Universal Cognitive Architecture (UCA), explained

Your organization speaks hundreds of incompatible languages: marketing in Notion, sales in Salesforce, engineering in GitHub. AI can’t synthesize “competitive analysis” and “market intelligence” even when they name the same thing.

Libraries solved this a century ago with universal classification. A book has a coordinate, and anyone who knows the system can find it without searching the building. Universal Cognitive Architecture (UCA) does the same for business knowledge: it gives every piece of what an organization knows a permanent address, so memory is retrieved by address, not search. It is how business becomes speakable to AI.

In linguistics, cognates are words that share the same root across different languages: doctor, docteur. Machines and humans need the same thing for business knowledge. UCA establishes 11 Organizational Cognates, categories universal to every organization, so your team and your AI share one classification system.

Each cognate has a two-letter code, and each answers one question about the business:

• Identity (ID): what you stand for.
• Market (MA): where you compete.
• Innovation (IN): what you’re building.
• Brand (BR): how you’re perceived.
• Productivity (PD): where you do your best work.
• Workflow (WF): how work moves.
• Process (PR): how work generates profit.
• Culture (CU): how you sustain performance.
• System (SY): how you’re configured.
• Language (LA): how you measure.
• Integration (IT): how intelligence flows.

Every business speaks many languages at once: sales talks pipeline, finance talks margin, your apps speak their own APIs, and every AI model reads in its own way. UCA is the Rosetta Stone between them. MA20 means your competitive position to your CMO, your CFO, Claude, GPT and every system that touches it. Store it once at that address, and every reader, human or machine, finds the same thing, with nothing to search and nothing to translate.

An address is a cognate, a number and a tier. Two letters say which of the eleven domains, the number says which shelf, and nothing else is needed to place it. MA20 is competitive position. MA21 is the business model. MA50 is the customer master. A human reads the names; a machine resolves the addresses.

Addresses extend. A proposal for a prospective customer sits under the customer at MA50_01, and the prospect’s own context node beneath it at MA50_01_SY20. The address is the meaning: where a document lives says what it is and what it belongs to.

Two documents cannot occupy the same coordinate. Where two are discovered at one address, which happens whenever two people file independently, the collision is surfaced, not resolved automatically. A person decides which one governs. A silent merge of two documents that each claimed to be authoritative is precisely the failure governed memory exists to prevent.

The number also places a memory in a tier. Canon is persistent memory: the standards and protocols the business runs on, such as MA21, the business model. Projects are active memory: the work folders in motion, such as MA50, the customer master. Patterns are learning memory: the lessons the business keeps, such as IN80, the Org Brain as head code.

Together they are an Org Library: three authored tiers, held as lossless truth in a repository you own, with a person governing every commit.

An address is unique inside an Org Library, not inside the world. MA20 is your competitive position in your library and somebody else’s in theirs, and the two never meet, because they are never under the same root.

The standard is universal in its vocabulary and local in its scope. Every organization uses the same eleven cognates and the same coordinates, which is what makes a standard worth having, and every organization’s memories live in its own library, which is what keeps them from colliding. Shared memories are a use case of the standard, not an exception to it: a partner’s node is re-addressed under the host and keeps its own coordinate as the final segment.

UCA 1.0 is published under Creative Commons Attribution 4.0. Nothing is gated and nothing is held back for customers. The specification carries everything an implementation needs: the eleven cognates, the four memory tiers, the address format, extension patterns, the collision and versioning policy, the governance model, and the conformance checklist.

It is given away so any size of business can confidently turn off AI memory. The download is one file: the specification, the fifteen starter files and a README. Every version stays published at its own address, and nothing is silently amended.

Start with the First Fifteen: the fifteen short documents the standard names, from Context Node to Product Roadmap, that make a business speakable to AI. A mature company can draft them in about an hour by handing a capable model the fifteen names and its existing record. With all fifteen in place, session-based AI knows enough to extrapolate the rest.

Then give the library an AI00, the page a model is told to read first, which tells it how to read and write the library. The free standard supplies a default, and it already carries the eight terms of the AI Code of Conduct, so every model reads them before it reads anything else.

Read the specification, or let Starling MX run it for you: the standard is the first of its five moves.