By industry

Partner-directory entity graph

The set of verifiable third-party profiles, from vendor partner directories to GitHub and Hugging Face, that an AI agency links from its Organization markup with sameAs to prove who it is.

By Shimon Carroll, Founder, SEO for AI Agents · Last updated

An entity graph is the web of consistent, linked facts that lets a search engine or AI system decide that a name refers to one specific, real organization. For an AI agency, the most persuasive nodes in that graph are the ones that are hard to fake: listings in vendor partner programs such as Google Cloud Partner Advantage, the AWS Partner Network, or the Microsoft AI Cloud Partner Program, plus any model-provider partner programs you genuinely belong to; the company's GitHub organization with real repositories; a Hugging Face organization with published models or datasets; and the company's LinkedIn and Crunchbase profiles.

The mechanics are simple. In the Organization JSON-LD on your homepage, list each of those profile URLs in sameAs. Make sure the company name, website, and description match on every profile, and that the profiles link back to your site where they allow it. Do the same for founders and principal engineers with Person markup, linking their GitHub, Google Scholar, conference talks, and LinkedIn.

Two cautions. Only list partner programs you are actually in; a claimed partnership that the directory contradicts is worse than none. And sameAs is corroboration, not magic: it helps an engine connect facts that already exist. Our audit checks the Organization and Person markup, tests that each sameAs link resolves, and flags name or URL mismatches across profiles.

How the audit measures this

Entity graph: can AI engines resolve your brand?Methodology →

Primary sources