ViduraCID Tools

Cost Analysis Capabilities

Vidura's CID tools give the agent access to AWS cost data through the Cost Intelligence Dashboard (CID) views stored in Athena. The agent can answer cost questions across EC2, RDS, storage, serverless services, and commitment instruments (Savings Plans and Reserved Instances).

EC2 Cost Analysis

  • Instance type breakdown: Cost per instance type (m5.xlarge, c5.2xlarge, etc.) with week-over-week trend comparison
  • Purchase option split: On-Demand vs. Reserved Instance vs. Savings Plan cost allocation, showing how much of your EC2 spend is committed
  • Week-over-week trends: Absolute and percentage change in EC2 spend by instance family or region
  • Rightsizing recommendations: Compute Optimizer suggestions for undersized or oversized instances, with estimated monthly savings per recommendation

Example query: "What are my top 5 most expensive EC2 instances this month?"

RDS Cost Analysis

  • Instance costs by engine and type: MySQL, PostgreSQL, Aurora, Oracle, and SQL Server cost breakdown by instance class
  • Multi-AZ vs. Single-AZ delta: The premium paid for Multi-AZ deployments vs. equivalent Single-AZ — helps identify over-provisioned availability configurations in non-production environments
  • RDS rightsizing: Compute Optimizer recommendations for RDS instances with low CPU or memory utilization

Example query: "Which RDS instances are over-provisioned according to Compute Optimizer?"

Storage Cost Analysis

  • S3 by storage class: Cost breakdown across Standard, Infrequent Access (IA), Glacier, and Glacier Deep Archive. Shows actual savings from tiered storage vs. all-Standard baseline.
  • EBS by volume type: gp2, gp3, io1, io2, st1, sc1 cost per volume with IOPS and throughput utilization context
  • EBS snapshot aging: Snapshots older than 90 days with estimated deletion savings — identifies forgotten snapshots from decommissioned instances
  • GP2 to GP3 migration: Volumes still on gp2 with estimated savings from migrating to gp3 (typically 20% cheaper for equivalent IOPS)

Example query: "Which S3 buckets are using Glacier and how much am I saving vs. Standard?"

Lambda and Serverless

  • Memory rightsizing: Lambda Compute Optimizer recommendations for over-allocated memory configurations
  • Function-level cost breakdown: Top Lambda functions by invocation cost and duration cost, with duration billed in GB-seconds
  • Architecture savings: ARM (Graviton2) vs. x86 cost comparison for functions that can be migrated

Savings Plans and Reserved Instances

  • Current SP/RI coverage rate: Percentage of eligible spend covered by committed pricing instruments. Low coverage means more spend at on-demand rates.
  • Unutilized capacity: Reserved Instances or Savings Plans that are not being fully utilized — wasted commitment spend
  • Purchase recommendations: Based on trailing 30-day usage patterns, the agent recommends new SP/RI commitments with estimated 1-year and 3-year savings

Example query: "What is our current Savings Plan coverage rate and what should we buy?"

Agent Response Format

For all cost analysis responses, the agent follows its system prompt structure: Executive Summary, Key Findings table, Top Opportunities ranked by savings, KPI Scorecard, and Recommended Actions prioritized by ROI. All savings estimates are shown in both $/month and $/year.

Example Agent Interaction

User QueryTools Called
"Why did our AWS bill spike this week?"query_cid_view(summary_view) → query_cid_view(cid_ec2_running_cost) → query_cid_view(co_ec2_rightsizing)
"What's our Savings Plan coverage?"query_cid_view(kpi_sp_coverage) → query_cid_view(kpi_ri_coverage)
"Find storage optimization opportunities"query_cid_view(s3_view) → query_cid_view(ebs_view) → query_cid_view(snapshot_view)