WeaknessFinder AI

• 

  The website interface

Weakness Finder AI was inspired by the need to turn failure into feedback, making the learning process feel less like an exam and more like a partnership.

Instead of a standard quiz app that simply says "Wrong Answer", this app acts like a personal tutor that explains why you are wrong and which specific concepts you need to fix.

To build Weakness Finder AI, we will use Python because it is simple and powerful. We will use a library called Streamlit for the website (frontend) and Google Gemini API for the brain (backend).

The Problem: Sometimes, the AI acts too confident. You might ask for a "Physics Quiz," and it creates a question where none of the options are actually correct, or the explanation is slightly wrong.

Simplicity by Design:- Despite the complex technology running in the background, we maintained a "Sweet and Simple" user interface.

The Power of Prompt Engineering We learned that AI is not magic. It is a tool that requires precise instructions. Writing code was only half the battle; the real challenge was crafting the perfect "System Prompt".

"Upload Your Syllabus" (RAG Integration) Currently, the AI generates questions based on its general training. In the future, we plan to implement Retrieval-Augmented Generation (RAG). This will allow users to upload their own PDF textbooks or lecture notes. The AI will then generate quizzes specifically from that document, ensuring the questions are perfectly aligned with their college syllabus.

Built With

• 1.5
• cloud
• community
• controlgithub
• flash
• gemini
• modelgoogle
• python