Cost Optimization

Reduce cloud costs with automated waste detection, right-sizing recommendations, and reserved instance optimization.

Overview

The cost optimization engine analyzes your knowledge graph to identify waste, right-sizing opportunities, and reserved instance optimization strategies. By combining resource metadata from discovery with usage patterns from connected monitoring tools, it provides actionable recommendations to reduce infrastructure costs.

Waste detection

The system automatically identifies common sources of cloud waste:

Waste category	Detection method
Orphaned resources	Resources with no relationships for 30+ days
Idle compute	Low CPU/memory utilization over 14 days
Unused storage	Volumes unattached to any instance
Over-provisioned DB	Database instances with low connection counts
Stale snapshots	Snapshots older than 90 days
Unattached IPs	Elastic IPs not associated with any resource

Right-sizing

Right-sizing recommendations analyze resource utilization against available instance types:

Downsize candidates -- resources consistently using less than 20% of capacity
Upsize candidates -- resources consistently at 80%+ utilization
Family migration -- moving to a more cost-effective instance family (e.g., Graviton)
Savings estimate -- projected monthly savings for each recommendation

# Right-sizing recommendation
{
  "resource_id": "i-abc123",
  "current_type": "t3.xlarge",
  "recommended_type": "t3.large",
  "utilization": {
    "cpu_avg": 12.4,
    "memory_avg": 34.2,
    "network_avg": 8.1
  },
  "monthly_savings": 42.50,
  "annual_savings": 510.00,
  "confidence": "high",
  "risk": "low"
}

Reserved instances

The RI optimization module analyzes usage patterns and recommends optimal reserved instance purchases:

Coverage analysis -- current RI coverage vs. optimal coverage
Purchase recommendations -- which instance types, regions, and terms to purchase
Payment options -- no upfront, partial upfront, all upfront comparisons
Savings plans -- compute savings plan vs. instance type RI analysis
Expiration tracking -- RI expiration dates with renewal recommendations

Cost trends

Cost trends track spending patterns over time:

Monthly spend -- cost breakdown by provider, service, and scope
Growth rate -- month-over-month and year-over-year cost growth
Budget tracking -- actual vs. budgeted spend with alerts
Anomaly detection -- unexpected cost spikes flagged automatically

Cost allocation tags

Ensure your cloud resources have proper cost allocation tags. Knowledge Tree can detect resources missing required cost tags and flag them for remediation.

Savings report

The savings report provides a summary of identified opportunities:

# Monthly savings report
{
  "period": "2025-06",
  "total_potential_savings": 3240.50,
  "implemented_savings": 1850.00,
  "by_category": {
    "right_sizing": { "potential": 1200.00, "implemented": 800.00 },
    "waste_removal": { "potential": 890.50, "implemented": 650.00 },
    "ri_optimization": { "potential": 1150.00, "implemented": 400.00 }
  },
  "top_recommendations": [
    {
      "resource": "i-xyz789",
      "action": "downsize to t3.medium",
      "savings": 85.00,
      "status": "pending"
    }
  ]
}