ChatPDF is a fully functional, production-ready full-stack application that enables users to upload PDFs and interact with them using AI-powered question answering. This project is built with a modern stack and showcases my capability to architect, develop, and deploy an end-to-end system.
| Layer | Technology Used |
|---|---|
| Frontend | React (App), ShadCN Components |
| Backend | Node.js, Express |
| NLP & RAG | OpenAI Embeddings + Pinecone Vector DB |
| File Parsing | pdf-parse |
| Database | MongoDB (for users, PDFs, queries) |
| Storage | Uploadthing |
| Deployment | Vercel (Frontend), Render (Backend API) |
I've been using Node.js extensively and chose it here to demonstrate my expertise in building scalable, production-level applications using JavaScript/TypeScript across the stack.
To implement document Q&A, I used Retrieval-Augmented Generation (RAG) — a modern architecture where PDF content is split, embedded using OpenAI, and stored in Pinecone, a vector database. This enabled:
- Accurate semantic similarity search
- Scalable, blazing fast response times
- A structured way to retrieve relevant content per question
This project is more than a tech demo — it's a real-world architecture designed to be deployable and extensible.
- Upload any PDF document
- Ask natural-language questions about the document
- Get smart AI answers using GPT + similarity search
- Previous session history
- Drag & Drop uploader
- Responsive and intuitive chat interface
- View all registered users
- Monitor uploaded resources
- Inspect user query sessions
- View platform usage analytics (charts + KPIs)
- Search/filter/sort data
- Designed for future scalability
🧪 Admin Credentials:
| Credentials | |
|---|---|
| username | [email protected] |
| password | admin1234 |
- User signups over time (line chart)
- Resources uploaded per day (bar chart)
- Queries generated daily (line chart)
- Top active users by query volume (pie chart)
- Resource upload breakdown per user
These insights help monitor platform growth and performance in real-time.
/frontend # Frontend (Landing page, Chat UI, Admin dashboard)
/backend # Backend (Node.js APIs, PDF parser, Pinecone integration)
- Add authentication and billing with Stripe for paid usage
- Support DOCX/CSV in addition to PDFs
- More granular analytics per document
- Document summarization on upload
- Real-time streaming response
I'm a full-stack developer with strong proficiency in Node.js, React, MongoDB, and AI integrations. This project showcases my ability to:
- Architect scalable systems
- Write clean backend logic
- Build engaging frontend interfaces
- Integrate third-party AI & DB services
- Deploy to modern cloud platforms (Vercel, Render)
This project is open for evaluation purposes. All content and code belong to the developer unless otherwise stated.
If you'd like to discuss this project or collaborate:
Email: [email protected] LinkedIn: [Your LinkedIn] GitHub: [Your GitHub]
Let me know if you want a README badge section or an ASCII logo at the top — or if you want me to convert this into a downloadable file.