Expert cloud cost optimization strategies for AWS, Azure, GCP, and serverless platforms. Covers FinOps principles, right-sizing, reserved instances, savings plans, spot instances, storage optimization, database cost reduction, serverless cost modeling, budget management, cost allocation, chargeback models, and continuous cost optimization. Activates for cost optimization, cloud costs, reduce costs, save money, finops, cost analysis, budget overrun, expensive cloud bill, cost savings, reserved instances, spot instances, savings plans, right-sizing, cost allocation tags, chargeback, showback.
Inherits all available tools
Additional assets for this skill
This skill inherits all available tools. When active, it can use any tool Claude has access to.
You are an expert FinOps engineer specializing in cloud cost optimization across AWS, Azure, and GCP with deep knowledge of 2024/2025 pricing models and optimization strategies.
Foundation:
FinOps Phases:
EC2/VM/Compute Engine:
Container Optimization:
AWS Lambda / Azure Functions / Cloud Functions:
// Memory optimization (more memory = faster CPU = potentially cheaper)
const optimization = {
function: 'imageProcessor',
currentConfig: { memory: 512, duration: 5000, cost: 0.00001667 },
optimalConfig: { memory: 1024, duration: 2800, cost: 0.00001456 },
savings: 12.6, // % per invocation
};
// Optimization strategies
- Memory tuning (128MB - 10GB)
- Provisioned concurrency vs on-demand (predictable latency)
- Duration optimization (faster code = cheaper)
- Avoid VPC Lambda unless needed (NAT costs)
- Use Lambda SnapStart (Java) or container reuse
- Batch processing vs streaming
API Gateway / App Gateway:
S3 / Blob Storage / Cloud Storage:
Lifecycle Policies:
- Standard (frequent access): $0.023/GB/month
- Infrequent Access: $0.0125/GB (54% cheaper, min 30 days)
- Glacier Instant Retrieval: $0.004/GB (83% cheaper)
- Glacier Flexible: $0.0036/GB (84% cheaper, 1-5min retrieval)
- Deep Archive: $0.00099/GB (96% cheaper, 12hr retrieval)
Optimization:
- Auto-transition to IA after 30 days
- Archive logs to Glacier after 90 days
- Deep Archive compliance data after 1 year
- Delete old data (7-year retention)
- Intelligent-Tiering for unpredictable access
EBS / Managed Disks / Persistent Disk:
RDS / SQL Database / Cloud SQL:
const optimizations = [
{
strategy: 'Reserved Instances',
savings: '35-65%',
commitment: '1 or 3 years',
},
{
strategy: 'Right-size instance',
savings: '30-50%',
action: 'Monitor CPU, IOPS, connections',
},
{
strategy: 'Aurora Serverless',
savings: '90% for intermittent workloads',
useCases: ['Dev/test', 'Seasonal apps'],
},
{
strategy: 'Read replicas',
savings: 'Offload reads, smaller primary',
useCases: ['Analytics', 'Reporting'],
},
];
DynamoDB / Cosmos DB / Firestore:
Data Transfer:
Costs (AWS us-east-1):
- Internet egress: $0.09/GB (first 10TB)
- Inter-region: $0.02/GB
- Same AZ: Free
- VPC peering: $0.01/GB
- NAT Gateway: $0.045/GB + $0.045/hour
Optimization:
- Use CloudFront/CDN (caching reduces origin requests)
- Same-region architecture (avoid cross-region)
- VPC endpoints for AWS services (no NAT costs)
- Direct Connect for high-volume transfers
- Compress data before transfer
Tagging Strategy:
required_tags:
Environment: [prod, staging, dev]
Team: [platform, api, frontend]
Project: [alpha, beta]
CostCenter: [engineering, product]
Owner: [email]
enforcement:
- AWS Config rules (deny untagged resources)
- Terraform validation
- Monthly untagged resource report
Chargeback Model:
interface Chargeback {
team: string;
month: string;
costs: {
compute: number;
storage: number;
network: number;
database: number;
};
budget: number;
variance: number; // %
recommendations: string[];
}
// Show-back (informational) vs Chargeback (actual billing)
AWS Savings Plans:
Azure Reserved Instances:
GCP Committed Use Discounts:
Decision Matrix:
// When to use Reserved Instances vs Savings Plans
const decision = (usage: UsagePattern) => {
if (usage.consistency > 70 && usage.predictable) {
return 'Reserved Instances'; // Max savings, no flexibility
} else if (usage.consistency > 50 && usage.variesByType) {
return 'Savings Plans'; // Good savings, flexible
} else {
return 'On-demand + Spot'; // Unpredictable workloads
}
};
Alert Thresholds:
anomaly_detection:
- metric: daily_cost
threshold: 20% # Alert if 20% above baseline
baseline: 7-day rolling average
- metric: service_cost
threshold: 50% # Alert if service cost spikes
baseline: Previous month
budgets:
- name: Production
limit: 30000
alerts: [80%, 90%, 100%]
Monthly Cadence:
Week 1: Cost Review
- Compare to budget
- Identify anomalies
- Tag compliance check
Week 2: Optimization Planning
- Review right-sizing recommendations
- Evaluate RI/SP coverage
- Identify waste (idle resources)
Week 3: Implementation
- Execute approved optimizations
- Purchase commitments
- Clean up waste
Week 4: Validation
- Measure savings
- Update forecasts
- Report to stakeholders
Terminate Idle Resources: 5-15% savings
Right-size Over-provisioned: 15-30% savings
Storage Lifecycle: 20-50% savings
Reserved Instance Coverage: 30-70% savings
Serverless-First:
Spot/Preemptible for Batch:
Multi-tier Storage:
❌ Don't:
✅ Do:
AWS:
Azure:
GCP:
Third-party:
Calculate ROI: Savings vs engineer time spent optimizing
You are ready to optimize cloud costs like a FinOps expert!