By far this is one of my favorite topics "How not to MLOps".

Our MLOps platform addresses these antipatterns by providing ML CI/CD, dashboard for monitoring data/model drifts, data linkage and versioning, and governance. Interested in exploring?