Traditional brand guidelines were written for people to read. They give designers, art directors, agencies, and marketing teams direction on how the brand should look and feel.

They weren’t designed to govern AI. An AI tool needs more than broad descriptions such as “confident” or “approachable.” It needs clear direction on approved claims, messaging priorities, language to avoid, and decisions that can’t vary from one output to another.

Thomas calls this a brand constitution. It is a central source of truth that sets out the governing logic of the brand: what it stands for, the rules it follows, and the choices it makes when priorities compete.

The constitution should be specific enough to guide every type of work, from an AI-generated customer email to a complex proposal. It gives teams room to tailor content while setting boundaries around the parts of the brand that must stay consistent. 

“Brand guidelines are built for human interpretation. A constitution has to be machine-readable. It needs much more detail and much more specificity about what we do, what the trade-offs are, and when they apply.” 

Thomas Marzano,

Agentic Branding Architect, Founder, Marzano Consulting 

A brand constitution is the starting point. To apply it in real work, AI also needs access to approved knowledge and a secure way to retrieve that knowledge within a workflow.

Thomas’s framework has three layers:

1. Brand constitution
   The governing law. It defines what the brand is, what it stands for, and the rules that guide its behavior.
2. Knowledge graph
   A searchable layer of brand knowledge. It includes approved guidelines, templates, assets, messaging frameworks, product information, customer stories, and proof points. Thomas also calls this the brand brain.
3. API and MCP
   The connection layer. An API, or application programming interface, allows systems to share information. Model Context Protocol, or MCP, gives AI tools a standard way to access approved systems, data, and actions.

“The constitution is the governing law. The knowledge graph is everything the brand knows. Then, through API and MCP, you expose that information to the workflows and agents.” 

Thomas Marzano,

Agentic Branding Architect, Founder, Marzano Consulting 

Most enterprises already have the content and information AI needs. The problem is that it often sits in different systems, belongs to different teams, and is difficult to find when someone needs it.

Brand assets might be in a digital asset management platform. Sales content may sit in SharePoint. Product details, customer stories, and proof points may be buried in old presentations or individual folders. That makes it difficult for employees to find approved material, and impossible for an AI agent to retrieve it reliably.

The knowledge graph solves that problem by making company knowledge structured and retrievable. In Thomas’s framework, it is the brand brain around the constitution. It holds the information people and AI agents need to represent the organization correctly.

Importantly, this isn’t a one-time content migration. The knowledge graph needs clear ownership so outdated messaging can be removed, new proof points can be added, and the material available to AI remains trustworthy.

“If a personal productivity AI isn’t connected to a knowledge base, and that knowledge isn’t structured correctly, your AI agent isn’t going to be able to give you many valuable answers.”

Thomas Marzano,

Agentic Branding Architect, Founder, Marzano Consulting 

Brand rules only protect the organization when people can use them without changing their flow of work.

When employees have to hunt for the right template, copy approved messaging into a prompt, or manually check requirements before sending a document, quality depends on time, memory, and individual effort. Those are weak controls at enterprise scale.

API and MCP connections make the brand constitution and knowledge graph available inside everyday tools. This gives AI the context it needs before it generates content, rather than relying on employees to provide every instruction themselves. Templafy’s MCP connects AI platforms to a governed document layer that applies approved content, templates, and compliance rules.

This allows central teams to manage the rules and information that must remain consistent, while local teams can adapt approved material for their customers, regions, and use cases. 

“With an agentic brand system, we’re putting brand governance upstream into the system. Therefore, downstream, in a local or regional workflow, the governance is already embedded and taken care of.”

Thomas Marzano,

Agentic Branding Architect, Founder, Marzano Consulting 

Most organizations have already given employees access to AI tools. People use them to summarize meetings, draft emails, or improve a first draft. That can save time, but it doesn’t change the wider process around the work. 

Organizations need to move beyond personal productivity use cases. Start with a high-value job that takes significant time and involves several steps, systems, and people. Then redesign that job from end to end with AI, using the brand constitution and knowledge graph as part of the workflow.

An agentic workflow is the result. It is a workflow where an AI agent can use approved knowledge, follow business rules, and complete several connected steps toward a defined outcome. 

For example, creating a sales proposal usually involves finding customer information, selecting the right template, locating approved messaging and product details, adding required legal language, and preparing the document for review. An agentic workflow connects those steps so the seller can start with the outcome they need, while the system brings together the right inputs.

This is how AI moves from a helpful writing tool to an enterprise capability: it reduces the manual work around a business process while keeping the output governed.

“The distance between intent and fulfillment has vanished. Now, you express your intent to AI, and it’s executed instantly.”

Thomas Marzano,

Agentic Branding Architect, Founder, Marzano Consulting 

Once the brand constitution, knowledge graph, and workflows are connected, the organization needs a way to improve the system over time.

Thomas’s agentic operating system describes that loop. It takes insight from the market, turns it into decisions, supports execution, and feeds the results back into the knowledge graph. 

1. Agentic insight
   Gather signals from performance data, customer feedback, social monitoring, competitors, market activity, and generative engine optimization, or GEO.
2. Agentic strategy
   Turn those signals into priorities, actions, and resourcing decisions.
3. Execution
   Create and deliver the documents, messaging, visual assets, campaigns, or experiences required to act.
4. Learning
   Feed results back into the knowledge graph so future work starts with better information.

This matters because a brand system can’t remain static while the market, customers, and business change. The knowledge and guidance behind AI need to improve as the organization learns what works. 

“The brands of the future will be the ones able to deliver their meaning in context, at infinite scale, in the moments that matter.”

Thomas Marzano,

Agentic Branding Architect, Founder, Marzano Consulting 

AI can draft, retrieve, assemble, and format. People remain responsible for deciding what information AI can use, which rules apply, and whether the final output is safe to send.

Thomas expects creative and brand teams to shift toward curation. They’ll help define the constitution, manage the knowledge graph, review outputs, and improve the workflows that use them. IT and business leaders will be needed to connect these systems securely and make sure they work across the organization.

This also requires enterprises to control AI use with intention. Tokens are units of text that AI models read and generate, and higher usage generally means higher cost. Rather than allowing every employee to build separate automations, organizations should focus on shared workflows that solve significant business problems and create measurable value. 

“Token usage goes completely out of hand when you enable your entire team to build their own workflows. But if you focus the team on workflows you’re going to tackle as a team, then you can constrain it and control it.”

Thomas Marzano,

Agentic Branding Architect, Founder, Marzano Consulting 

Documents are one of the first places enterprises need to govern AI at scale.

Employees are already using AI to create proposals, presentations, reports, and customer communications. These documents are high-stakes: they need to reflect the brand, use current and approved information, include the right customer or commercial details, and meet any legal or compliance requirements.

That makes documents a clear test of the framework Thomas describes. A brand constitution defines the rules AI must follow. A knowledge graph provides the approved content and context. The final step is making both available in the workflow where employees are actually creating. 

Templafy brings those elements together in document workflows. Its Document Agents connect AI with company-approved templates, content, brand assets, business data, and rules, helping employees create editable, business-ready documents without having to rebuild brand context in every prompt. 

“It’s no longer about delegated control. It’s about having a central system that, with an MCP, connects to your workflows. As soon as you’ve connected the brand API to the workflow, the brand governance is taken care of in that system.”

Thomas Marzano,

Agentic Branding Architect, Founder, Marzano Consulting