STEM Mentor AI

Inspiration

Many students struggle with STEM subjects because they often receive only the final answer without understanding the reasoning behind it. I wanted to build a tool that acts like a personal tutor, helping students learn concepts step by step instead of simply solving homework for them.

As a computer science student, I was also interested in exploring how AI and computer vision could be used to improve education.

What it does

STEM Mentor AI is an AI-powered Telegram learning assistant for students.

The bot can:

Answer STEM questions with detailed explanations Generate personalized study plans Create topic-specific quizzes Solve math and physics problems from photos Store user activity and learning history Provide step-by-step solutions instead of only final answers

Students can interact with the bot directly through Telegram on any device

How we built it

I built the project using:

Python Aiogram Google Gemini API Gemini Vision MongoDB Atlas Telegram Bot API Render Cloud Platform The Telegram bot handles user commands and messages. Google Gemini generates explanations, study plans, and quizzes. Gemini Vision analyzes uploaded photos and explains problems step by step. MongoDB stores user activity and learning history. The application is deployed online using Render

Challenges we ran into

One of the biggest challenges was deploying the bot to the cloud and configuring all environment variables correctly.

I also encountered issues with Python package compatibility during deployment and had to troubleshoot dependency errors, API configuration issues, and database connections.

Another challenge was making AI responses educational instead of simply providing final answers

Accomplishments that we're proud of

I successfully built and deployed a complete AI-powered educational assistant that combines:

Conversational AI Computer Vision Cloud Deployment Database Integration Telegram Automation

I am especially proud that users can upload a photo of a STEM problem and receive a detailed explanation generated by AI

What we learned

Through this project I learned:

Telegram bot development with Aiogram Cloud deployment using Render MongoDB Atlas integration AI application development with Google Gemini Prompt engineering Debugging and deploying production-ready applications

What's next for STEM Mentor AI

Future improvements include:

PDF document support Voice questions and responses Personalized learning recommendations Study analytics dashboard Calendar reminders Additional STEM subjects and learning tools

My goal is to continue developing STEM Mentor AI into a complete AI learning platform that helps students learn more effectively