🏠 Advanced Real Estate Price Prediction MLOps Pipeline

An end-to-end production-ready machine learning pipeline to predict California housing prices, built using ZenML, MLflow, BentoML, and GitHub Actions — following modern MLOps best practices: CI/CD, model tracking, experiment management, scalable deployment, and robust monitoring.

---

🎯 Key Highlights

✅ Modular ZenML pipelines for ingestion, preprocessing, training, evaluation, and deployment

✅ MLflow integration for experiment tracking & model registry

✅ Train and compare multiple models (Linear, Ridge, Lasso, Random Forest)

✅ Automated evaluation & comparison

✅ BentoML service for real-time REST API predictions

✅ Fully automated CI/CD with GitHub Actions (linting, testing, deploying, rollback)

✅ Code quality checks & health monitoring

---

🧠 Tech Stack

Tool	Purpose
ZenML	Pipeline orchestration
MLflow	Experiment tracking & artifact logging
BentoML	Model packaging, serving & API
GitHub Actions	CI/CD automation
Scikit-learn	ML models & preprocessing
Pytest + Black	Testing & code formatting

---

🏛️ High-Level Architecture

graph TD
  A[Raw Data] --> B[ZenML Pipeline]
  B --> C[MLflow Tracking]
  C --> D[Trained Model]
  D --> E[BentoML Service]
  E --> F[REST API Endpoint]
  F --> G[Client / User]

Loading

---

📂 Project Structure

property-price-prediction/
├── src/
│   ├── data_loader.py
│   ├── feature_engineering.py
│   └── utils.py
├── steps/
│   ├── preprocessing.py
│   ├── model_trainer.py
│   ├── evaluator.py
│   └── deployment.py
├── bentoml/
│   ├── bentoml_service.py
│   └── bentofile.yaml
├── .github/
│   └── workflows/
│       └── mlops-pipeline.yml
├── zenml_pipeline.py
├── enhanced_zenml_pipeline.py
├── training_pipeline.py
├── requirements.txt
├── pyproject.toml
└── README.md

---

⚙️ Local Development

# 📥 1️⃣ Clone the repo & set up virtual environment
git clone https://github.com/priyanshumishra610/Advanced-Real-Estate-Price-Prediction.git
cd property-price-prediction
python -m venv venv && source venv/bin/activate  # Windows: venv\Scripts\activate
pip install -r requirements.txt

# 🔗 2️⃣ Initialize ZenML & run pipeline
zenml init
python zenml_pipeline.py
# or
python enhanced_zenml_pipeline.py

# 📊 3️⃣ Start MLflow UI for experiment tracking
mlflow ui --host 0.0.0.0 --port 5000
# Visit: http://localhost:5000

# 🚀 4️⃣ Serve REST API with BentoML
bentoml serve bentoml_service:PropertyPriceService
# Visit Swagger UI: http://localhost:3000

---

🔁 CI/CD Workflow

Automated GitHub Actions pipeline on each push:

✅ Code linting & formatting checks

✅ Unit tests with Pytest

✅ Run ZenML pipelines

✅ Build BentoML service

✅ Health checks

✅ Deploy on success

✅ Rollback on failure if needed

Workflow: .github/workflows/mlops-pipeline.yml

---

🌐 Roadmap

Phase	Description	Status
Phase 1	Core ZenML pipelines, MLflow setup	✅ Completed
Phase 2	CI/CD integration with GitHub Actions	✅ Completed
Phase 3	BentoML deployment & serving	✅ Completed
Phase 4	Monitoring, scaling & cloud deployment	🚧 In Progress
Phase 5	Extend with advanced models, feature store	🗂️ Planned

---

💡 Future Enhancements

• Add cloud-native orchestration with Kubernetes & Helm
• Integrate Prometheus + Grafana for monitoring
• Use Feast for feature store management
• Build a simple Streamlit dashboard for user interaction

---

👨‍💻 Author

Priyanshu Mishra

“Building world-class AI pipelines with robust MLOps, one repo at a time.”

🔗 GitHub

---

⭐️ Support & Share

If you find this template valuable — star ⭐️ it, share it, and feel free to fork & contribute!

---

📄 License

MIT License — free for personal & commercial use. Use it. Improve it. Share it.