Azure Kubernetes Cost Optimization Strategy

I run several workloads on Azure Kubernetes Service and the monthly compute bill has started to creep up. I want a seasoned Azure practitioner to review the way my cluster is consuming compute, with a particular emphasis on how we monitor resource usage. At the moment we rely on the default metrics surfaced in the portal, but that only tells part of the story.

Your job is to examine my existing AKS node pools, pod scheduling, and any horizontal or vertical scaling rules, then design a concrete optimisation plan. I am specifically interested in:

• Setting up or refining a robust monitoring stack (Azure Monitor, Prometheus, Grafana, Metrics-Server—whichever combination makes most sense) so I can see at a glance which nodes or pods are consistently under- or over-provisioned.
• Translating those insights into right-sizing recommendations, whether that means different VM SKU choices, mixed-size node pools, or smarter cluster-autoscaler settings.
• Highlighting quick wins I can implement within days and outlining longer-term adjustments that might require code or architecture changes.

Deliverables I expect
1. A written report (PDF or Markdown) that walks through current findings, estimated monthly savings, and step-by-step actions.
2. A lightweight spreadsheet or dashboard snapshot showing the key metrics you used to reach your conclusions.
3. A 30-minute screen-share session where you present the plan, answer questions, and make sure I can carry it forward.

If you can demonstrate prior success reducing AKS compute spend through better monitoring and tuning, please tell me briefly what tools and techniques you used. I’m ready to start as soon as you are.