In this proposed workflow, an AI terminology model receives source passages and the approved Workspace concept entry, then proposes UI wording patches with source references. The entry retains Project for historical scope and project_id for the API contract. Literal-token checks and writer/API-owner review verify the proposed changes before the source and localization instructions are updated.
Read the business caseSupply the AI terminology model with current UI strings, literal API contracts, versioned source passages, and the approved concept entry. Product and localization owners define the release boundary and any unresolved meanings before patch approval.
Use the approved concept entry to constrain AI wording proposals. Discovery collects passages, owners approve meanings, and the AI model proposes a contextual patch for writer review and deterministic literal-token validation.
Content team: The content team collects Project and Workspace passages, including UI instructions, API examples, and historical guidance, as inputs for the AI wording review. Extraction finds candidates rather than declaring them equivalent.
Depends on: Product and API owners confirm whether a string names the same concept or a distinct one. Sensitive or customer-specific vocabulary requires permission-aware handling.
Content team: A glossary entry assigns a stable concept ID, definition, preferred UI label, deprecated UI label, literal API exception, scope, and effective release. The content team records examples of correct and incorrect usage.
Depends on: Product terminology owners resolve contested meanings; localization specialists approve target-language terms. The taxonomy uses the concept ID to group related labels. A knowledge graph is useful if many concepts have explicit renamed-from or part-of relationships across products.
Content team: The AI model proposes “Open your Workspace” for the current UI passage and leaves {"project_id":"…"} unchanged, citing the glossary’s literal API exception. A deterministic diff check flags any changed API token before writer/API-owner review. Writers review each change and update source topics, support macros, and localization instructions from the same approved entry.
Depends on: The CMS, string repository, and translation platform must accept the term record or an export. Failed syncs are logged; a global find-and-replace is not an approval path.
Inspect the AI diff for the intended UI change and unchanged project_id, then validate the API example against its contract. Retain historical wording where applicable and route ambiguous proposals to the term owner.
Keep historical quotations and older-version procedures labeled as such. If an API name conflicts with a UI label, document the exception in the term entry. Escalate terms with competing product meanings instead of letting a model choose by frequency.
Re-scan after UI and API releases; retire deprecated labels only when dependent channels have changed. Review localization feedback and maintain owner, date, and decision history for each concept.
Keep model, prompt, passage set, and effective release fixed while comparing glossary inputs. Save AI diffs and reviewer decisions, including protected API tokens and ambiguous contexts. Validate published examples and compare equivalent topic/language effort; a high scan score can still hide a wrong concept mapping.
Sample contextual uses for Approved term adherence, classify review returns for Terminology rework rate, and track issue-to-decision dates for Ambiguity resolution time. Compare Content cost across similar topic and language batches after accounting for complexity and volume.
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 impactWe can guide your team, work alongside it, or deliver the work for you. Combine these approaches as your needs and capacity change.
Build capability
We guide your people through the work, building capability that stays with your team.
Add specialist backup
Your team owns the work. We provide experienced backup and specialist help where you need it.
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.
Intuitive Stack can help give AI wording proposals a scoped glossary entry and literal API exceptions, with reviewable patches and checked published examples.
Intuitive Stack can help give AI wording proposals a scoped glossary entry and literal API exceptions, with reviewable patches and checked published examples.