How Jetscale Works

Transform cloud costs into savings with AI-powered optimization

Overview

Jetscale automatically discovers cost-saving opportunities in your cloud infrastructure and delivers them as production-ready Terraform code. This guide explains the Jetscale platform workflow: what happens when you connect your cloud account and how you get actionable savings.

The Jetscale Process

graph LR
    A[Connect Cloud] --> B[Discover Resources]
    B --> C[Analyze Usage]
    C --> D[AI Generates Recommendations]
    D --> E[Review & Select]
    E --> F[Deploy via Terraform]
    F --> G[Track Savings]

3-Step Workflow

1. Connect → Grant read-only cloud access
2. Review → Browse AI-generated recommendations with savings estimates
3. Deploy → Get Terraform code via pull request, merge, and save

1. Connect Your Cloud Account

What You Do:

• Grant Jetscale read-only access to your AWS or Azure environment
• No write permissions. Jetscale never modifies your infrastructure directly
• Takes 5 minutes with our step-by-step guides

How Security Works:

• AWS: Cross-account IAM role with External ID verification
• Azure: Service Principal with certificate authentication
• Zero credentials stored: Everything uses temporary, scoped tokens
• Read-only permissions: Jetscale can only view resource metadata and metrics

What Jetscale Needs:

• AWS: IAM Role ARN and External ID
• Azure: Tenant ID, Subscription ID, Service Principal credentials

AWS Setup Guide | Azure Setup Guide

2. Discover Resources

What Jetscale Finds:

Compute Resources:

• EC2 instances (all types and families)
• Azure Virtual Machines
• Auto Scaling Groups
• Reserved Instance coverage gaps

Databases:

• RDS instances (all engines: MySQL, PostgreSQL, SQL Server)
• Aurora clusters
• Azure SQL databases
• Read replicas and multi-AZ configurations

Storage:

• EBS volumes (all types: gp2, gp3, io1, io2)
• Snapshots and backups
• Azure Managed Disks

Caching:

• ElastiCache (Redis, Memcached)
• Azure Cache for Redis

Discovery Process:

• Scans all regions in your account automatically
• Catalogs resource configurations and relationships
• Identifies tagging patterns and organizational structure
• Updates daily to catch new resources and changes

3. Analyze Usage Patterns

Historical Data Collection:

Jetscale pulls comprehensive usage data from:

• AWS CloudWatch or Azure Monitor metrics
• Cost Explorer or Azure Cost Management APIs
• Performance metrics (CPU, memory, network, disk I/O)
• Cost data (current spend, RI utilization, Savings Plans)

Performance Metrics Tracked:

• CPU utilization (average, p50, p95, p99, max)
• Memory utilization and pressure
• Network throughput and packet rates
• Disk IOPS and throughput
• Database connections and query patterns
• Cache hit/miss ratios

Pattern Recognition:

Jetscale's AI classifies workloads:

4. AI-Generated Recommendations

What Makes Jetscale Different:

Unlike generic cost tools that simply alert you to low CPU usage, Jetscale's AI understands the nuances of each cloud service:

For Databases:

• Multi-AZ failover patterns and capacity requirements
• Read replica lag tolerances
• Connection pooling behavior
• Query workload characteristics

For Compute:

• Burstable instance credit balance patterns
• Auto Scaling Group headroom requirements
• Load balancer health check configurations
• Graviton migration compatibility

For Storage:

• IOPS burst patterns vs sustained throughput
• Volume type performance characteristics
• Snapshot lifecycle optimization

Recommendation Types:

Every Recommendation Includes:

• Estimated monthly savings: Detailed cost breakdown
• Performance impact assessment: Zero-downtime or minor adjustments noted
• Implementation risk level: Low, Medium, or High with specific risks
• Before/after comparison: Side-by-side configuration view
• Supporting evidence: Historical usage charts and metrics
• Rollback instructions: Easy path back if needed

Safety Validation:

Before recommending any change, Jetscale validates:

• Minimum 20% CPU headroom above historical peak
• Minimum 15% memory headroom above peak usage
• High-availability requirements preserved (multi-AZ, replicas)
• Performance SLAs maintained
• Sufficient capacity for traffic spikes and growth

5. Review & Select Recommendations

Dashboard View:

Recommendations are organized by:

• Potential savings (highest to lowest)
• Risk level (prioritize low-risk quick wins)
• Resource type (EC2, RDS, EBS, etc.)
• Cloud account (if managing multiple accounts)

Recommendation Details:

Click any recommendation to see:

• Summary: What's changing and why
• Evidence: Historical usage charts
• Configuration Comparison: Current vs recommended settings
• Savings Calculation: Detailed cost analysis
• Impact Assessment: Performance and availability considerations
• Implementation Notes: Prerequisites and special considerations

Bulk Selection:

• Select multiple recommendations for batch implementation
• Filter by criteria (e.g., "all low-risk recommendations > $100/month")
• Preview combined savings across selections

6. Deploy Changes

Production-Ready Terraform:

When you approve recommendations, Jetscale generates:

• Standard HCL format: Works with Terraform 0.13+
• Lifecycle blocks: Safe deployments with create-before-destroy
• Documentation: Inline comments explaining each change
• Rollback instructions: Easy recovery if needed
• Savings tracking: Tags for ROI reporting

Example Generated Code:

# Jetscale Recommendation: REC-2025-001
# Estimated Monthly Savings: $248.16
# Implementation Risk: LOW

resource "aws_instance" "web_server_01" {
  instance_type = "t3.large"  # Changed from t3.xlarge

  lifecycle {
    create_before_destroy = true
  }

  tags = {
    Name                      = "web-server-01"
    JetscaleRecommendation    = "REC-2025-001"
    JetscaleMonthlySavings    = "248.16"
    JetscalePreviousType      = "t3.xlarge"
  }
}

Integration Options:

GitHub/Bitbucket:

• Automatic pull request creation in your repository
• PR includes full documentation, cost analysis, and testing checklist
• Reviewers auto-assigned based on your team settings
• Links back to Jetscale dashboard for detailed evidence

Jira:

• Automatic ticket creation for tracking
• Custom fields for savings and ROI
• Status syncing with pull requests
• Executive reporting integration

Manual Download:

• Download Terraform files as ZIP
• Apply using your existing Terraform workflow
• Full flexibility for custom processes

GitHub Integration Guide | Jira Integration Guide

7. Track Savings

Continuous Monitoring:

After deployment, Jetscale tracks:

• Actual savings vs projected: Verify accuracy of recommendations
• Performance metrics post-change: Ensure no degradation
• Resource utilization trends: Spot new optimization opportunities
• Cumulative savings over time: Demonstrate ROI to stakeholders

Reporting & Analytics:

• Monthly savings dashboard with trend analysis
• Year-to-date totals and ROI calculations
• Resource-level savings attribution
• Export data for executive reporting and FinOps

Continuous Discovery:

• Daily scans for new optimization opportunities
• Alerts when high-value recommendations are available
• Trend analysis to identify growing waste
• Seasonal pattern detection for scheduled optimizations

Security & Trust

What Jetscale Stores

Data We Keep:

• Resource metadata (instance IDs, types, configurations)
• Aggregated usage metrics (configurable retention period)
• Recommendation history and approval status
• User preferences and team settings

Data We Never Store:

• Application data or business logic
• Database contents or customer data
• Source code or intellectual property
• IAM credentials (only temporary tokens used)

Encryption & Compliance

Data Protection:

• At rest: AES-256 encryption
• In transit: TLS 1.3 for all connections
• Key management: Separate encryption keys per customer

Compliance Certifications:

• SOC 2 Type II certified
• GDPR compliant (data residency options available)
• CCPA compliant
• Regular security audits by third-party firms

Access Control

Read-Only Permissions:

• Jetscale only needs read access to resource configs and metrics
• No write, delete, or modify permissions granted
• All changes reviewed and deployed by you
• Your code stays in your repositories

Audit Trail:

• All API calls logged and available for review
• CloudTrail/Azure Activity Log integration
• Access logs provided on request for security reviews

Why Jetscale Saves You Time

Without Jetscale:

• Manual CloudWatch/Azure Monitor analysis: 8-12 hours/month
• Researching right-sizing options: 4-6 hours/month
• Writing Terraform code: 6-8 hours/month
• Testing and validation: 4-6 hours/month
• Total: 22-32 hours/month per engineer

With Jetscale:

• Review recommendations: 30 minutes/month
• Approve and merge PRs: 30 minutes/month
• Monitor results: 15 minutes/month
• Total: 75 minutes/month
• Time saved: 95%

Next Steps

Get Started:

1. Connect AWS Account (5 minutes)
2. Connect Azure Account (5 minutes)
3. Configure Integrations (GitHub, Jira, Slack)

Learn More:

• AI Analysis Deep Dive - How our AI generates recommendations
• Supported Services - What Jetscale optimizes
• FAQ - Common questions answered

Need Help?

• Email: support@jetscale.ai
• Live chat in dashboard
• Schedule Demo

Last Updated: January 29, 2025