Drift

Model 05
HTTP/2gRPCgRPCPartner apps20 req/sAPI server1 vCPU · 2 GBModel (CPU)2 vCPU · 4 GBModel train…1 vCPU · 1 GBModel regis…1 vCPU · 1 GB
HTTP/2gRPCgRPCPartner apps20 req/sAPI server1 vCPU · 2 GBModel (CPU)2 vCPU · 4 GBModel train…1 vCPU · 1 GBModel regis…1 vCPU · 1 GB

Fix the board. Turn every goal green.

Traffic
20 req/s · reads only
Limit
API server answers about 3,300 req/s
Bill
  • API server8 USD/mo
  • Model (CPU)16 USD/mo
  • Model trainer4 USD/mo
  • Model registry4 USD/mo
32 of 200 USD/mo

Goals

Not tested yet

Time from a fault until an alert fires.The share of requests that fail or get no answer.What the servers and services cost each month.

A model predicts which orders will ship late. It was trained once, before day 0. Over 90 days the order mix changes twice. Keep accuracy above 90 percent over the run and find each change within 3 days. You can make 3 edits.

Day 41, 09:00. Late-shipping model accuracy is 0.90 and has fallen for 26 days. No alert fired. Graph: daily accuracy.

One change feed keeps serving in stepTap or press Space to change the write paths

Two write paths · 4 of 16 served features differ from training