Tutor Copilot analyzes student writing, identifies strengths and weaknesses, and provides actionable feedback to help students improve
Students can submit essays and receive instant rubric-based feedback designed to strengthen writing skills.

Inspiration

Writing is one of the most important academic skills, yet many students struggle to receive timely and detailed feedback on their work. Teachers often manage dozens of essays at once, making it difficult to provide personalized guidance for every draft. While AI tools have become more common, many focus on generating essays rather than helping students improve their own writing.

I was inspired to build Tutor Copilot through my work with SEA Literacy, a nonprofit organization that supports refugee and multilingual students. I saw firsthand how students often needed more individualized writing support than tutors and teachers could realistically provide. I wanted to create a tool that empowers students to improve their writing while still maintaining ownership of their ideas and voice.

What it does

Tutor Copilot is an AI-powered writing feedback platform that helps students improve essays through rubric-based analysis and actionable revision guidance.

Students can submit essays and receive detailed feedback on areas such as:

Ideas and supporting evidence
Organization and structure
Writing conventions and clarity
Overall effectiveness

Rather than writing essays for students, Tutor Copilot focuses on helping them revise and strengthen their own work through personalized feedback and revision suggestions.

How I built it

Tutor Copilot was built as a full-stack web application.

Frontend:

Next.js
React
JavaScript

Backend:

FastAPI
Python

Deployment:

Vercel
Render

The platform uses AI-powered analysis to evaluate student writing against structured rubrics and generate meaningful feedback that students can use to revise their work.

Challenges I ran into

One of the biggest challenges was designing feedback that was actually useful for students. Early versions of the system often produced feedback that was either too generic or too overwhelming.

Another challenge was balancing AI assistance with academic integrity. I wanted the platform to support learning rather than simply generate content for students. This required careful prompt design and a strong focus on revision-oriented feedback.

Building and deploying a full-stack application also required solving issues related to API communication, deployment, rate limiting, and user experience.

Accomplishments that I'm proud of

Successfully built and deployed a working AI-powered educational platform.
Conducted a pilot with students through SEA Literacy.
Analyzed dozens of student essays.
Created a system focused on improving writing rather than replacing it.
Presented related research at an international conference and received recognition in STEM and AI competitions.

What I learned

Through this project, I learned how to design and deploy a full-stack application, integrate AI into an educational workflow, and collect meaningful user feedback.

More importantly, I learned that technology is most impactful when it addresses a real problem faced by real people. Working with students helped shape the project into a practical educational tool rather than just a technical demonstration.

What's next for Tutor Copilot

Future development plans include:

Teacher dashboards and classroom analytics
Student progress tracking across drafts
Enhanced rubric customization
Support for additional writing formats
Expanded access for schools, tutoring programs, and nonprofit organizations

The long-term goal is to make high-quality writing feedback more accessible to students everywhere while helping educators provide more personalized support at scale.