🎙️ OffRecord — Privacy-First Voice Notes

Powered by Google Gemini AI

---

📁 What Each File Does

OffRecordAI/
│
├── .gitignore          ← Git ignore file
├── README.md           ← This file!
├── requirements.txt    ← List of Python packages to install (one command)
├── data/               ← Data storage
│   └── db/             ← Database files
│       └── notes.json  ← Where your notes are saved (created automatically)
├── env/                ← Environment configuration
│   ├── .env            ← Your API key (not committed to git)
│   └── .env.example    ← Template for API key setup
├── backend/            ← Backend code (Flask app)
│   ├── app.py          ← The brain — Python server that handles everything
│   ├── static/         ← Static assets (CSS, JS)
│   │   ├── app.js      ← Makes buttons work in the browser
│   │   └── style.css   ← Makes the app look beautiful
│   └── templates/      ← Jinja templates
│       └── index.html  ← The web page you see in the browser
├── frontend/           ← Frontend code (for future expansion, e.g., React)
└── product/            ← Product documentation
    └── engineering/
        └── spec/
            └── mvp.md

---

🛠️ Setup — Step by Step

---

STEP 1 — Install Python

1. Go to: https://www.python.org/downloads/
2. Download Python 3.10 or newer
3. Run the installer
4. ⚠️ IMPORTANT: During install, check the box that says "Add Python to PATH"
5. Click Install

Check it worked — open Terminal (Mac) or Command Prompt (Windows) and type:

python --version

You should see something like Python 3.11.x

---

STEP 2 — Get Your Gemini API Key

OffRecord uses Google's Gemini AI to:

• Transcribe your voice into text
• Remove names and personal info
• Summarize into structured insights

---

STEP 3 — Open the Project in VS Code

1. Open VS Code
2. Go to File → Open Folder
3. Select the OffRecordAI folder
4. You'll see all the files in the left sidebar

---

STEP 4 — Set Up Your API Key

1. In VS Code, find the file called env/.env.example
2. Right-click it → Rename → change name to .env (Just remove the word "example" — the dot at the start is important!)
3. Click on env/.env to open it
4. Replace your-gemini-api-key-here with your actual key:

   GEMINI_API_KEY=AIzaSyYourActualKeyHere
   

5. Save the file (Ctrl+S on Windows, Cmd+S on Mac)

---

STEP 5 — Open the Terminal in VS Code

1. In VS Code, go to Terminal → New Terminal (or press Ctrl+ on Windows / Ctrl+ on Mac)
2. A panel will appear at the bottom — this is where you type commands

---

STEP 6 — Install Required Packages

In the terminal, type this and press Enter:

pip install -r requirements.txt

Wait for it to finish. You'll see packages downloading. This only needs to be done once.

If you see "pip not found", try: pip3 install -r requirements.txt

---

STEP 7 — Run the App!

In the terminal, type:

python backend/app.py

You should see:

══════════════════════════════════════════════════
🎙️  OffRecord is running!
📱  Open your browser → http://127.0.0.1:5000
🔒  Privacy-First | Powered by Gemini
══════════════════════════════════════════════════

---

STEP 8 — Open in Your Browser

1. Open Chrome, Safari, or Firefox
2. In the address bar, type: http://127.0.0.1:5000
3. Press Enter
4. OffRecord loads! 🎉

---

🎤 How to Use OffRecord

1. Click the amber button to start recording
2. Speak your thoughts — ideas, notes, anything
3. Click again to stop recording
4. Wait ~10-15 seconds — Gemini AI is:
   • Transcribing your voice to text
   • Removing any names or personal info
   • Generating structured insights
5. Your note appears with:
   • 💡 Key Ideas — the main points you said
   • ✅ Action Points — things you mentioned doing
   • 🧠 Important Thoughts — notable insights
6. Click any note on the left to read it
7. Use the search bar to find old notes by keyword

---

🔒 Privacy Features

Feature	Status
Raw audio stored permanently	❌ Never
Audio deleted after processing	✅ Always
Names removed from text	✅ Automatically
Audio uploaded to Gemini	✅ Temporarily, then deleted
Final insight saved	✅ Locally on your computer

---

❓ Common Problems & Fixes

Problem	Fix
"pip not found"	Try pip3 instead of pip
"python not found"	Try python3 instead of python
"API key invalid"	Check your .env file — make sure there are no spaces around the =
"Microphone access denied"	Click "Allow" when the browser asks
"No speech detected"	Speak for at least 3-4 seconds
Page doesn't load	Make sure python app.py is still running
Very slow processing	Normal! Gemini is doing 3 tasks — give it 15-20 seconds

---

🛑 To Stop the App

In the terminal, press: Ctrl + C

To run it again: python backend/app.py

---

💡 Tips for Best Results

• Speak clearly for at least 5 seconds
• You can ramble — AI will extract the key points
• Try speaking about a meeting, idea, or plan
• Search notes with any keyword from the content
• Notes are saved in data/db/notes.json — open it to see raw data

---

🚀 What This Prototype Shows

OffRecord MVP demonstrates:

Voice → AI Processing → Anonymous Insights
(raw audio is never permanently stored)

The full product would add:

• Local Whisper model (no internet needed for transcription)
• Voice pitch shifting (anonymize the audio itself)
• End-to-end encryption
• Mobile app
• Team workspaces

---

Built with: Python · Flask · Google Gemini 2.5 Flash