Readify - Document Intelligence Assistant

Readify is a premium RAG (Retrieval-Augmented Generation) pipeline designed for intelligent, context-aware document exploration. It enables users to upload multiple document formats and engage in deep, citation-backed conversations with their data.

✨ Premium Features

• Multi-Format Mastery: Support for PDF, DOCX, TXT, and Markdown.
• Advanced RAG Engine: Optimized dual-stage retrieval with MongoDB Atlas Vector Search.
• Real-time Vectorization: Parallel processing for fast ingestion of large documents.
• Strict Session Isolation: Data remains private and isolated per session.
• Verifiable Citations: Every answer includes precise source and page referencing.
• Sleek UX: Modern dark-themed interface with smooth Framer Motion transitions.

📺 Product Preview

---

🏗️ System Architecture

Readify uses a modular architecture that separates the ingestion pipeline from the query pipeline to ensure maximum performance and accuracy.

For a deep dive into the technical details, read the Full Architecture Guide and Technical Strategy.

---

🛠️ Technology Stack

Layer	Technology
Frontend	Next.js 14, Tailwind CSS, Framer Motion, Lucide Icons
Backend	FastAPI, LangChain, PyMongo
Database	MongoDB Atlas (Vector Search & Metadata Filtering)
Models	Gemini 2.5 Flash (LLM), Text-Embedding-004 (Google)

---

🚀 Getting Started

1. Prerequisites

• Node.js 18+
• Python 3.10+
• MongoDB Atlas Cluster

2. Setup & Installation

# Clone the repository
git clone https://github.com/pranavsinghpatil/Readify.git
cd Readify

# Install Backend Dependencies
pip install -r backend/requirements.txt

# Install Frontend Dependencies
cd frontend && npm install && cd ..

3. Environment Configuration

Create a .env file in the backend/ directory based on .env.example:

GOOGLE_API_KEY=your_key_here
MONGODB_URI=your_mongodb_atlas_uri
DB_NAME=readify_db
COLLECTION_NAME=documents

4. MongoDB Atlas Search Index

Configure an Atlas Search Index named vector_index on your collection with this JSON:

{
  "fields": [
    { "type": "vector", "path": "embedding", "numDimensions": 768, "similarity": "cosine" },
    { "type": "filter", "path": "session_id" },
    { "type": "filter", "path": "source" }
  ]
}

5. Launch

Simply run the included batch file to start both servers:

.\launch.bat

---

📄 Documentation & Media

• Architecture Guide - Deep dive into system design.
• Technical Strategy - RAG optimizations and technical decisions.
• 🎬 Product Demo Video - Watch the app in action.
• ⚙️ Architecture Walkthrough - Technical video overview.

---

Created by Pranav Patil. 🚀