Evolve AI

• 

  Evolve AI

• 

  Dashboard

Inspiration

We noticed that students have access to decades of past exam papers but often use them passively, by simply re-reading them. The inspiration for Evolve AI was to create a tool that transforms this passive resource into an active, intelligent, and personalized study experience. We wanted to build something much more than just a chatbot—a true study partner that understands a student's weaknesses and helps them improve.

What it does

Evolve AI is a web platform that allows students to upload their past year exam questions. Our AI then gets to work: Analyzes: It processes the documents to identify high-frequency topics, key concepts, and question patterns. Personalizes: It generates interactive quizzes and flashcards from the material, specifically targeting areas where the student is weakest. Guides: It creates a dynamic study plan that adapts over time, ensuring students focus their efforts where it matters most for maximum impact.

How we built it

We built Evolve AI on a modern, robust tech stack designed for scalability and a great user experience: Framework: Next.js 14 with the App Router for a full-stack, server-rendered application. Language: TypeScript for type safety and improved developer experience. Styling: Tailwind CSS for a utility-first approach to building a clean and responsive UI. UI Components: Shadcn/UI provided a library of accessible, pre-built components that accelerated our development and ensured a professional look and feel. AI & Backend: For this initial prototype, we've mocked the AI processing and database interactions, creating a seamless user flow that is ready for integration with a real large language model (LLM) backend and a database like Supabase or PostgreSQL.

Challenges we ran into

Our primary challenge was designing a system that can accurately parse a wide variety of document formats (.pdf, .docx, .txt) and correctly extract individual questions and their corresponding answers. Another significant hurdle was conceptualizing the adaptive learning algorithm; creating logic that can genuinely identify a user's weak points and generate a truly dynamic study plan is complex. Finally, ensuring the user interface remained simple and intuitive, despite the complex processes happening in the background, was a constant focus.

Accomplishments that we're proud of

We are incredibly proud of creating a complete, end-to-end user experience, from the polished landing page and seamless sign-up process to the interactive dashboard. Building a clean, maintainable, and scalable architecture using Next.js and Shadcn/UI is a major accomplishment. We successfully implemented a core feature—the file upload and processing workflow—which serves as the foundation for the entire platform's intelligence.

What we learned

This project reinforced the importance of a solid foundation. Choosing the right tech stack and project structure (like using the src/ directory) from the start saved us significant time later on. We learned how powerful modern UI libraries like Shadcn/UI are for rapidly building beautiful and functional interfaces. Conceptually, we learned that the future of AI in education lies not just in information retrieval, but in creating structured, personalized learning pathways.

What's next for Evolve AI

The future for Evolve AI is exciting, and our roadmap is focused on bringing our intelligent features to life: Full AI Backend Integration: Connect the frontend to a powerful LLM (like the Gemini API or OpenAI's GPT) to perform real-time analysis and question generation. Database & User Persistence: Implement a full-fledged database to store user data, uploaded documents, and performance analytics securely. Advanced Analytics Dashboard: Build out the performance dashboard with rich visualizations and actionable insights to help students track their mastery over time. Spaced Repetition System: Integrate an algorithm to re-introduce questions and concepts at optimal intervals for long-term memory retention.

Built With

• groqapi
• next.js
• react
• shadcn
• supabase
• tailwindcss
• typescript