Inference Dashboard

After the model has been deployed, it becomes available via API, ready to receive and respond to requests in real time. All inferences executed in the live environment from that moment onward are shown in the deployment dashboard.

It shows inferences split according to the outcome class, across a variety of timescales. These are the present and previous day, week, and month. There is also a table of the values alongside the visual display.

The bar chart view can be toggled between stacked and grouped, with stacked bars being the default setting.