Intelliparse – Smart PDF Analyzer & RAG Chat

Intelliparse is an AI-powered web application that allows users to upload PDF documents, extract knowledge, and interact with AI in a RAG (Retrieval-Augmented Generation) chat interface.

It supports:

• Authentication via Clerk (Google login)
• Vector storage and retrieval using Qdrant
• Context-aware AI chat leveraging uploaded PDFs, text, or URLs

Perfect for students, professionals, and anyone who wants fast, context-aware insights from documents.

---

🚀 Features

• Upload PDFs, text, or URLs and generate knowledge embeddings
• Generate smart summaries and FAQs from documents
• RAG chat interface to ask questions based on uploaded sources
• Vector storage with Qdrant for efficient retrieval
• Clerk authentication with Google SSO
• Interactive FAQ / Chat UI with code & link highlighting
• Modern UI with animations and responsive design

---

📸 Screenshots

Home / Upload Page

Summary & FAQs

RAG Chat

---

🛠 Tech Stack

• Frontend: Next.js, React, TypeScript, TailwindCSS, Framer Motion
• Backend: Next.js API routes, Axios
• Authentication: Clerk (Google SSO)
• Vector DB: Qdrant for storing and retrieving embeddings
• AI / Embeddings: OpenAI, Google Generative AI, LangChain
• Syntax Highlighting: Highlight.js

---

⚡ Installation

1. Clone the repository:

git clone https://github.com/icodervivek/intelliparse.git
cd intelliparse

2. Install dependencies:

npm install --force
# or
yarn install --force

3. Create a .env file in the root:

GOOGLE_API_KEY=your_google_generative_ai_key
QDRANT_URL=your_qdrant_host
QDRANT_API_KEY=your_qdrant_api_key
NEXT_PUBLIC_CLERK_PUBLISHABLE_KEY=clerk_publishable_key
CLERK_SECRET_KEY=your_secret_key
NEXT_PUBLIC_CLERK_SIGN_IN_URL=/sign-in
NEXT_PUBLIC_CLERK_SIGN_UP_URL=/sign-up
NEXT_PUBLIC_CLERK_AFTER_SIGN_IN_URL=/
NEXT_PUBLIC_CLERK_AFTER_SIGN_UP_URL=/

4. Run the development server:

npm run dev
# or
yarn dev

5. Open http://localhost:3000 in your browser.

---

💡 Usage

Authentication

• Sign in using Google via Clerk.
• Only authenticated users can upload documents or chat in the RAG feature.

Upload & Summarize PDF

1. Go to Home / Summarise PDF.
2. Click Select PDF and upload your source.
3. Click Upload to generate summaries, and FAQs.

RAG Chat

1. Navigate to the Home / Try Our RAG.
2. Click on the attachment icon and upload PDFs, URLs or text from your source.
3. Ask AI questions — it retrieves relevant context from uploaded PDFs, text, or URLs using Qdrant embeddings.
4. Answers include references to the source documents.

Managing Contexts

• Clear uploaded sources and chat history using the Clear Chat button.
• Vector embeddings are stored in Qdrant for faster retrieval.

---

🤝 Contributing

Contributions are welcome!

• Open an issue to suggest features or report bugs.
• Submit a pull request to add improvements.

---

Made with ❤️ by Vivek