Study Buddy

A modern, end-to-end LLMOps quiz and study system that leverages large language models and advanced system design patterns for maintainability, extensibility, and robustness. This project covers the full lifecycle from LLM-powered question generation to deployment and monitoring, following best practices in LLMOps.

🚀 Features

• End-to-End LLMOps: Complete workflow from prompt engineering, LLM-powered question generation, evaluation, to deployment and monitoring.
• LLM-Powered Quiz Generation: Uses large language models to generate multiple-choice and fill-in-the-blank questions on any topic and difficulty.
• System Design Patterns: Incorporates Factory, Strategy, and Template Method patterns for clean, modular, and extensible code.
• Beautiful Streamlit UI: Interactive, user-friendly web interface for quiz taking and results review.
• Session Management: Uses Streamlit session state for persistent, multi-step workflows.
• Result Export: Download quiz results as CSV files.

🏗️ System Design

• Factory Pattern: Centralized question creation logic for different question types.
• Strategy Pattern: Pluggable evaluation and UI logic for each question type.
• Template Method Pattern: Modular quiz workflow steps (generate, attempt, evaluate).
• Separation of Concerns: Business logic, UI, and LLM integration are cleanly separated.

🛠️ Tech Stack

• Python 3.12
• Streamlit (frontend/UI)
• LangChain (LLM integration)
• Pydantic (data validation)
• Kubernetes (container orchestration)
• GCP VM (cloud hosting)
• Jenkins (CI pipeline)
• ArgoCD (CD pipeline)

⚡ Setup

1. Clone the repo:

   git clone https://github.com/yourusername/study_buddy.git
   cd study_buddy

2. Create and activate a virtual environment:

   python3 -m venv venv
   source venv/bin/activate

3. Install dependencies:

   pip install -r requirements.txt

4. Set up environment variables:

   • Copy .env.example to .env and fill in your API keys and settings.

5. Run the app locally:

   streamlit run app.py

☸️ Deployment

• Kubernetes:
  • Deploy the app using the provided Dockerfile and Kubernetes manifests (manifests/).
• GCP VM:
  • Host your Kubernetes cluster or Jenkins/ArgoCD runners on Google Cloud Platform VMs.
• Jenkins (CI):
  • Automated testing, linting, and build pipeline.
• ArgoCD (CD):
  • GitOps-based continuous deployment to Kubernetes.

📁 Project Structure

src/
  models/           # Pydantic models
  generator/        # LLM question factory
  strategy/         # Evaluation and UI strategies
  utils/            # QuizManager and helpers
  prompts/          # Prompt templates
  common/           # Logging, exceptions, constants
  config/           # Settings
app.py              # Streamlit entry point
manifests/          # Kubernetes manifests
Dockerfile          # Container build
.pre-commit-config.yaml

🤝 Contributing

Pull requests are welcome! Please lint and format your code with pre-commit before submitting.

📜 License

MIT License