AI That Reads PDFs Like a Topper 😮 | Built with Algolia + OpenAI

#devchallenge #algoliachallenge #webdev #ai

Hey dev

here i post 🧠 AI Study Assistant (PDF Q&A)
This is my submission for the Algolia MCP Server Challenge

🚀 Project Overview
📘 A personal AI-powered study assistant where you can:

📄 Upload your PDF notes
🔍 Use Algolia MCP server to index and chunk your content
💬 Ask natural-language questions
🤖 Get intelligent answers powered by OpenAI
🎥 Demo
▶️ Watch on Loom

💻 GitHub Repo
🔗 Rajyadav81/ai-study-assistant

🧪 How It Works
User uploads a PDF.
The text is extracted using PDF.js.
Text is chunked into 1000-character segments.
Chunks are pushed to Algolia index using MCP-style format.
User asks a question.
Algolia retrieves relevant chunks.
OpenAI API answers based on retrieved context.
🧰 Technologies Used
PDF.js
Algolia MCP Server
OpenAI API (GPT-3.5)
HTML, CSS, JavaScript
Express.js (Node backend)
🔐 MCP Integration
I used Algolia's Model Context Protocol (MCP) approach to store and query contextual information.

This allows OpenAI to respond more intelligently by grounding responses in user-uploaded notes.

✅ Submission Categories
🏅 Ultimate User Experience

🧠 Backend Data Optimization

🧠 What I Learned
This challenge helped me:

Understand chunking and contextual memory storage
Build a full-stack app with AI-powered interaction
Work with Algolia's search APIs and OpenAI integration
🙌 Thank You
Thanks to the DEV team and Algolia for this amazing challenge!

Excited to explore more with MCP and intelligent search 🤖