The app extracts text from the uploaded PDF, chunks it into searchable segments, indexes them using Algolia MCP, and then uses OpenAI's API to generate answers based on relevant context.
Demo (27s): https://www.loom.com/share/f012303fb1794e5eb1989c656fc5e2b8?sid=70fccf0b-1350-43fa-8d15-cc22b17635ae Code: https://github.com/Rajyadav81/ai-study-assistant
Built with: - PDF.js (for extracting content) - Algolia MCP (for indexing/search) - OpenAI GPT-3.5 (for answers) - Vanilla HTML/JS + Express backend
Would love your feedback — especially around: - How to improve retrieval accuracy - Making it faster or lighter - Any open-source AI tools to replace OpenAI in future?
This is my submission to the DEV x Algolia MCP Server Challenge.
Thanks!
sksrbWgbfK•4h ago