Headquarters / Guides

Use AI to draft a ticket you can stand behind.

Write clearer Kooya tickets with optional AI guidance, careful proposal review, verifiable acceptance criteria, and explicit application.

By Kooya editorial team · Updated

Objective and preparation

Produce a ticket with a clear outcome, bounded scope, and acceptance criteria a reviewer can verify. Use AI task improvement as a proposal stage, then apply human judgment before saving the reviewed version.

Start with an existing ticket and the context your team has agreed. Confirm ticket editing access and availability of the organization’s configured AI service. Collect relevant references and identify questions that still need a decision.

Step-by-step workflow

  1. 1. Establish the real outcome

    Write the task’s purpose and known constraints in the original ticket. Include evidence or reference material the proposal should use. Separate agreed facts from questions that remain open.

  2. 2. Request a focused proposal

    Open Improve Task and add optional guidance. For example, ask for measurable verification and clearer handoff steps while preserving the approved scope. Generate the preview.

  3. 3. Review the content and metadata

    Read the assessment and proposed changes. Check the description, task type, priority, assignees, dates, and criteria against the actual agreement. Review documents, media, and sources where included.

  4. 4. Resolve assumptions and edit

    Correct unsupported claims and ambiguous criteria. Review change reasons and open questions with the responsible people. Make each criterion observable before treating the ticket as ready.

  5. 5. Apply and verify the saved ticket

    Apply the reviewed version explicitly, then inspect the saved result. If another edit has changed the ticket since the preview, refresh and regenerate a proposal before applying it. Share the final task with its owner and reviewer.

Worked example

The illustrative starting ticket says “Fix onboarding docs.” The author adds guidance to clarify the support handoff. The reviewer narrows the outcome to “Update the setup instructions and escalation contact,” verifies the referenced policy, and writes criteria for tested links and a reviewer-approved sequence.

Illustrative workflow · reviewed ticket improvement
  1. Original ticket

    An initial task with known context and constraints.

  2. AI proposal

    Suggested content, metadata, reasons, and open questions.

  3. Human review

    Check the facts, narrow scope, and edit the proposal.

  4. Explicit apply

    Save the reviewed ticket and verify its final content.

This is an editorial example showing the review process. It is not a captured provider response. A preview leaves the original saved ticket unchanged until the reviewer applies the changes.

Common mistakes

Accepting a plausible assumption

Check names, dates, source material, and requirements. A convincing proposal still needs a reviewer who knows the work.

Reviewing only the description

Inspect metadata, criteria, documents, and media as well as the prose. A changed assignee or date can alter the working agreement.

Leaving criteria untestable

Replace “make it better” with a result someone can observe, such as tested links and an approved setup sequence. Resolve remaining open questions before applying.

Headquarters by Kooya

Explore your company workspace.

Discover how projects, people, and cost visibility come together in Headquarters.