Skip to content

Brando::Representing Brand-as-Code

Brand Oracle (Brando) Vocabulary v1.3

Brando® – the Brand Oracle — is a proposed linked-data standard and Brand Directive Vocabulary (BDV). It models brand identity, expression, governance, and product/persona structures as machine-actionable semantics, providing an ontological layer that LLMs, agents, and multimodal systems can interpret, enforce, and perform with precision.

Brando Schema: AI-native vocabulary for brands

Why Brando exists

Advanced Analytica has been pioneering the use of Schema.org and JSON-LD (Schema Markup) as a control layer for LLMs and generative AI. Our research into foundation model training and inference showed that JSON-LD is an AI-native, graph-based representation that models interpret reliably, making it a natural fit for encoding brand rules and constraints and enforcing them as a controlled, policy-aware brand oracle at runtime.

Brando, short for Brand Oracle, is a formal extension of Schema.org from a descriptive SEO layer into an AI-native operative layer for generative and agentic systems. Instead of replacing existing Schema Markup, Brando reuses and extends it – minimising rework, accelerating implementation, and simplifying internal adoption. If your sites already publish Schema Markup, you already have the substrate that Brando turns into Brand-as-Code.

This site documents the latest version of the Brando Vocabulary v1.3:

  • the JSON-LD vocabulary under the brando: namespace,
  • the Brand Directive Vocabulary (BDV) patterns used at runtime,
  • the YAML profile for configuration and ops,
  • and reference integration patterns for Brand Operating Systems (Brand OS), RAG pipelines, and agent runtimes.

Brando is designed to be both:

  • an open vocabulary and ontology, and
  • the schema layer for Machine-Actionable Policy Graphs that AI runtimes can execute in a Brand Operating System.

What is Brando?

Schema.org defines what things are — the entities in your world: products, brands, audiences, articles, places, actions.
It is primarily a descriptive vocabulary: it helps machines recognise and index entities.

GS1 Web Vocabulary extends this descriptive layer for commerce: trade items, locations, parties, logistics, and supply chain attributes.

Brando sits one layer up as a Brand Directive Vocabulary (BDV):

  • where Schema.org and GS1 describe facts about things,
  • Brando BDV encodes how the brand expects systems to behave around those things.

In other words:

  • Schema.org / GS1 → descriptive layer (“what this product/brand is”).
  • Brando BDV → directive layer (“how any AI or agent must speak and act about it”).

Brando is:

  • a linked data vocabulary and ontology for brands (JSON-LD, Schema.org extension),
  • a Brand Directive Vocabulary (BDV) that encodes tone, language, constraints, guard rails, personas, and automation rules, and
  • a taxonomy / control layer for brand and product categories aligned with GS1, UNSPSC, and Google Product Taxonomy.

Brando lives at:

  • Namespace: https://brandoschema.com/
  • Preferred prefix: brando

It:

  • extends schema:Brand and schema:Intangible,
  • aligns cleanly with GS1 Web Vocabulary for products, brands, GTINs and GLNs,
  • is published canonically in JSON-LD, and
  • is profiled for YAML configuration and TypeScript types.

You can:

  • embed Brando JSON-LD in public sites for AI and web discovery,
  • manage Brando YAML in repositories as configuration for a Brand Operating System (Brand OS), or
  • load Brando instances into a Brand Policy Graph that powers multiple runtimes with the right runtime architecture, execute Brando instances as Machine-Actionable Policy Graphs that direct how AI systems behave.

Internally, we talk about Brando as:

Brando BDV – the Brand Directive Vocabulary that tells AI systems what a product or brand is, and how they are allowed to talk, reason, and act in relation to it.

What this documentation covers

Use the navigation, or jump directly into:

Brando in practice: from vocabulary to Policy Graph

The Brando vocabulary is designed to be used as the schema layer for:

  • Brand Knowledge Graphs – modelling brands, contexts, tokens, personas, and policies as linked data;
  • Machine-Actionable Policy Graphs – encoding governance logic (what AI may / must / must not do) as data, not prose;
  • Brand Operating Systems – AI-native runtimes where LLMs, agents, and workflows query the Policy Graph and BDV at inference time.

Advanced Analytica’s Brando Implementation Service uses this vocabulary to:

  • map brand strategy and governance into Brando classes,
  • author JSON-LD / YAML Policy Graphs and BDV instances,
  • and deploy them as a live AI-native Brand OS.

For implementation details, see:

Brando as a brand vocabulary for AI

Traditional brand governance is built around static, human-only artefacts:

  • PDF guidelines and slide decks,
  • brand portals and DAM folders,
  • unstructured “system prompts” and playbooks for AI.

These do not translate cleanly into probabilistic systems like LLMs and agents.

The Brando vocabulary applies linked data and semantic web patterns to brand governance:

  • reusing JSON-LD and Schema.org patterns, and aligning where useful with GS1 product and brand descriptions,
  • modelling brand intent, rules, personas, and tokens as data (not prose),
  • and using that data as a control layer that AI runtimes can load, combine, and enforce.

Foundation models are heavily trained on JSON-LD and Schema.org–style structures.
Brando reuses this structure so that models can interpret brando: classes and properties zero-shot, without custom training.

Why AI has to help manage the Brand Policy Graph

A Brando-based Brand Policy Graph is intended to be living:

  • brands, products, and portfolios change,
  • channels and surfaces proliferate,
  • policies evolve with regulation and risk.

At any meaningful scale – multiple brands, markets, campaigns, and channels – maintaining this graph purely by hand becomes prohibitively hard:

  • too many contexts (brando:Context) to keep in sync,
  • too many tokens (Brando:VerbalIdentity, Brando:VisualIdentity, Brando:AudioIdentity) to update,
  • too many personas (brando:SyntheticPersona, brando:ProductPersona) to govern,
  • too many policies (brando:Policy, brando:AutomationRule) to refactor as requirements change.

Brando is therefore AI-manageable by design:

  • the vocabulary is compact, with clear domains and ranges, so agents can propose changes (e.g. new contexts, updated tokens, refined personas) in JSON-LD or YAML;
  • LLMs can generate schema-conformant Brando data zero-shot because the structure mirrors Schema.org and linked data patterns;
  • properties like brando:reviewWorkflow, brando:enforcementLevel, and brando:riskTag support human-in-the-loop governance, not bypass it.

In practice:

People define intent and governance.
AI does the mechanical work of keeping the Brand Policy Graph up to date, under auditable human control.

Advanced Analytica designs and iterates Brando with this AI-manageable governance model in mind, treating the vocabulary as a long-term, evolving asset for brands.

Core concepts and classes

Brando v1.3 defines twelve core classes under the brando: namespace:

Class Informal purpose
brando:Brand Core brand identity node extended with governance, architecture, and AI-ready semantics.
brando:Context Activation context describing audience, channel, jurisdiction, and operational conditions.
brando:BrandExpression Abstract superclass for reusable expression tokens across verbal, visual, and audio modalities.
Brando:VerbalIdentity Reusable configuration of the brand’s voice, language, and behavioural prompt rules.
Brando:VisualIdentity Reusable configuration of visual system: logo, colour, typography, imagery, and motion.
Brando:AudioIdentity Reusable configuration of sonic identity: voice, sonic logos, and audio cues.
brando:Policy Governance node for guard rails, refusals, compliance tags, and retrieval visibility.
brando:BrandedCategory Brand-specific product/service category aligned with GS1, UNSPSC, and Google Product Taxonomy.
brando:Campaign Time-bound campaign that may temporarily adjust brand rules.
brando:AutomationRule Automated governance rule reacting to metrics or events to adjust brand behaviour.
brando:ProductPersona Product-scoped AI persona aligned with schema:Product / schema:Brand / schema:Organisation and GS1 identifiers for that product.
brando:SyntheticPersona Governed AI persona type encoding identity, tone, competencies, constraints, and policy bindings aligned with schema:Product / schema:Brand / schema:Organisation .

All Brando classes ultimately extend schema:Intangible, and brando:Brand additionally extends schema:Brand and can be related to GS1 brand/product concepts in implementation profiles via schema:Product.

High-level relationships

Key structural properties include:

  • brando:hasContext
    – brando:Brand → brando:Context (activation contexts in which the brand operates)

  • brando:usesVerbalIdentity / brando:usesVisualIdentity / brando:usesAudioIdentity
    – brando:Brand or brando:Context → expression tokens for different modalities

  • brando:hasPolicy / brando:hasAutomationRule
    – attach governance and automation to brands and contexts

  • brando:hasProductCategory
    – brando:Brand → brando:BrandedCategory nodes with GS1/UNSPSC/Google mappings

  • brando:hasSyntheticPersona / brando:hasProductPersona
    – connect brands and products to their governed AI personas

See Types reference and Properties reference for normative definitions (domains, ranges, comments).

Canonical form and implementation profiles

Brando is JSON-LD–first:

  • the JSON-LD vocabulary defines all classes, properties, and metadata (publisher, license, version info);
  • implementations are encouraged to publish JSON-LD on the public web where possible.

Two implementation profiles are documented:

  1. JSON-LD profile

For:

  • public publication on brand domains,
  • ingest into knowledge-graph platforms,
  • direct consumption by AI runtimes.

See:

Same semantics, represented as YAML, suitable for:

  • configuration repositories,
  • CI/CD pipelines,
  • infrastructure-as-brand setups.

See:

A TypeScript model is provided as a non-normative convenience layer for strongly-typed implementations
(see TypeScript types).

Example: minimal brand + context in JSON-LD

The example below shows a small Brando graph using canonical JSON-LD:

{
  "@context": {
    "schema": "https://schema.org/",
    "brando": "https://brandoschema.com/"
  },
  "@id": "https://example.com/brand/northstar",
  "@type": "brando:Brand",
  "schema:name": "Northstar Bank",
  "brando:missionStatement": "Financial clarity with absolute trust.",
  "brando:coreValues": [
    "Clarity over complexity",
    "Do the right thing",
    "Long-term alignment with customers"
  ],
  "brando:hasContext": {
    "@id": "https://example.com/brand/northstar/context/retail-chatbot",
    "@type": "brando:Context",
    "brando:audienceSegment": "Retail banking customers",
    "brando:domainContext": "{\"channel\":\"chatbot\",\"jurisdiction\":\"UK\",\"productLine\":\"current-accounts\"}"
  },
  "brando:usesVerbalIdentity": {
    "@id": "https://example.com/brand/northstar/tokens/verbal/default",
    "@type": "Brando:VerbalIdentity",
    "brando:toneOfVoice": "calm, empathetic, and precise",
    "brando:approvedTerms": [
      "interest rate",
      "fees",
      "savings goals"
    ],
    "brando:prohibitedTerms": [
      "guaranteed returns",
      "get rich quick"
    ]
  },
  "brando:hasPolicy": {
    "@id": "https://example.com/brand/northstar/policies/regulated-claims",
    "@type": "brando:Policy",
    "brando:enforcementLevel": "mandatory",
    "brando:riskTag": [
      "financial-regulation",
      "consumer-protection"
    ],
    "brando:refusalStrategies": [
      "Decline to provide personalised investment advice.",
      "Redirect to a human advisor for complex suitability questions."
    ]
  }
}

The same structure can be authored as YAML using the Brando YAML profile, then emitted as JSON-LD at build or deployment time.

The example below shows a small Brando graph using YAML:

@context:
  schema: https://schema.org/
  brando: https://brandoschema.com/

@id: https://example.com/brand/northstar
@type: brando:Brand

schema:name: Northstar Bank
brando:missionStatement: Financial clarity with absolute trust.
brando:coreValues:
  - Clarity over complexity
  - Do the right thing
  - Long-term alignment with customers

brando:hasContext:
  @id: https://example.com/brand/northstar/context/retail-chatbot
  @type: brando:Context
  brando:audienceSegment: Retail banking customers
  brando:domainContext: '{"channel":"chatbot","jurisdiction":"UK","productLine":"current-accounts"}'

brando:usesVerbalIdentity:
  @id: https://example.com/brand/northstar/tokens/verbal/default
  @type: Brando:VerbalIdentity
  brando:toneOfVoice: calm, empathetic, and precise
  brando:approvedTerms:
    - interest rate
    - fees
    - savings goals
  brando:prohibitedTerms:
    - guaranteed returns
    - get rich quick

brando:hasPolicy:
  @id: https://example.com/brand/northstar/policies/regulated-claims
  @type: brando:Policy
  brando:enforcementLevel: mandatory
  brando:riskTag:
    - financial-regulation
    - consumer-protection
  brando:refusalStrategies:
    - Decline to provide personalised investment advice.
    - Redirect to a human advisor for complex suitability questions.

Brando in a Brand Operating System

Brando is designed to act as the data and ontology layer for a Brand Operating System:

Identity & architecture

Model house-of-brands, branded houses, and hybrids with multiple brando:Brand nodes and properties such as:

  • brando:brandArchitectureRole
  • brando:inheritsPolicies
  • brando:inheritsTokens

Context-aware behaviour

Use brando:Context to represent channels, audiences, regulatory regimes, and operational conditions that change how the brand behaves.

Multimodal expression tokens

Attach Brando:VerbalIdentity, Brando:VisualIdentity, and Brando:AudioIdentity to brands and contexts using:

  • brando:usesVerbalIdentity
  • brando:usesVisualIdentity
  • brando:usesAudioIdentity

Personas and product-linked assistants

Use:

  • brando:SyntheticPersona for governed, cross-surface AI personas, and
  • brando:ProductPersona for GTIN/GLN-scoped assistants that speak for specific products or ranges,

linking them back to brands, products, contexts, and policies via brando:hasSyntheticPersona, brando:hasProductPersona, and brando:appliesToProduct.

Governance and lifecycle

Encode rules and risk in brando:Policy and brando:AutomationRule with:

  • brando:enforcementLevel (e.g. "mandatory", "advisory", "conditional"),
  • brando:riskTag,
  • brando:reviewWorkflow,
  • plus temporal metadata (dcterms:created, dcterms:modified, optionally schema:validFrom / schema:validThrough in instance data).

Runtime automation and integrations

Use:

  • brando:AutomationRule, brando:triggerType, brando:automationAction, brando:schedulingRule for automated checks and behaviours;
  • brando:externalSystem and brando:externalSystemId to connect Brando nodes to external applications such as brand management platforms, content and design tools, ad platforms, CMS/DXP systems, and internal line-of-business tools;
  • brando:downloadableFile and brando:supportingFile to point AI systems at the right assets in a DAM or Brand OS.

For architecture and runtime patterns, see:

Classification alignment

brando:BrandedCategory nodes provide a bridge between a brand’s own taxonomy and external standards.

Relevant properties include:

  • brando:gpcCategoryCode / brando:gpcCategoryDescription – GS1 Global Product Classification (GPC),

  • brando:unspscCode – UNSPSC category identifiers,

  • brando:googleProductCategoryId – Google Product Taxonomy identifiers.

This lets you:

  • keep brand-facing labels and semantics in your own language,
  • while exposing machine-readable hooks to GS1 / UNSPSC / Google for commerce, ads, and marketplaces.

See also Brand portfolios & architecture for portfolio- and category-level modelling patterns.

Versioning and licensing

From the vocabulary header (owl:Ontology):

  • Title: Brando Schema Vocabulary v1.3
  • Namespace (prefix): brando: → https://brandoschema.com/
  • Ontology version info: owl:versionInfo = "1.4"
  • Created: dcterms:created = "2025-11-22"
  • Modified: dcterms:modified = "2025-11-22"
  • Publisher: Advanced Analytica
  • License: CC BY 4.0 via cc:license

Implementations should treat:

  • the JSON-LD vocabulary file and
  • this documentation set

as the normative reference for Brando Schema Vocabulary v1.3.

Any “future ideas” or experimental patterns mentioned in the docs are explicitly labelled as:

  • non-normative guidance, or
  • potential future extensions, or
  • profile-level recommendations,

and are not part of the core v1.3 vocabulary.

Stewardship and support

Brando is originated, owned, and stewarded by Advanced Analytica.

Advanced Analytica:

  • defines and maintains the core Brando vocabulary,
  • publishes the reference JSON-LD and YAML profiles,
  • and leads ongoing research, design, and innovation around Brand OS patterns and AI-native brand governance.

The intent is to:

  • keep the core vocabulary stable and minimal,
  • allow organisations and platforms to define extensions for their own needs,
  • and evolve the spec in collaboration with brands, agencies, and technology providers.

For information about stewardship, trademark usage, and terms, see:

From vocabulary to running Brand OS

Brando is designed to be used, not just read.

Most teams follow a simple spec → graph → runtime path:

  1. Model with the Brando vocabulary

    • Define brando:Brand, brando:Context, expression tokens, personas, and policies.
    • Start with JSON-LD or YAML using the official profiles.

  2. Build a Machine-Actionable Policy Graph

    • Load Brando instances into a Brand Knowledge Graph or config repo.
    • Attach policies, campaigns, personas, and categories to the right brands and contexts.

  3. Integrate at runtime

    • Expose the Policy Graph and BDV to LLMs, agents, and APIs.
    • Use it as the control layer that governs prompts, retrieval, generation, and approvals.

  4. Monitor and evolve

    • Treat the graph as a living asset: update products, contexts, personas, and rules as the brand changes.
    • Use AI to propose structured changes, with humans approving and merging.

For many organisations this journey is incremental: start with a single brand and use case, then extend to portfolios, campaigns, automation rules, and product-level personas.

Implementation support (optional)

Brando is an open vocabulary: you can adopt it independently using this documentation.

If you want a fully managed path from brand book to Brand OS, Advanced Analytica offers the Brando Implementation Service, covering:

  • mapping strategy and governance into Brando classes and BDV patterns,
  • authoring JSON-LD / YAML Machine-Actionable Policy Graphs and personas,
  • deploying them as a live Brand Operating System for your AI stack.

Learn more at:

Next steps

If you are evaluating or implementing Brando:

Step 1. Read the overview

Step 2. Clone a minimal example

Step 3. Choose your primary representation

Step 4. Model your first brand, context, and persona

  • brando:Brand
  • brando:Context
  • Brando:VerbalIdentity
  • brando:SyntheticPersona
  • brando:Policy

Step 5. Plan your Brand OS integration

From there, extend into portfolios, campaigns, product personas, automation rules, and classification alignment – all grounded in the same Brando vocabulary and managed as a living, AI-assisted Brand Policy Graph.