Subham Nayak

I am a machine learning engineer who builds production-grade ML pipelines, AI research agents, and scalable backend systems. My work blends model development, robust MLOps, and practical engineering to deliver reliable, observable systems.

I designed end-to-end solutions including XGBoost-based churn prediction with MLflow tracking, Optuna hyperparameter tuning, and FastAPI/Gradio serving, ensuring train-serve parity and reproducible model artifacts.

I build agentic AI workflows using LangChain and LangGraph integrated with Google Gemini and external search APIs to extract structured intelligence, and I implement verification and automation features like SMTP verification and Google Sheets integrations.

I complement ML work with backend and infrastructure expertise: Docker, Kubernetes, PostgreSQL, Redis, BullMQ, Prisma, CI/CD, and monitoring with Prometheus/Grafana to keep systems scalable, observable, and resilient.