🧠 StudyFlow AI

From the intention to learn to real progress, guided by AI.

StudyFlow AI is an agentic learning coach that transforms vague educational goals into structured, personalized study plans with actionable tasks, curated resources, and adaptive progress support.

✨ Features

• 🎯 Smart Goal Analysis — Enter any learning goal and get assessed
• 📝 Diagnostic Quiz — AI-generated questions to gauge your level
• 📋 Personalized Study Plans — Milestone-based roadmaps tailored to your time and level
• 📚 Curated Resources — Ranked learning materials from Khan Academy, MIT OCW, 3Blue1Brown, and more
• ✅ Task Tracking — Daily tasks with progress monitoring
• 🔄 Adaptive Planning — Automatic plan adjustment when you get stuck or advance faster
• 🧠 AI Coaching — Encouragement and guidance from a progress-aware coach

🏗️ Architecture

StudyFlow AI uses a multi-agent system powered by LangGraph:

Agent	Role
Planning Agent	Interprets goals, assesses level, creates milestones
Research Agent	Finds and ranks learning resources
Coaching Agent	Converts milestones into daily actionable tasks
Progress Agent	Monitors completion, detects bottlenecks, adapts the plan

🛠️ Tech Stack

• LLM: Qwen3-30B-A3B (via Hugging Face Inference API)
• Agent Orchestration: LangGraph
• Vector DB: Qdrant (in-memory)
• Embeddings: FastEmbed (BGE-small-en-v1.5)
• Backend: FastAPI
• Frontend: Gradio
• Database: SQLite
• Deployment: Hugging Face Spaces

🚀 Quick Start

1. Clone the repository

git clone https://github.com/your-username/StudyFlow-AI.git
cd StudyFlow-AI

2. Create a virtual environment

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies

pip install -r requirements.txt

4. Set up environment variables

cp .env.example .env
# Edit .env and add your HF_TOKEN

5. Run the application

python app/main.py

Open http://localhost:7860 in your browser.

📁 Project Structure

StudyFlow-AI/
├── app/                    # Gradio frontend
│   ├── main.py            # Entry point
│   ├── theme.py           # Custom theme + CSS
│   └── components/        # UI components
├── backend/               # FastAPI backend
│   ├── server.py          # REST API
│   ├── models/            # Pydantic schemas + DB
│   └── services/          # LLM client + session mgmt
├── agents/                # LangGraph agent system
│   ├── graph.py           # Workflow definition
│   ├── state.py           # Shared state schema
│   └── nodes/             # Agent implementations
├── rag/                   # RAG + Qdrant
│   ├── embeddings.py      # Embedding model
│   ├── indexer.py         # Resource indexing
│   └── retriever.py       # Semantic search
└── data/                  # Seed data
    └── sample_resources.json

📄 License

MIT License

👥 Team

Built for the AMD Hackathon 2026.