Documentation media enrichment · implementation

Verify an AI-proposed asset before enriching the procedure

In this proposed workflow, an AI media assistant interprets the cartridge orientation step and proposes an existing library asset with its source ID and revision evidence. Metadata filters limit eligible assets; service engineering verifies revision B geometry. Writers add labels and equivalent text, reviewers approve the asset, and technicians attempt the enriched task.

Read the business case

Business goal: Reduce costs
Business metric: Self-service rate

Prerequisites and scope

AI asset retrieval requires an approved library with source IDs, rights, approval state, and cartridge/housing revisions, plus a maintained link to the orientation step. Name service engineering as the reviewer of the proposed view.

The working sequence

The AI media assistant proposes an existing source asset from the orientation step and revision context. Filters restrict eligibility, engineering verifies the physical view, and people produce and approve the accessible annotation.

Select the needed view

Content team: The content team asks the AI media assistant for an existing revision B view showing the inlet, sealing edge, and latch. The proposal includes a library ID, source reference, revision, and approval state for inspection. It places the visual next to the orientation step and keeps the equivalent instruction in text.

Depends on: Service engineering confirms that the chosen cartridge revision and camera angle show the correct assembly. The AI assistant ranks candidate assets by their relevance to this procedure step; engineering certifies the chosen view’s physical accuracy before annotation.

Produce accessible media

Content team: Writers add focused labels, alt text conveying orientation, and captions or a transcript if motion is shown. They avoid decorative screenshots that repeat the paragraph without clarifying the action.

Depends on: Accessibility reviewers test meaning and contrast; a technician confirms that the labeled view supports the task. Automatic checks can detect missing text tracks, not whether a description is sufficient.

Publish with version links

Content team: The asset record carries part revision, product version, source, owner, approval date, and the procedure step that uses it. The publishing workflow places the approved variant in every relevant channel.

Depends on: The CMS must keep links intact and invalidate outdated cached media. Stable IDs and taxonomy tags for part revision help retrieval; a graph is useful only if shared parts create complex cross-document dependencies.

Validation and exceptions

Include an obsolete but similar asset in the AI retrieval test. Confirm that reviewers can reject it using revision evidence, then have a technician identify the sealing edge and install the revision B cartridge using the approved visual and procedure.

If no approved image exists, do not fabricate a product view. Use a verified diagram or improve the text while sourcing an asset. If the visual and the physical part disagree, withdraw the visual and route the issue to engineering before reuse.

Keep it current

Review assets whenever the housing or cartridge changes. Check alt text, captions, and translations with each revision. Archive superseded media without breaking guidance for older installed equipment.

Evidence of progress

Save AI asset IDs, proposed reasons, model, prompt, and library metadata snapshot. Engineering scores revision fit and landmark visibility, including obsolete lookalikes. Run matched text-only/enriched task sessions with approved imagery; count completed installations, errors, and help requests rather than asset views.

Audit assets for Media accessibility pass rate and compare current-version approval records for Media freshness coverage. Test matched task sessions for Task completion rate, then assess task-specific Self-service rate across the same support entry points.

These are proposed measures for the illustrative workflow. A business result needs observed, comparable operating data and a review of other causes.

Return to the business impact

How we work
with your team

We can guide your team, work alongside it, or deliver the work for you. Combine these approaches as your needs and capacity change.

Lead

Build capability

We guide your people through the work, building capability that stays with your team.

Support

Add specialist backup

Your team owns the work. We provide experienced backup and specialist help where you need it.

Do

Delegate delivery

We handle work that is one-time, specialized, or outside your core business, without adding an ongoing burden to your team.

Build your team's skills with standard and custom courses.

More about our approach

Support for this implementation

Intuitive Stack can help connect AI asset proposals to procedure steps, part revisions, and approval evidence, then test the verified visual with technicians.

Plan a controlled pilot