I build production-ready machine learning models using Python, Scikit-learn, and PyTorch — from EDA to deployment via Docker. My churn model on 500K+ records achieved AUC 0.84 and reduced churn 12% in 2 quarters. I deliver clean, documented code, model metrics, and actionable insights your team can act on immediately.
✅ Logistic regression, ensemble models, neural nets
✅ Time-series: ARIMA, Prophet
✅ A/B testing & causal inference
✅ Docker-containerised delivery
Starting at: $500/project