ML Model Serving Architectures — Vocabulary

Practise vocabulary for ML serving topologies: online vs. batch inference, model gateways, streaming inference, and hardware choices.