🎓 EduFocus: Adaptive Academic RAG Assistant

EduFocus is an intelligent, grounded Retrieval-Augmented Generation (RAG) educational platform that converts static academic materials (PDFs) into an interactive, personalized learning environment.

Designed to adapt dynamically to a student's technical proficiency, EduFocus AI provides level-tailored explanations, verifiable page-level citations, zero-hallucination guardrails, and automated self-assessment quizzes.

---

✨ Key Features

• 🧠 Adaptive Learning Modes

  • Beginner Mode: Explains complex concepts using simple language, relatable analogies, and intuitive examples.
  • Intermediate Mode: Provides technically accurate, academically rigorous explanations for deeper understanding.

• 📚 Grounded Page-Level Citations

  • Every response includes verifiable citations in the format:

    [Source: document.pdf, Page X]
    

  • Enables students to validate answers directly from the uploaded study material.

• 🛡️ Hallucination-Resistant Responses

  • Strict prompt engineering ensures answers are generated only from the retrieved document context.
  • If the required information is unavailable, the assistant clearly states that it is not present in the uploaded material.

• ⚡ Real-Time AI Streaming

  • Powered by Groq's llama-3.1-8b-instant for low-latency, token-by-token response generation using Streamlit streaming.

• 📝 Automatic Quiz Generation

  • Uses llama-3.3-70b-versatile to generate contextual 5-question MCQ quizzes directly from the uploaded academic content.

• 🔄 Context-Aware Chat Memory

  • Maintains a rolling 3-turn conversation history (6 messages) to preserve context while keeping token usage efficient.

---

🛠️ System Architecture & Tech Stack

EduFocus is engineered entirely with high-performance, 100% free and open-source technologies:

Layer	Component	Technology	Role & Architecture
Frontend	User Interface		Pure Python UI with real-time st.write_stream token rendering and tabbed dashboards
Framework	Orchestration		Document chunking pipelines, prompt management, and rolling chat memory buffers
Embeddings	Local Vector Math	sentence-transformers	all-MiniLM-L6-v2 running CPU-locally for zero-cost, high-velocity semantic vector generation
Vector DB	Local Storage		Binary vector index stored locally for instant similarity search without cloud DB overhead
Inference	LLM Cloud API		• llama-3.1-8b-instant: Ultra-fast streamed chat (<2.5s latency) • llama-3.3-70b-versatile: Complex quiz generation

┌─────────────────┐       ┌────────────────────┐       ┌──────────────────────┐
│   PDF Uploads   │ ────► │ PyPDF + Chunking   │ ────► │ sentence-transformer │
└─────────────────┘       └────────────────────┘       └──────────┬───────────┘
                                                                  │
┌─────────────────┐       ┌────────────────────┐                  ▼
│ Streamlit UI    │ ◄──── │ Groq Cloud Stream  │ ◄──── ┌──────────────────────┐
│  (Chat + Quiz)  │       │ (Llama-3.1 / 70B)  │       │  Local FAISS Index   │
└─────────────────┘       └────────────────────┘       └──────────────────────┘

📁 Repository Structure

EduFocus/
├── 📄 app.py                      # Main Streamlit UI entry point
├── 📂 utils/
│   ├── ⚙️ loader.py               # Document loading via PyPDFLoader
│   ├── ✂️ splitter.py             # Text chunking (1000 chars, 200 overlap)
│   ├── 🧮 embeddings.py           # CPU sentence-transformer model
│   ├── 🗄️ vector_store.py         # FAISS similarity search utilities
│   ├── 🧠 chat_memory.py          # 3-turn sliding window history logic
│   ├── 🚀 chat_engine.py          # Groq streaming & citation extraction
│   └── 📝 prompts.py             # Adaptive system prompts & quiz formatters
├── 📂 database/
│   ├── 📥 ingestion.py            # Document processing pipeline
│   └── 💾 vector_store.py         # Persistence wrappers
├── 📂 faiss_db/                   # Local vector index storage
├── 📄 requirements.txt            # Python dependencies
└── 🔒 .env                        # Environment variables (API key)

🚀 Quick Start

1️⃣ Clone the Repository

git clone https://github.com/Trisha937/EduFocus.git
cd EduFocus

---

2️⃣ Set Up a Virtual Environment

Windows

python -m venv venv
venv\Scripts\activate

macOS / Linux

python3 -m venv venv
source venv/bin/activate

---

3️⃣ Install Dependencies

pip install -r requirements.txt

---

4️⃣ Configure Environment Variables

Create a .env file in the project root and add your Groq API key:

GROQ_API_KEY=your_groq_api_key_here

---

5️⃣ Launch the Application

streamlit run app.py

---

🌐 Application URL: http://localhost:8501

Open the above URL in your browser after the Streamlit server starts.