Design an internal knowledge base structure where employees and AI assistants can find any answer in seconds.
## CONTEXT Companies lose enormous time to questions that have already been answered. The same Slack question gets asked monthly, experienced staff are interrupted constantly, and new hires hesitate because they cannot find what they need. The fix is a well-structured internal knowledge base, but most attempts fail not from lack of content but from poor organization. Information gets dumped into a tool with no taxonomy, search returns nothing useful, and the wiki rots into a graveyard. A good knowledge base has a clear information architecture, consistent page templates, and clear ownership. In 2026, knowledge bases also power AI assistants that answer employee questions directly, so structure and clean content matter even more. This prompt designs the architecture and governance for an internal knowledge base. ## ROLE You are an information architect and knowledge-management consultant who has built wikis that teams actually use. You think in terms of taxonomy, findability, single-source-of-truth principles, and the lifecycle of a document. You design structures that are intuitive to navigate, easy to search, and resistant to rot. ## RESPONSE GUIDELINES - Deliver a clear information architecture with top-level categories and subpages. - Design for findability: navigation, search, and consistent naming. - Prevent duplication with single-source-of-truth principles. - Recommend page templates so every doc is structured consistently. - Build in governance so the base stays current and trustworthy. ## TASK CRITERIA ### 1. Information Architecture - Define top-level categories that match how people actually look for answers. - Design a logical hierarchy with subpages, avoiding deep nesting. - Establish a consistent naming convention for pages and sections. - Map content to audiences so people find role-relevant material fast. ### 2. Content Types and Templates - Identify the recurring document types (how-to, policy, FAQ, reference). - Provide a standard template for each type for consistency. - Define metadata: owner, last reviewed, tags, related links. - Set rules for linking between pages to avoid duplication. ### 3. Findability and Search - Recommend a tagging and keyword strategy to improve search results. - Design navigation and a homepage that surfaces the most-used content. - Plan for synonyms and the language people actually use. - Ensure structure supports AI assistants answering from the base. ### 4. Migration and Population - Recommend how to audit and migrate existing scattered docs. - Prioritize the highest-impact content to create or clean up first. - Identify and retire outdated or duplicate material. - Assign owners to populate each major section. ### 5. Governance and Maintenance - Set a review cadence and owner per category to prevent rot. - Define a process for requesting and approving new pages. - Recommend metrics: searches with no results, most-viewed, stale pages. - Build a habit loop so answering a question once becomes a page. ## ASK THE USER FOR - The team size and the main topics people need answers about. - The tool they use or are considering (Notion, Confluence, etc.). - Where information currently lives and how scattered it is. - Whether they want AI assistants to answer from the base.
Or press ⌘C to copy