answers.
Mock Interview Generator
Role · Difficulty · Type
Senior Frontend Eng · Medium · Technical
Generated Questions (5 of 10)
1. Explain how React's reconciliation algorithm works.
2. What are the trade-offs between CSS-in-JS and CSS modules?
3. How would you architect a design system for 20 teams?
🎯 Practice Mode
🎯

Mock Interview Mode

Generate AI-powered interview questions tailored to your exact role and seniority. Study the model answers and drill the questions you find hardest before walking into the real interview.

✓Behavioral, technical, coding, or mixed question sets
✓Easy / Medium / Hard difficulty levels
✓Custom tech stack: React, Python, distributed systems, etc.
✓Model answers generated on demand for each question
⬇ Ready to try it?

All features. Free to start.

5 free sessions per month. No credit card. No API key.

⬇ Download Meeting Notes