Learning Intelligence System (LIS)

Adaptive AI-Powered Learning Platform with Cognitive Knowledge Tracing, Spaced Repetition, Dynamic PDF Ingestion, and Grounded Material Recommendations.

---

🌟 Overview

The Learning Intelligence System (LIS) is a modern, responsive educational web platform designed to adapt dynamically to every learner's cognitive pace. Built with Next.js 15, React 19, TypeScript, and Google Gemini, it combines cognitive science algorithms (Bayesian Knowledge Tracing and SuperMemo SM-2) with generative AI to deliver personalized curriculums, instant error remediation, synchronized multimedia lessons, and weekly behavioral analytics.

---

🚀 Key Features

1. 🧠 Cognitive Mastery & Bayesian Knowledge Tracing (BKT)

• Models student mastery $P(L_t)$ per Knowledge Component based on correct/incorrect responses, slip, and guess probabilities.
• Re-evaluates mastery in real-time and personalizes downstream lesson pathways.

2. ⚡ Instant AI Error Remediation

• When a learner answers a quiz question incorrectly, Gemini identifies the root-cause misconception, provides a tailored real-world analogy, and generates a targeted follow-up practice problem.

3. 📄 Dynamic PDF Course Ingestion

• Upload any textbook or document (e.g. Criminal-Psychology.pdf) to automatically parse chapters with PDF.js and synthesize complete, multi-module structured courses with key takeaways and quizzes.

4. 🌐 Live Web Search & Material Recommender

• Uses Google Search Grounding via the Gemini API to search the live web for books, academic papers, and publishers matching user queries.
• Cross-references queries with active course content to provide direct "Jump to Lecture" navigation.

5. 📊 Weekly Behavioral Analytics & Broader Goalsets

• Weekly Analytics: Week-over-week velocity trends, diagnostic habit flags, and 7-day study cadences.
• Broader Goalset Tracker: Track custom certification goals (e.g. CFBA, State Bar Evidence Specialist), add custom subjects/topics, set target hours, and monitor mastery readiness.

6. 🗂️ Spaced Repetition Flashcards (SM-2)

• Interactive flashcard review deck using the SuperMemo SM-2 algorithm to schedule review intervals and optimize memory retention.

7. 🎬 Multimedia Lessons & Synchronized Transcripts

• Supports rich text lessons and video masterclasses with interactive video playback, click-to-seek synchronized transcripts, and formative quizzes.

8. 🔑 Client-Side Gemini BYOK & Active Model Selection

• Safe client-side Bring-Your-Own-Key (BYOK) architecture.
• Whitelist of active models: Gemini 3.7 Flash (default), Gemini 3.5 Flash Lite, Gemini 3.1 Pro Preview, Gemini 2.5 Pro, and Gemini 2.5 Flash.

---

🛠️ Technology Stack

• Framework: Next.js 15 (App Router) & React 19
• Language: TypeScript 5.7
• Styling: Custom Vanilla CSS Design System (sharp micro-geometry, warm alabaster canvas #FAF9F6, terracotta #EA580C, electric cobalt #0052FF, veridian #059669)
• AI SDK: Official Google GenAI SDK (@google/genai) with Google Search Grounding
• Document Processing: pdfjs-dist
• Icons: lucide-react
• Infrastructure: 100% Free Tier compatible (Vercel, client-side storage, Google AI Studio)

---

⚡ Quick Start

1. Install Dependencies

npm install

2. Run Development Server

npm run dev

Open http://localhost:3000 in your browser.

3. Build for Production

npm run build
npm run start

---

☁️ Deployment (Vercel)

Deploy to Vercel in 1 click:

npx vercel

Or connect your repository directly in the Vercel Dashboard. No server-side environment variables are strictly required since API keys and profile settings are securely configured client-side by each learner.