Wiki & Knowledge Base
Organizational memory that actually works. Making explicit what was implicit, and making hard-to-find information easy to find.
Part of The Substantial Brain
The Brain Wiki & Knowledge Base is organizational memory that actually works.
It makes explicit what was implicit. It makes hard-to-find documents and scattered bits of information easy to find. And it solves the problems that traditional company wikis have always struggled with: nobody uses them, they go stale immediately, they're impossible to navigate, and they end up full of conflicting information.
The traditional wiki is a graveyard. It requires massive manual effort to build, decays almost immediately, and people bypass it entirely in favor of DMing the nearest expert on Slack. That's the reality for most organizations.
An AI-native knowledge base shifts the paradigm from "Where is that written down?" to "What is the current answer, and what's the context behind it?"
Every mid-market company we've talked to has the same problem: critical knowledge lives in people's heads, in random Slack threads, buried in Google Drive folders, or scattered across tools that don't talk to each other. When someone leaves, that knowledge walks out the door. When someone new joins, onboarding is a scavenger hunt.
This is the foundational layer of the Brain. If we get organizational memory right, everything else we build (meeting intelligence, workflow automation, operational dashboards) gets smarter because it has context to draw from.
The simplest version that delivers real value on day one.
What we're building:
What this solves: Someone asks "What's our PTO policy?" or "How do we handle client SOWs?" and gets an accurate answer with a link to the source doc in seconds, instead of hunting through Drive or bugging someone on Slack.
Key technical direction:
Turn the knowledge base into something people can actually browse, not just query.
What we're building:
What this solves: People don't always know what to ask. Sometimes they need to browse, explore, and stumble into knowledge they didn't know existed. The chat sidebar per document means you can read a policy and immediately ask "Does this apply to contractors?" without switching contexts.
Key technical direction:
The knowledge base becomes a living, team-maintained system.
What we're building:
What this solves: This is where traditional wikis die. Maintaining docs is tedious, so nobody does it, and the wiki rots. Lowering the friction to update (especially via chat) means docs stay alive. Version history means people can update confidently without fear of breaking things.
Key technical direction:
These are features on the horizon. Each one extends the Brain's capabilities and could be built independently as the core levels mature.
A few guiding principles for how we build this:
Trust over magic. Every answer cites its source. If the system isn't confident, it says so. People need to trust this before they'll rely on it.
Meet people where they are. Slack-first for queries. Web for browsing. Voice for capture. Don't force people into a single interaction model.
Lower the friction of maintenance. The #1 killer of wikis is the effort required to keep them current. Every design decision should ask: does this make it easier or harder for someone to update a doc?
Start narrow, go deep. Better to have 20 well-indexed, well-maintained docs that give great answers than 200 docs dumped in with mediocre retrieval. Quality of the knowledge base matters more than quantity.
Build for the mid-market. Our users are companies with 50-500 people. They don't have dedicated knowledge management teams. This needs to be simple enough that an ops person or EA can manage it, not just engineers.