CourseFlow

AI-powered course builder that creates custom roadmaps and adapts based on your progress, quizzes, and weaknesses — like having a personal learning coach.

Comment 1

Inspiration

We noticed that most self-paced learning platforms offer static courses with no real-time feedback or adaptation. Learners often struggle to stay motivated, lose track of progress, and waste time on irrelevant materials. Inspired by the idea of having a personal tutor available 24/7, we set out to build a system that not only generates custom learning roadmaps but also evolves with each learner’s unique journey.

What it does

  • Generates Personalized Roadmaps: Breaks any skill goal into modules, subtopics, micro-projects, and free online resources.
  • Auto-Generates Quizzes: Creates multiple-choice quizzes for each module to assess understanding.
  • Adaptive Feedback Loop: After each quiz, collects user scores and reflections, then uses AI to identify weaknesses, suggest extra resources, and adjust the roadmap.
  • Progress Dashboard: Visualizes strengths, weaknesses, and next steps, keeping learners motivated and on track.
  • Export & Share: Allows users to export their customized roadmap and progress report as PDF or shareable link.

How we built it

  • Frontend: React.js (Next.js) for a responsive, component-driven UI.
  • Backend: Node.js with Express to handle API endpoints and business logic.
  • Database: Supabase for user auth, roadmap storage, quizzes, and progress tracking.
  • AI Models: Llama 3 70B served via Groq Cloud for fast, cost‑efficient inference.
  • Deployment: Vercel for frontend, Railway for backend — both on free tiers.
  • Visualization: Mermaid.js for interactive roadmaps.

Challenges we ran into

  • Prompt Consistency: Fine‑tuning AI prompts to always output valid JSON and consistent structure.
  • Adaptive Logic: Designing algorithms to meaningfully adjust a learner’s roadmap based on quiz results and self‑reported confidence.
  • Resource Curation: Ensuring AI reliably suggests high‑quality, free resources without recommending paywalled content.
  • Rate Limits: Managing API usage within free credits during intense hackathon development.

Accomplishments that we're proud of

  • End‑to‑End AI Flow: Seamless integration from user input → AI roadmap generation → quiz creation → adaptive feedback.
  • Real‑Time Adaptation: Demonstrated dynamic roadmap adjustments live during the demo.
  • Zero‑Code Deploy: Full deployment on free tiers without any paid services.
  • Strong UX: Simple, intuitive interface that users can pick up in seconds.

What we learned

  • The power of feedback loops in AI‑driven education.
  • How to craft system prompts that balance creativity with strict structure.
  • Best practices for integrating modern LLMs (Groq Cloud) into a full‑stack app.
  • Importance of user reflection in personalization and motivation.

What's next for CourseFlow

  • Team Collaboration: Enable shared roadmaps and progress tracking for study groups.
  • Version Comparison: Show “before vs. after” knowledge maps to highlight improvement.
  • Scheduled Reminders: Automated scheduling and calendar integration for study planning.
  • Mobile App: Lightweight iOS/Android companion for on‑the‑go learning.
  • Voice Support: Add voice‑guided modules and hands‑free quiz taking.

Built With

Share this project:

Updates