GENAI AREDEZ
GENAI AREDEZ · AI Governability · MMXXVI · AI-Operable · Page 3 / IX

AI-Operable

What it means, in concrete terms, for your business to be operable by artificial intelligence.

Subtitle

It is the most-used term and the least-explained. Here we ground it: without it, what follows is smoke.

§ 01 · The definition, without detours

Operable: one word, four requirements.

A business is AI-operable when its processes, data and permissions are described in a structured way —machine-readable— so that an artificial intelligence system can act on them with traceability.

Fig. 05 The four layers of operability
01 Layer · Processes

Described processes.

What steps exist, in what order, with what exceptions. Not "in an employee's head".

02 Layer · Data

Readable data.

What the datum is, where it comes from, what can be done with it, what quality it has.

03 Layer · Permissions

Defined permissions.

What the AI may touch and what it may not; what it decides autonomously and what goes to a person.

04 Layer · Traceability

Traceability.

Every action leaves a record: what it did, with what data, why. Auditable afterwards.

The consequence

The difference between a copilot that suggests and a system that operates with accountability.

Without those four, an AI "works" in a demo and fails in production — or it works, but no one can say what it did. With them, the AI operates and lets itself be seen.

§ 02 · What it is not

Not a folder of PDFs.

The vast majority of corporate documentation is made to be archived, not operated: PDFs, slide decks, manuals no one opens. A machine does not read that, and an agent does not use it.

What we deliver is the opposite: operable documentation —structured, interlinked, with stable identifiers— that an AI agent queries, traverses and acts upon, and that a person understands. And, on top of that foundation, the artificial intelligence that operates the process — we build it or we refine it.

§ 03 · The proof

The proof is in this very project.

This is not theory. This very project —its strategy, its positioning, its sources, its decisions— is built as documentation operable by agents: stable identifiers, a navigable map, orchestration rules readable by a machine.

An AI agent picks it up, traverses it and extends it without anyone summarising anything for it. We show you the pattern because it is exactly the one we leave in your company.

GENAI-AREDEZ/
  01-DESIGN.md   — design system, tokens and rules
  02-design-tokens.css   — variables: palette, typography, spacing
  SITE.md   — site architecture, page map, order
  pages/
    01-inicio.md   → /en/
    02-metodo.md   → /en/method
    03-operable.md   → /en/operable ◀  you are reading this
    

§ 04 · Next step

Operable, governed, demonstrable.

Once the foundation is operable, what follows is giving it a defensible shape —for you and for a regulator— and the method to get there. The two following pages explain it in detail.

Operable is not a word: it is the condition for everything else.

Let us talk → One conversation. No form, no demo.