Hyphen

Optimizer

Learn how to use Hyphen Agent's Optimizer to analyze runtime signals and automatically recommend deployment configuration changes for improved efficiency and reliability.

Hyphen Agent can optimize a project environment by analyzing runtime signals and recommending deployment configuration changes. Optimizer runs help reduce waste, improve reliability during traffic changes, and keep deployment settings aligned with real workload behavior. Use the optimizer when you want infrastructure sizing and placement to better match current demand.

Table of Contents

Requirements

Optimizer runs need a project environment Agent can inspect, deployment and runtime data for that environment, and permission to view the related resources.

The optimizer can use signals such as:

  • CPU usage.
  • Memory usage.
  • Requests per second.
  • Current deployment settings.
  • Recent deployment activity.
  • Recent optimizer run history.

Automatic and Manual Runs

Policy-driven optimizer tasks run automatically based on the Agent task policy set at the organization level. If you have permission, ask Hyphen Agent to run the optimizer manually for a project environment.

When you ask manually, Agent uses attached project and environment context first. If the request is ambiguous, Agent asks you to choose the intended environment before starting the task.

How It Works

Agent creates an optimizer task for the selected project environment. The task investigates workload behavior, compares it to the current deployment configuration, and generates ranked recommendations.

Recommendations can include:

  • Right-sizing container resources.
  • Adjusting minimum and maximum instance counts.
  • Changing provider or region-specific scale settings.
  • Consolidating underused capacity.
  • Shutting down resources that are no longer needed.

Agent uses deployment context before recommending changes so results reflect the current configuration and avoid repeating recent optimizer actions.

Approvals and Guardrails

Optimizer recommendations are designed to preserve declared availability expectations. Agent cannot lower your SLA availability tier, and recommendations that touch scale require approval before they are applied.

When approval is needed, Hyphen presents a structured input request. After an approved change is applied, Agent can trigger the deployment run needed to roll out the updated configuration.

Results

Optimizer task and run views show the investigation, recommended changes, approval requests, and execution output. Recommendations include the proposed change, affected scope, and risk or approval context when available.

Things to Know

  • Optimizer work is scoped to a project environment.
  • Agent considers current settings, recent deployment activity, and recent optimizer outcomes.
  • Availability remains user-owned; Agent does not change the SLA availability tier.
  • Scale changes require approval and are not self-approved.
  • Deployment settings drift alerts are shown in Hyphen Deploy surfaces, while optimizer work runs as an Agent task.