End-to-End Event-Driven MLOps Pipeline on Kubernetes. This project automates the entire ML lifecycle: from experimentation and model tracking with MLflow, to automated retraining triggered by MinIO S3 events using Argo Workflows & Argo Events, and seamless CI/CD deployment with GitHub Actions.

A 6-week hands-on masterclass in production MLOps engineering. Build a file-backed experiment tracker, a containerized model registry, an inference server with dynamic batching, a cached pipeline DAG, automated drift detectors (PSI/KS), and high-performance LLM serving infra (KV-cache/continuous batching) from scratch.