ML Model Serving (D2)
Online inference with a feature store and monitoring.
Rendering…
Make it your own.
client -> gateway: request
gateway -> model_server: features
model_server -> model: predict
model -> model_server: score
model_server -> gateway: response
gateway -> client
model_server -> feature_store: fetch
model_server -> monitor: log
feature_store.shape: cylinder
monitor.shape: page