🧠 Second Brain

RAG-Powered Note Taking App — Chat with your notes using AI

📖 Overview

Second Brain is a local-first, AI-powered note-taking application that lets you "chat" with your notes using Retrieval Augmented Generation (RAG). Built with Flutter for cross-platform support, it combines the power of local storage with optional cloud sync for a seamless note-taking experience.

✨ Features

• ✅ Notes CRUD — Create, read, update, and delete notes with search functionality
• 🔍 Semantic Search — Find notes using natural language queries powered by embeddings
• 💬 RAG Chat — Have conversations with your notes using AI (Google Gemini)
• 🎤 Voice-to-Note — Convert speech to text for quick note capture
• ☁️ Cloud Sync — Optional synchronization with Supabase (pgvector for embeddings)
• 🌓 Dark Mode — Beautiful Material 3 theme with light and dark modes

🛠️ Tech Stack

Category	Technology
Framework	Flutter 3.38.4
Language	Dart 3.10.3
State Management	Riverpod 2.6+
Local Database	SharedPreferences (Local-first)
AI Framework	LangChain.dart
LLM	Google Gemini API
Cloud Backend	Supabase (pgvector)

🏗️ Architecture

Second Brain follows Clean Architecture principles combined with MVVM pattern in a feature-based folder structure:

lib/
├── app.dart                    # App root widget
├── main.dart                   # Entry point
├── core/                       # Core utilities
│   ├── constants/             # App constants
│   ├── errors/                # Error handling
│   ├── theme/                 # App theme
│   └── utils/                 # Extensions & helpers
├── features/                   # Feature modules
│   ├── notes/                 # Notes feature
│   │   ├── data/             # Data sources & models
│   │   ├── domain/           # Entities & use cases
│   │   └── presentation/     # UI & state
│   ├── chat/                  # AI chat feature
│   ├── search/                # Semantic search
│   └── voice/                 # Voice input
└── shared/                     # Shared widgets & providers

Key Architecture Decisions

• Local-First: All data stored locally using SharedPreferences for instant access
• Optional Cloud Sync: Supabase integration for cross-device synchronization
• Clean Separation: Domain logic isolated from UI and data layers
• Testable: Architecture enables comprehensive unit and integration testing

🚀 Getting Started

Prerequisites

• Flutter SDK 3.38.4 or higher
• Dart SDK 3.10.3 or higher
• Android Studio / VS Code with Flutter extensions
• Git

Installation

1. Clone the repository

   git clone https://github.com/Muhammad-Bilal-03/second_brain.git
   cd second_brain

2. Install dependencies

   flutter pub get

3. Run the app

   flutter run

Running Tests

flutter test

Code Generation

For Riverpod code generation:

dart run build_runner build --delete-conflicting-outputs

🗺️ Roadmap

Phase 1: Foundation (Notes CRUD) ✅ Complete

• Project setup with clean architecture
• Core dependencies and folder structure
• CI/CD with GitHub Actions
• Notes CRUD with SharedPreferences
• Material 3 UI with search & empty states
• Riverpod state management (AsyncNotifier pattern)

Phase 2: Intelligence Layer (Embeddings + Vector Search) ⬅️ Next

• Text embedding generation
• Vector similarity search
• Semantic search UI

Phase 3: RAG Chat (LangChain.dart + Gemini)

• LangChain.dart integration
• Google Gemini API setup
• RAG pipeline implementation
• Chat UI with conversation history

Phase 4: Cloud Sync (Supabase + pgvector)

• Supabase backend setup
• pgvector for cloud embeddings
• Sync engine implementation
• Conflict resolution

Phase 5: Voice-to-Note

• Speech-to-text integration
• Voice recording UI
• Real-time transcription

Phase 6: Polish & Ship

• Performance optimization
• Comprehensive testing
• User documentation
• App store deployment

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

👨‍💻 Author

Muhammad Bilal

• GitHub: @Muhammad-Bilal-03
• LinkedIn: muhammad-bilal-bsse

---

Built with ❤️ using Flutter