🌌 Lyra AI

The shortest path from curiosity to clarity. Lyra AI is a high-velocity research engine that synthesizes the open web into precise, cited insights. Driven by advanced LLMs, it transforms how you discover, learn, and build.

---

✨ Features

• 🧠 Elite Research Intelligence: Powered by Mistral AI, Lyra provides deep, long-form answers with professional Markdown formatting.
• 🧵 Thread Management: Organize your research with persistent chat history. Switch, rename, or delete threads with ease.
• 📱 Fully Responsive UI: A premium, glassmorphic design that works perfectly on desktop and mobile.
• ⚡ Real-time Synthesis: Powered by Socket.io for instant, streaming-like communication between the interface and AI brain.
• 🔍 Tavily Research Mode: Real-time web search capabilities optimized for detailed, cited insights.

---

🛠 Tech Stack

Layer	Technology
Frontend	React + SCSS (Glassmorphic Design)
State	Redux Toolkit (Thunk-based architecture)
Backend	Node.js + Express
Real-time	Socket.io
Database	MongoDB
AI Brain	Mistral AI + Tavily (LangChain integration)

---

🚀 Getting Started

1. Prerequisite

• Node.js (v18+)
• MongoDB (Local or Atlas)
• Mistral AI API Key
• Tavily API Key (for Research Mode)

2. Clone the Repository

git clone https://github.com/Ramiz1323/Lyra.git
cd Lyra

3. Backend Setup

cd Backend
npm install

Create a .env file in the Backend directory:

PORT=3000
MONGO_URI=your_mongodb_uri
MISTRAL_API_KEY=your_mistral_api_key
TAVILY_API_KEY=your_tavily_api_key
JWT_SECRET=your_jwt_secret

Run the server:

npm run dev

4. Frontend Setup

cd ../Frontend
npm install

Create a .env file in the Frontend directory:

VITE_API_URL=http://localhost:3000/api

Start the development server:

npm run dev

---

📄 License

This project is licensed under the MIT License. Developed with ❤️ by Ramiz1323.