MLOps / LLMOps Platform with Drift Monitoring

A complete machine-learning operations platform that runs on a laptop CPU. A LightGBM model predicts New York green taxi trip durations and is tracked and versioned in MLflow, served by a FastAPI service with Prometheus metrics, and watched month by month with Evidently drift reports. When drift or a rising error appears, a challenger is trained and promoted only if it passes a quality gate. A local LLM writes the monthly monitoring notes, and its prompts and outputs are versioned and checked like any other model.

Category: Machine Learning Engineering / MLOps / LLMOps

CSE Project 20 from ProTechSC Academy, with graphical abstract, video demonstration and project guidance.