📝 Memoir

Your AI-Powered Voice Notebook

One tap to record • Double tap to transcribe • Long press to talk with AI

---

"Your ideas are waiting for you."

📖 Overview

Memoir is an AI-powered smart voice pen built around the ESP32-C3. It allows you to instantly capture thoughts, transcribe them into text using AI, and even converse with an AI assistant.

Designed with an offline-first architecture, Memoir combines embedded systems, BLE communication, a Progressive Web App, and modern AI services into one seamless experience.

---

✨ Features

🎙️ Voice Recording

• Single tap to start/stop recording
• High-quality I2S MEMS microphone
• Automatic recording storage on MicroSD card
• Haptic and LED feedback

---

📝 AI Transcription

• Double tap to transcribe
• Powered by Whisper
• Beautiful dark-mode transcript viewer
• Works directly from the PWA

---

🤖 AI Voice Assistant

• Long press to activate
• Chat naturally with AI
• Powered by Groq Llama 3
• Voice responses using Web Speech API

---

📱 Progressive Web App

• Install like a native mobile app
• Works offline
• IndexedDB storage
• BLE connectivity
• Live notifications
• Apple Notes inspired interface

---

🔋 Power Features

• Deep Sleep (<5µA)
• Touch wake-up
• USB-C charging
• 400mAh LiPo battery

---

🏗️ System Architecture

                        MEMOIR ECOSYSTEM

              BLE                       WiFi
      ┌────────────────┐          ┌─────────────────┐
      │                │          │                 │
      │   ESP32-C3     │──────────│       PWA       │
      │   Smart Pen    │          │ (Phone/Desktop) │
      │                │          │                 │
      └───────┬────────┘          └────────┬────────┘
              │                            │
              │                            │
      ┌───────▼────────┐          ┌────────▼────────┐
      │   SD Card      │          │   IndexedDB     │
      │ Audio Storage  │          │ Offline Storage │
      └────────────────┘          └────────┬────────┘
                                           │
                                           │ HTTP
                                           ▼
                                   ┌─────────────────┐
                                   │ Flask Backend   │
                                   │ Whisper + Groq  │
                                   └─────────────────┘

---

🎮 Gesture Controls

Gesture	Action
📌 Right-side Up + Single Tap	Start / Stop Recording
📌 Right-side Up + Double Tap	Transcribe
🔄 Upside Down + Single Tap	Play Last Recording
🔄 Upside Down + Double Tap	Delete Recording
⏱️ Long Press	AI Assistant

---

🛠️ Tech Stack

Hardware

Component	Model
MCU	ESP32-C3
Microphone	INMP441
Amplifier	MAX98357
Speaker	8Ω 1W
IMU	MPU6050
Touch Sensor	TTP223
Storage	MicroSD SPI
Battery	400mAh LiPo
Charging	TP4056
LED	WS2812B NeoPixel

---

Software

Layer	Technology
Embedded	C++, Arduino Framework, PlatformIO
Backend	Python, Flask
Frontend	HTML, CSS, JavaScript
Database	IndexedDB
Communication	BLE + WiFi
AI	Whisper + Groq Llama 3

---

📂 Project Structure

Memoir
│
├── dashboard/
│   ├── index.html
│   ├── api.js
│   ├── manifest.json
│   └── sw.js
│
├── memoir-backend/
│   ├── app.py
│   ├── requirements.txt
│   └── .env
│
├── src/
│   ├── main.cpp
│   └── Config.h
│
├── lib/
│   ├── AudioManager/
│   ├── GestureManager/
│   ├── FileManager/
│   ├── VirtualButton/
│   └── ...
│
├── docs/
├── examples/
├── LICENSE
└── README.md

---

🚀 Getting Started

Clone Repository

git clone https://github.com/mm-black65/Memoir.git

cd Memoir

---

ESP32 Firmware

pio run -t upload

---

Backend

cd memoir-backend

python -m venv venv

source venv/bin/activate
# Windows
venv\Scripts\activate

pip install -r requirements.txt

Create a .env

GROQ_API_KEY=your_api_key

Run

python app.py

---

PWA

cd dashboard

python -m http.server 8080

Open

http://localhost:8080

---

🌐 Deployment

Backend

Deploy easily on Render

Build Command:
pip install -r requirements.txt

Start Command:
python app.py

---

Frontend

Deploy using Vercel

npm install -g vercel

vercel

---

🔌 Backend API

Endpoint	Method	Purpose
/health	GET	Server status
/transcribe	POST	Whisper STT
/chat	POST	AI chat
/assistant	POST	Voice Assistant
/providers	GET	Available AI providers
/history	GET	Chat history
/history/clear	POST	Clear history

---

📊 Project Status

Module	Progress
UI/UX	✅ 100%
BLE Integration	✅ 100%
IndexedDB	✅ 100%
Offline Sync	✅ 100%
Whisper Integration	✅ 100%
AI Assistant	✅ 100%
Hardware Assembly	🟡 30%
3D Enclosure	⚪ 0%

---

💰 Bill of Materials

Component	Cost
ESP32-C3	₹250
INMP441	₹250
MAX98357	₹200
Speaker	₹80
MPU6050	₹120
TTP223	₹40
MicroSD Module	₹300
NeoPixel	₹50
Vibration Motor	₹80
LiPo Battery	₹300
TP4056	₹50
Misc Components	₹400

Estimated Total

≈ ₹2,500

---

🤝 Contributing

1. Fork the repository

2. Create your feature branch

git checkout -b feature/amazing-feature

3. Commit changes

git commit -m "Added amazing feature"

4. Push

git push origin feature/amazing-feature

5. Open a Pull Request

---

🙏 Acknowledgements

• ESP32
• PlatformIO
• Hugging Face Whisper
• Groq
• Web Bluetooth API
• Flask

---

👨‍💻 Author

Mahi Ahalawat

B.Tech Robotics & AI
JIIT Noida

GitHub: @mm-black65

---

⭐ If you like this project, give it a star!

Made with ❤️, Embedded Systems, AI, and a lot of coffee ☕