Project Story: Smart Grade Assistant
💡 Inspiration
I was inspired by the main theme of the Digital Hackathon to Improve the Educational Environment. I read the hackathon challenges and noticed a big problem: Smart and Automated Assessment.
The document mentioned the "heavy burden on faculty members in correcting assignments and exams". It also highlighted the "difficulty of analyzing student performance comprehensively and quickly". I wanted to create a solution that helps teachers save time and gives students faster feedback. This fits perfectly into the First Track: AI in Improving the University Educational Environment.
Smart Grade Assistant App will help teachers to speed up the productivity of correcting assignments and exams by letting the app to compare between the model answers (which has the correct answers) with students' answers
⚙️ How We Built It
We focused on building a Minimum Viable Product (MVP) that is ready to work.
The Brain (AI): We used a Large Language Model (LLM) to act as the "Smart Tutor". We designed a prompt that compares a student's answer to a model answer.
The Interface: We used React to create a simple web page. This allows teachers to upload files easily.
🚧 Challenges We Faced
Accuracy: It was hard to make the AI grade exactly like a human. We had to adjust our instructions many times to ensure the "Smart Assessment" was fair.
Time Management: We only had a short time to build the prototype before the hackathon on February 15. We had to keep the features simple.
Data Handling: We wanted to make sure we could analyze educational data effectively.
📚 What We Learned
Problem Solving: We learned how to take a real educational challenge and turn it into a working digital solution.
AI Application: We learned how to use "ready-made AI tools and models" to improve the educational process.
Teamwork: Working fast to meet the "hybrid style" of the hackathon taught us how to collaborate efficiently.
Log in or sign up for Devpost to join the conversation.