Exploring common use cases for Paver

Review the following common use cases and example goals to see how Paver assists across your infrastructure lifecycle. When you prompt Paver, provide your goals, specific constraints, and ask for a plan before applying changes.

You can interact with IBM® Paver in natural language to handle routine operational tasks, investigate complex environment issues, and execute governed updates across your infrastructure.

Deploy infrastructure changes#

Describe the change that you need, and get a plan that you can review before anything is created or modified. Paver builds execution plans, identifies dependencies, and surfaces proposed updates for your approval before making changes.

Example goals:

• Deploy multitier applications across availability zones with backup and compliance constraints.
• Expand compute and storage capacity based on forecasted utilization.
• Roll out configuration updates across container worker pools without downtime.

Example prompt:

I want to deploy my 3-tier web application (React frontend, Java backend, PostgreSQL database) with high availability. All data must reside in the eu-de region and include automated daily backups. Generate a deployment plan in plan.md for my review before creating any resources.

Investigate and resolve incidents#

Tell Paver what's slow or failing, and get a likely cause and next steps. Paver correlates system logs, performance metrics, active alerts, and recent environment changes to identify probable root causes and guide you through resolution.

Example goals:

• Triage sudden latency spikes across production microservices.
• Determine the root cause of service errors following a recent configuration update or deployment.
• Correlate container restart loops with platform metric anomalies and logs.

Example prompt:

Production API latency has increased by 40% over the last hour for the checkout service. Review recent deployment events, correlate error logs from @logs.txt, and summarize the probable root causes before suggesting mitigation steps.

Remediate security vulnerabilities#

Point Paver at your environment to find what's exposed, and get a prioritized fix plan that you can review. Paver assesses severity, generates actionable remediation plans, and guides or executes approved fixes to secure affected resources.

Example goals:

• Identify and prioritize resources that are impacted by critical CVEs or outdated base images.
• Remediate misconfigured security groups or publicly exposed storage buckets.
• Rotate compromised credentials and verify that dependent services remain healthy.

Example prompt:

Scan our public-facing subnets in the production VPC for open administrative ports (such as SSH and RDP). Identify noncompliant security group rules and draft a remediation plan with the exact rule changes before applying them.

Audit access and permissions#

Ask who has access to what, and see where permissions are broader than they need to be. Paver evaluates active permissions and role bindings, highlights compliance gaps, and recommends least-privilege configurations.

Example goals:

• Enforce least-privilege access by identifying inactive or overly permissive service IDs and role bindings.
• Prepare compliance evidence and audit reports for access control reviews.
• Validate separation of duties across production and development environments.

Example prompt:

Review all IAM service IDs with Administrator or Operator roles in our production account. Flag any IDs inactive for over 90 days and generate an audit summary table with recommended privilege reductions.

Optimize cost and resource consumption#

Ask where you're spending more than you need to, and get a list of savings opportunities. Paver evaluates usage patterns and spots idle or oversized resources so you can reduce costs without compromising availability or performance.

Example goals:

• Locate unattached storage volumes, idle compute instances, and unassociated floating IPs.
• Rightsize underutilized worker nodes or database instances based on 30-day utilization metrics.
• Analyze billing trends across resource groups to identify cost anomalies.

Example prompt:

Analyze compute and storage utilization across the staging environment over the past 30 days. Identify idle or oversized instances and provide a cost optimization plan with estimated monthly savings and impact analysis.

Assess resiliency and failure impact#

Ask what would break if a component or zone went down, before it happens. Paver maps resource relationships and models disruption scenarios so you can find single points of failure, understand the blast radius, and improve fault tolerance.

Example goals:

• Map cross-service dependencies to identify single points of failure across workloads.
• Simulate an availability zone outage to evaluate failover readiness and disaster recovery procedures.
• Inspect storage redundancy and cross-region replication status across critical databases.

Example prompt:

Analyze the architecture of our payment processing workload across availability zones in us-south. Identify any single points of failure, verify cross-zone database replication status, and create a resiliency assessment report.