6. AI workflows and experimentation
6.1 Capturing a reusable documentation process
I developed reusable AI workflows for help authoring, Jira research, peer review, feature-level content comparison, visual diffs, and Umbraco export. They connect requirements, existing help, designs, topic changes, review, and validation. Generalized Jira and peer-review skills are available for other writers to customize.
The AI Help workflow examines feature requests, comments, attachments, linked implementation issues, and relevant Figma designs before identifying the documentation impact. I used it to prepare targeted changes and new draft content for the 2027.0 help, with self-review and image checks built into the process.
I shared the workflow through a team email and training guide. The process helps writers reach a strong starting draft, then test, review, refine, and verify it. Repository synchronization and validation keep generated content compatible with the publishing system. The AI authoring project account explains this release-help workflow.
6.2 Learning from actual review decisions
I mined approximately 2,500 peer-review comments from Git commit history to shape this skill's review guidance. I refined it when an automated review missed something I subsequently corrected. The work included direct openings, grammar, title and UI-control treatment, and the distinction between adding review comments and improving a generated draft. The project account explains the reusable skill.
The skill makes recurring editorial decisions easier to explain, apply, and refine through human review.
6.3 Integrations and access boundaries
My Jira MCP business case requested read-only access for research into recurring support issues, troubleshooting patterns, product behaviour, and documentation opportunities. It distinguished that research access from create/edit/transition permissions and retained Technical Writer review before publication.
I also explored Jira Data Center and Excel integration, attachment handling, previews, and Copilot Studio workflows. I considered the access each task needed, distinguishing research from actions that create or modify content.
The integrations also include Figma design retrieval, API/export investigations, and packaging workflows for different AI environments. I prepared public Jira and peer-review copies with installation paths for both Codex and Claude.
6.4 Tool evaluation and governance
In the Education Services AI Committee, I compared Copilot, Claude, Open Web UI, and other tools for documentation work. I shared my experience using Claude for migration development and discussed which tools suited particular tasks, alongside quality, cost, and governance considerations.
The committee discussed measurable outcomes, use cases for different tools, licence costs, and consolidated business cases. We evaluated potential reductions in manual effort alongside quality and cost considerations.
I also explored embedding a custom AI worker through Open Web-based APIs, including its business case and pricing, and examined the case for a Figma Dev seat to address export/API limitations.
6.5 Pulse AI and visual context
A September 3, 2026 proof-of-concept discussion addressed better image handling in Pulse AI answers. I suggested focusing on a specific toggle or option in a workflow and asked whether UI/UX already had reusable labelled material. A colleague and I were assigned to help label and draw bounding boxes for a SOTI XSight Live View console page.
I also investigated Pulse AI responses where draft documentation appeared unexpectedly and the data-source or repository path needed review. This proof of concept examined what an AI system consumes and exposes alongside my work using AI to draft content and build tools.
6.6 Refining workflows through practice
I use AI across research, authoring, peer review, integrations, migration, and troubleshooting. Repeated use helps me identify where a workflow needs clearer instructions, additional checks, or a human decision.
I refine instructions and checks as I learn from new features, review feedback, and the behavior of the tools themselves.