Custom solution · pre-build
Stop the same questions climbing upward
In many mid-sized companies, important operating knowledge lives in the heads of a small leadership team. Managers ask the same questions repeatedly, new leaders take months to learn how decisions are made, and the reasoning behind a decision can leave with the person who made it.
Org Brain is a proposed second stage after Personal AI setups are working for individual leaders. Each leader would choose a curated subset of decisions, standing positions, ownership and context to publish. Those approved subsets would form a shared company memory. Staff could ask it a question and receive an answer that names the leader who owns the position and the date it was last updated.
The asset is judgment, not a pile of files
The useful material is written leadership context: what was decided, who owns it, when it changed and why. Org Brain would not ingest every document and call that knowledge. Each leader chooses what to publish from their Personal AI setup, and the shared version carries only that approved material.
Visibility follows the audience
The proposed design uses three audience rings: leadership only, management and all staff. A user sees only material published for their ring. A private leader vault is never exposed as a whole, and unpublished links must not be displayed or completed by the system.
Answers inform; leaders decide
An answer would carry its owner and last-updated date so accountability remains visible. Org Brain would not make policy, approve credit, commit the company or replace the person who owns the decision. When the current source is missing, the right result is an unanswered question—not a confident guess.
A custom build in stages
What would need to be designed and tested
Org Brain is an offer concept, not a completed public product. The engagement starts only after Personal AI setups provide maintained private knowledge for the participating leaders.
1. Published leadership subsets
Define the decisions, standing positions, ownership and context each leader may publish. Add a named owner, update date and intended audience to every approved article.
2. Audience enforcement
Build and test server-side access for leadership, management and staff. The same question can return different permitted material depending on who asks it.
3. Sourced answer experience
Provide a company-specific question interface. Every answer must expose the accountable owner and date, while refusing to invent material from a private or missing source.
4. Unanswered-question queue
Record questions the approved knowledge cannot answer. A monthly view would group repeated gaps so each leader can choose what to write or update next.
Important boundary
Not file search, not full visibility, not an autonomous brain
Org Brain is not a search index over everyone's documents. It is not a connection that exposes every leader's Personal AI setup, and it is not a system that decides on behalf of the company. It is a proposed shared memory fed by carefully published subsets, with access controls and human accountability built into the design.
Hosting, identity, audience enforcement, retention and the first useful question set would be scoped for each organisation. There is no list price because the work depends on the number of leaders, the selected knowledge, the access model and the systems involved.
Start with the foundation
Build the Personal AI knowledge first
If your leadership team repeatedly answers the same operating questions, we can first map the private Personal AI setups and the small approved subset an Org Brain could use.
Questions people ask
Is Org Brain available as a finished product today?
No. It is a custom solution at the pre-build stage. It would be scoped and tested for one company after the participating leaders' Personal AI setups are in place.
Does it expose a leader's whole knowledge base?
No. Each leader would publish only a selected subset, with an audience set for each approved article.
How is Org Brain different from document search?
It centres on curated decisions, standing positions, ownership and context, rather than indexing every company file.