All work
Cloud Infrastructure & SRE

Cloud Cost Governance & Resource Hygiene Automation

Read-only reporting tools across Azure and Google Cloud that automatically surface cost-optimization opportunities and abandoned/unused cloud resources.

Azure + Google Cloud

Cloud platforms audited

Zero — fully read-only

Mutating calls made

Markdown + CSV reports

Deliverables

The Problem

Cloud spend and resource sprawl are easy to lose track of at scale — unused disks, oversized VMs, and forgotten projects quietly accumulate cost and attack surface. Doing this analysis by hand in the portal doesn't scale and doesn't get done regularly.

My Approach

  • Built an Azure cost optimization report generator: a read-only tool that queries the Cost Management API, Azure Resource Graph, and Azure Advisor, then synthesizes results into a single Markdown report — cost breakdown by service and resource group, month-over-month trend, and a prioritized list of actionable savings opportunities (right-sizing/decommissioning underutilized VMs, deleting orphaned managed disks, reservation/savings-plan opportunities). Explicitly designed as read-only: no mutating calls, safe to run anytime.
  • Built a GCP project inventory & cleanup analysis script that pulls the full list of cloud projects, analyzes creation dates, naming patterns, and lifecycle state, and produces a CSV audit deliverable identifying likely abandoned or proof-of-concept projects that are candidates for decommissioning.
  • Framed every recommendation with an explicit confidence caveat (e.g., "Advisor estimate — validate before purchasing a reservation") so the report supports a decision rather than pretending to make one automatically.

Stack

Cloud APIs

Azure Cost Management APIAzure Resource GraphAzure AdvisorGoogle Cloud Resource Manager API

Languages

Python

Output/Tooling

Markdown report generationCSV/JSON exportpandas-style data processing

Practices

Read-only/non-mutating automation designFinOps reporting

Skills Demonstrated

  • ▸FinOps: translating raw billing/usage data into prioritized, actionable recommendations for non-technical stakeholders
  • ▸Safe-by-design automation (read-only tooling for sensitive environments)
  • ▸Comfort working directly with cloud provider billing and resource-graph APIs
  • ▸Communicating uncertainty/confidence honestly in a technical report