Grade.ai

• 

  This is the landing page of our application.

• 

  This is the evaluation and feedback output provided by our application.

• 

  Generating Score using LLMs.

• 

  Generating response using LLMs.

Inspiration: As students, we've often found ourselves grappling with assignments, pouring hours into our work, only to receive grades that don't quite reflect our efforts. We understand the importance of constructive feedback in the learning process.

What it does: It helps students grow, improve, and become better programmers. That's why we're driven to create a solution that not only benefits students but also eases the burden on educators. We believe that by enhancing the grading process, we can create a more rewarding and efficient learning experience for everyone involved.

How we built it: We've build an Autograder for coding, which analyzes the coding assignments written by students and provide reasonable feedbacks, while also generating study tips to keep in mind.

Challenges we ran into: Software problems while working with server, designing the application using newer technologies such as next.js, tailwind.css, tried multiple LLMs to try to get the required efficiency but couldn't.

Accomplishments that we're proud of: Got the right LLM to use towards the end and got the desired result which was super-satisfying, integrating multiple technologies under a time constraint and actually making it work and not giving up till the end!

What we learned: Time-management and to work under pressure while coding straight for 24 hours.

What's next for Grade.ai: Scale it to commercial level and make it a reliable tool for students and professors to use.

Built With

• apis
• codeqwen
• few-shot-learning
• flask
• google-cloud-platforms
• html
• hugging-face
• intel-developer-pc
• intel-hardware
• intel-smart-pcs
• javascript
• machine-learning
• ngrok
• one-shot-learning
• react
• tailwind.css