Grading Assistant - GradeMaster

• 

  Main Dashboard

Inspiration

Grading coding assignments manually takes hours and often lacks consistency. We wanted to build a tool that uses AI to automate grading while maintaining fairness and transparency for students and professors.

What it does

The Grading Assistant uses Gemini LLM to generate a reference solution for coding assignments, then automatically grades student submissions using rubric-based evaluation. It provides detailed feedback and displays results in a unified dashboard.

How we built it

We integrated Gemini API for LLM processing, used Python to parse and evaluate code files, and designed a web dashboard for professors using Flask and React.

Challenges we ran into

Handling multiple programming languages, aligning LLM evaluation with professor rubrics, and maintaining grading fairness were key challenges.

Accomplishments that we're proud of

Successfully automating the entire grading process from assignment upload to feedback generation, saving professors hours of manual work.

What we learned

We learned how to combine LLMs with rubric-driven evaluation, and how important user feedback loops are to ensure grading accuracy.

What's next for Grading Assistant

We plan to add plagiarism detection, multi-language support, and LMS integration for seamless adoption in universities.

Built With

• gemini
• javascript
• python
• react
• sqlite
• tailwind