Live view of EDU-AI app showing AI-powered doubt solving and note summarization using GPT-3.5 and Hugging Face models.

📖 Project Story – EDU-AI: Your Personal AI Study Mentor

💡 Inspiration

The idea for EDU-AI was born out of personal experience and observation. As students, we often struggle with doubts when studying alone — especially during late-night sessions or just before exams. Finding clear, instant help isn’t always possible, and large volumes of notes can feel overwhelming to revise.

We asked ourselves: What if every student had an always-available AI-powered study buddy to support them whenever they needed help? That question sparked the creation of EDU-AI — a minimal yet powerful tool that brings AI assistance directly into students' hands.

🔨 How We Built It

We built EDU-AI using a lightweight Python stack that runs locally and is beginner-friendly to set up. Here's how:

Frontend: Designed using Streamlit to keep the UI clean, responsive, and fast to develop.
AI Doubt Solver: Integrated OpenAI’s GPT-3.5 model via API to act as a conversational chatbot for solving academic doubts.
Notes Summarizer: Leveraged Hugging Face’s transformers library with the distilbart-cnn-12-6 model to summarize large text blocks.
Local Setup: Focused on making the project easy to run without heavy dependencies, ideal for students with limited resources.

The folder structure was organized as: edu_ai/ ├── app.py ├── summary_generator.py ├── chatbot.py ├── requirements.txt └── README.md

yaml Copy Edit

🧠 What We Learned

Throughout this project, we learned a lot about:

Building intuitive AI tools with human-centered design
Using pre-trained NLP models effectively in real-time apps
Working with Streamlit to prototype and deploy ideas quickly
Handling API keys, modularizing code, and creating clean Python architecture

🚧 Challenges We Faced

Model Compatibility: We encountered issues when loading Hugging Face models without PyTorch installed. Fixing those dependency issues taught us more about environment handling.
OpenAI API integration: Ensuring the chatbot works smoothly with limited API calls while maintaining quality answers was a tricky balance.
Streamlit layout: Designing a simple interface that doesn’t overwhelm new users while also being feature-rich required UI/UX tweaking.
Time Limit: Building a usable AI mentor from scratch within 48 hours meant tight prioritization and fast iteration.

🚀 The Impact

We believe EDU-AI has the potential to become a valuable learning companion for students, especially those in remote areas or preparing independently. It’s not about replacing teachers — it’s about enhancing self-learning with the power of AI.

This project reflects our belief that education should be accessible, smart, and personalized.