Why Your Brand Needs a Public AI Agent Profile

Product note (October 2026): AdMesh Public Profiles are in development. This article describes the intended experience; public brand pages and Agent endpoints are not yet available on the production site.

A customer sees your brand mentioned, opens a product page, and still has a question: Will this work for my situation? Another customer asks an AI assistant to compare options in your category. In both cases, the answer depends on more than a tagline. It needs current product information, clear limits, and a reliable place to go next.

Most brands already have that information, but it is spread across product pages, help articles, campaign copy, and internal documents. A public AI Brand Agent profile brings the information the brand chooses to share into one reviewed experience for people and for the agents they use.

What is a Public Profile?

An AdMesh Brand Agent holds the brand's broader context: its story, products, offers, audience, and guidance. The Public Profile is the part the brand deliberately prepares and publishes for others to use. Saving a draft does not change the live profile; the brand reviews both views and publishes when ready.

One publication has two presentations:

• For people: a visual brand page with a summary, selected products, current eligible offers, a customer action, and a Brand Assistant that can answer questions from the published context.
• For AI agents: a structured Agent view and a Markdown version with the same published brand information, product links, and the complete list of approved sources. When enabled, a separate query API lets a connected agent ask a question or retrieve structured information.

The Human page is designed to help a visitor decide what to do. The Agent view makes the brand's public information easier for a machine to inspect and cite. They draw from the same approved publication, so a brand does not have to maintain two conflicting versions of its story.

Why give a Brand Agent a public home?

Help customers move from interest to understanding

An ad, search result, or recommendation can introduce a product, but a person may still need to understand fit, availability, terms, or the difference between options. The Public Profile gives them a direct route to the brand's selected products and official destinations. They can ask the Brand Assistant a question before they browse the catalog or leave for another site.

An answer should use the information the brand has published. If the available material does not support a claim, the assistant should avoid inventing one and offer a useful next step, such as the official product page or a person on the brand's team.

Give AI agents information they can actually use

A person's shopping or research agent may arrive at the brand's Human page first. From there it can follow the Agent link, read the structured profile or Markdown, and inspect product and source URLs. An agent with the appropriate integration can also query the published Brand Agent directly.

This makes the brand easier to understand programmatically. It does not guarantee that every AI assistant will discover the page, call the API, recommend the brand, or repeat its claims. The external assistant controls its own retrieval and response.

Keep the public story under brand control

The brand chooses which products and offers to feature, which pages to approve as sources, and what action visitors should take. The profile can be saved privately, previewed, and published explicitly. Changes to the draft stay private until the brand publishes again. When a source or offer is no longer eligible, the public experience should stop presenting it as current.

That control matters because a plausible answer is not always a correct one. A product link, an offer's terms, and a policy page can change independently. The Public Profile gives the brand an operating process for reviewing the information behind answers rather than relying on a static prompt.

A simple example

Imagine a fictional outdoor brand, Ridgeway, with three hiking backpacks. Its team publishes a short brand summary, the three product pages, an approved sizing guide, a return-policy page, and a link to shop on its official site.

A visitor asks, “Which pack is better for a weekend trip?” The Brand Assistant can compare the published capacities and intended uses, point to the relevant products, and cite the sizing guide where it supports the answer. If the visitor asks whether their personal order has shipped, the public profile alone cannot know. That requires a separately authorized account lookup or a handoff to the brand's support team.

The same distinction applies when an AI agent works for that visitor. It can use the public profile to research Ridgeway's backpacks. Acting on the visitor's account or making a purchase requires customer authorization and a separate commerce connection; reading a public profile is not permission to buy.

How a brand prepares its profile

The setup begins with information already saved in the Brand Agent where available. The brand then reviews five areas:

1. Brand Profile: name, summary, logo, website, and the way the brand appears publicly.
2. Products: the items worth featuring and their official destinations.
3. Offers: the offers the brand wants considered, subject to current eligibility and terms.
4. Knowledge: official pages and other trusted sources to support answers.
5. Experience: the visitor's next action and which public question-answering capabilities to enable.

After saving, the team can open a private preview of the Human and Agent views. Publishing is a separate decision. A brand can also choose whether its published profile is listed in the AdMesh brand directory and made available for search-engine indexing.

Where other channels fit

The Public Profile is the approved public context, not an automatic connection to every messaging or advertising platform. A brand that uses ManyChat can separately configure an External Request so its Brand Agent helps answer eligible Instagram or WhatsApp conversations. ManyChat remains responsible for the channel connection and message delivery. Account-specific information or actions need customer verification and authorization by the brand's own connected system.

The same principle applies to AI shopping agents. Public product information can help an agent research a choice; checkout and payment need live merchant systems, explicit authority, and their own transaction flow. A Public Profile helps an agent understand the brand before those later steps.

Start with one useful customer question

A strong profile need not begin with the entire catalog. Choose a product group, add the pages that answer real pre-purchase questions, check the customer action, and preview what both a person and an agent will see. Then test a question the profile should answer and one it should decline.

The goal is straightforward: when someone asks about your brand, give them a credible path from a question to the right information and the right next step. Brand Agents Explained introduces the broader Brand Agent; AdMesh for brands shows where to begin.