Adaptly

# Description

Adaptly is a full-stack Learning Management System that integrates AI throughout the learning experience. Built for students, instructors, and admins, it features automated quiz generation from lecture content, context-aware chat assistance during lectures, and personalized remedial feedback after assessments.

# Tech Stack

• Next.js 16 with App Router, Server Components, and Server Actions
• TypeScript for type safety across the full stack
• MongoDB with Mongoose and aggregation pipelines for analytics
• Groq AI (llama-3.3-70b) via Vercel AI SDK for chat and quiz generation
• NextAuth.js for role-based authentication (students, instructors, admins)
• Server-side caching with unstable_cache and tag-based invalidation
• Tailwind CSS + shadcn/ui for accessible, responsive design
• Comprehensive testing suite with automated CI/CD pipeline

# Problem

Most LMS platforms are static—students watch lectures, take quizzes, and that's it. There's no adaptive assistance during learning, and post-assessment feedback is generic or non-existent. Students need help at different stages: clarification during lectures, independent testing, and understanding mistakes afterward.

# Solution

Built an AI-enhanced LMS where students get contextual help when they need it. During lectures, an AI chat assistant provides clarification based on the content. Quizzes are on a separate page (academic integrity). After quizzes, the AI returns with enhanced context—knowing which questions were missed and providing targeted explanations. Instructors can auto-generate quizzes from lecture content and track detailed student analytics.

# Results

Successfully launched with comprehensive testing and CI/CD. Achieved 4x performance improvement through server-side caching (340ms → 82ms on cache hits). Built efficient database queries using MongoDB aggregation pipelines to handle complex analytics across hundreds of records with sub-second load times. WCAG 2.1 AA accessible.