How do you monitor ML infrastructure cost?

medium

Answer

$(1) Tag resources by (team, project, model, environment).$
$/prediction. (3) Alerts on unexpected cost spikes (>2σ). (4) Budget alerts per team. (5) FinOps dashboard: idle GPU, over-provisioned pods, expensive-but-unused features. (6) Right-sizing recommendations: instance types + autoscale bounds. (7) Cost per experiment (training run). Tools: AWS Cost Explorer, GCP Billing, Kubecost. Business impact per$

Check yourself — multiple choice

Random
Tag by team/project/model/env + $/prediction per model + spike alerts (2σ) + budget alerts + FinOps dashboard (idle GPU/over-provision) + right-size + cost per experiment; AWS/GCP/Kubecost
Not needed
Just count

Cost monitoring: tag + $/prediction + spike alerts + FinOps + right-size.

#monitoring#infrastructure

Practise MLOps & Data Quality

215 interview questions in this topic.

Read the lesson Flashcards Take the quiz

Related questions