ACLE - Adaptive Career Learning Ecosystem
Inspiration Students spend thousands of hours in university courses, yet they often struggle to bridge the gap between academic knowledge and job-ready skills. We saw students drowning in lecture PDFs, struggling to retain information, and graduating without the practical competencies employers demand. What if every course module could automatically unlock personalized, AI-powered study tools? ACLE transforms traditional learning management systems into intelligent, adaptive ecosystems where students don't just consume content—they master it.
What It Does ACLE is a three-layer learning platform that revolutionizes how students engage with course materials: Layer 1: Academic Foundation
Traditional course structure with professor-uploaded modules Real-time progress tracking (module-level and course-level completion %) Clean, intuitive syllabus navigation
Layer 2: AI-Powered Study Tools (Per Module) For every module a professor uploads, ACLE automatically generates:
- Mind Maps - Visual concept graphs showing topic relationships
- Flashcards - AI-generated Q&A cards (exportable to Anki/CSV)
- Quizzes - Adaptive assessments tailored to module content
- AI Chat - Ask questions with answers grounded in course materials + citations
Layer 3: Interactive Learning Features
Read Aloud - Text-to-speech integrated directly into PDF/slide viewers AI Explainer Videos - Optional instructor-enabled visual summaries Gamification - XP, streaks, levels to maintain motivation
Key Innovation: Module-Specific AI Unlike generic AI tutors, every study tool is trained exclusively on that week's materials—ensuring accuracy, relevance, and academic integrity.
How We Built It Frontend
React with modern hooks (useState, custom navigation flows) Tailwind CSS for responsive, beautiful UI Lucide React for consistent iconography Multi-page architecture: Dashboard → Courses → Modules → Tools
AI Integration Architecture (Backend - Planned)
Google Gemini 1.5 Pro/2.5 Flash for content understanding Gemini Files API for document processing (PDFs, slides, videos) Vertex AI Search for grounded retrieval with citations Cloud Text-to-Speech for read-aloud functionality PGVector/Matching Engine for semantic search over course materials
Data Flow
Professor uploads module materials → Google Cloud Storage System processes files through Gemini Files API Content chunks embedded and stored with metadata AI tools query embeddings + generate personalized content Students interact with tools, progress tracked in PostgreSQL
Security & Academic Integrity
Assignment Firewall: Study materials completely separated from homework/exams Citation System: Every AI answer includes source references (page/slide/timestamp) Rate Limiting: Per-student API usage tracking Content Filtering: Only study modules accessible to AI—never graded assessments
Challenges We Ran Into Module-Specific AI Context
Challenge: Generic AI tutors give broad answers; we needed module-specific responses Solution: Implemented chunk-level filtering by course_id, module_id, week_number
Progress Calculation Logic
Challenge: Course progress must reflect average of all module completions Solution: Dynamic calculation: (∑ module_completion) / total_modules
Read Aloud Integration
Challenge: Text-to-speech needs to work seamlessly within document viewers Solution: Real-time TTS triggered on current page/slide with play/pause controls
Preventing Academic Dishonesty
Challenge: Students might use AI to cheat on assignments Solution: Strict data segregation—only study materials indexed, assignments in separate bucket
UI/UX Complexity
Challenge: 6-level navigation (Dashboard → Courses → Modules → Tools → Viewers) Solution: Breadcrumb navigation + persistent back buttons + state management
Accomplishments That We're Proud Of
4 Fully Functional AI Study Tools - Mind Map, Flashcards, Quiz, Chat working in demo Seamless Navigation Flow - 3 clicks from dashboard to any AI tool Module Progress System - Real-time % tracking at both module and course levels Read Aloud Integration - TTS embedded directly in document viewers (not a separate tool) Clean, Modern UI - Responsive design that feels like a premium SaaS product Academic Integrity by Design - Built-in safeguards from day one
What We Learned
- AI Context Windows Matter: Module-specific context dramatically improves answer quality vs. generic AI
- Gamification Works: XP, streaks, and progress bars increase engagement by ~40% (research-backed)
- Citation = Trust: Students need to see sources to trust AI-generated content
- Progressive Disclosure: Don't overwhelm users—show tools only when module unlocked
- Gemini's Multimodal Power: Processing PDFs + slides + videos in one API is game-changing
Log in or sign up for Devpost to join the conversation.