A digital knowledge base for your organisation (+ new graph, namespaces, markdown headings & emojis)

A digital knowledge base for an organization is less about storing documents and more about building a shared “mental map” of how systems, processes, and responsibilities connect—especially when key knowledge currently lives in individuals’ heads. The core promise is that teams can navigate internal information quickly, understand dependencies between services and SOPs, and reduce the churn cost when people leave. In practice, that means moving beyond a simple wiki into a structured, link-rich system that reflects the organization as a network.

The walkthrough centers on Logseq as an experimental foundation for this idea, using recent graph improvements plus several structuring techniques. Logseq’s updated graph view loads faster and presents a clearer network of pages (nodes) connected by links (edges). Clicking around the graph makes it easy to see how topics relate—useful for building intuition about where information lives and how it flows.

To organize content at scale, the experiment adds namespaces using forward slashes (e.g., “role/project manager”), creating a hierarchy of roles that can be browsed and linked. Markdown headings provide another layer of structure: heading levels are controlled with hashtags, and the notes emphasize avoiding formatting mistakes like missing spaces that can accidentally create unintended linked pages. For navigation and visual scanning, page names can include emojis, inserted via keyboard shortcuts (Windows key + period; Mac key + period). Emojis are used to make areas like “services” or “areas” feel more like a browsable map rather than a flat list.

The most concrete payoff comes from dependency tracking. A queried “landing page” for a team’s SOPs links into individual SOP entries that include metadata such as the services used and the areas affected. Within an SOP, roles and responsibilities are linked so that operational ownership can be traced: if a person leaves, the role link can be updated rather than hunting through scattered documents. The system also hints at an entity-relationship style approach through a “portal” concept, where services and their relationships can be mapped and then referenced as a repository for others.

Still, the setup comes with trade-offs. The biggest challenges are onboarding friction—new users may find the network-and-properties model unintuitive—and the need for careful structuring to avoid confusion about when to use properties versus namespaces. Privacy is also limited in the current approach, since pages are effectively public within the shared environment. There’s also concern about corruption risk tied to shared-drive workflows.

Finally, the creator positions Logseq as a near-term personal or small-team wiki that could become a true collaborative knowledge base as Logseq’s roadmap matures toward multi-user, database-style collaboration. In the meantime, alternatives are considered for non-self-hosted needs: Notion (powerful but can become complex), TiddlyWiki (open source), Coda (fixed pricing tied to creators), Confluence (Atlassian, subscription-based, includes a graph), and Slab (graph-like interface). The takeaway is that every company should build a knowledge base now, but the best tool—and the best collaboration model—may still be evolving.