Article — Jonomor
Entity Graphs Explained for Business Owners
By Ali Morgan · Published by Jonomor
What Is an Entity?
In the context of AI systems and structured data, an entity is anything that can be distinctly identified and described. A company is an entity. A person is an entity. A software product, a published work, a methodology — each of these is an entity with its own name, type, and characteristics.
What makes an entity useful to AI systems is not just that it exists, but that it is defined clearly enough for a machine to understand what it is. A company that calls itself "Jonomor" on its website, "jonomor.com" in its schema, and "Jono-Mor" in a directory listing has created three potential entities instead of one. The AI system cannot be certain they are all the same thing.
Entity definition is the process of establishing one canonical name, one type, one description, and one identity for each thing in your business. Once defined, that identity stays locked. Every reference to the entity — on your website, in your schema, on external platforms — uses the exact same form.
What Is a Graph?
A graph is a map of connections. In everyday terms, think of it as a diagram where each circle represents an entity and each line represents a relationship between entities.
Your business already has a graph — you may just not have defined it explicitly. There is a company. That company has a founder. The founder built several products. Each product is part of the company. The company operates in a specific category. These are all relationships, and together they form the entity graph of your business.
The difference between a business that has an implicit graph and one that has an explicit graph is the difference between a business that AI systems might recognize and one that AI systems definitely recognize. An explicit graph is declared in machine-readable structured data — JSON-LD — so that AI systems do not have to infer the relationships from unstructured text. They can read them directly.
A Worked Example: The Jonomor Entity Graph
The Jonomor ecosystem consists of six entities, each with a defined name, a specific Schema.org type, and explicit relationships to the other entities in the graph. This is not theoretical — it is the actual entity graph that is deployed in production across five domains.
- JonomorOrganization
The parent entity. Everything else connects to this. It has a canonical name, a URL, a description, and a list of products it contains.
- Ali MorganPerson
The founder. Connected to Jonomor through a 'worksFor' relationship. Connected to every product through 'creator' relationships. This person entity is the authorship anchor for all published content.
- XRNotifySoftwareApplication
A product entity. Connected to Jonomor through 'isPartOf'. Jonomor declares it through 'hasPart'. The bidirectional connection means AI systems encountering either entity can verify the relationship from both sides.
- MyPropOpsSoftwareApplication
A product entity operating in the property management domain. Same bidirectional connection pattern as XRNotify — isPartOf from the product, hasPart from the parent.
- Guard-ClauseDefinedTermSet
A methodology entity — not typed as software, but as a defined methodology. This type distinction changes which AI retrieval contexts the entity appears in. Type selection is an architectural decision, not a cosmetic one.
- The Neutral BridgeCreativeWork
A research publication entity. Typed as CreativeWork, connected to Jonomor as publisher and Ali Morgan as author. The type signals 'research output' rather than 'software product' or 'service'.
Every relationship in this graph is declared bidirectionally. Jonomor declares that it has XRNotify as a part (hasPart). XRNotify declares that it is part of Jonomor (isPartOf). Both declarations exist in machine-readable JSON-LD schema on their respective domains. An AI system encountering either declaration can verify the relationship from both sides.
This is what a production entity graph looks like. It is not a visualization or a marketing diagram. It is a set of structured data declarations deployed across every domain in the ecosystem, governed by an entity registry that locks every name, type, and relationship.
Why Entity Graphs Matter for AI Visibility
AI answer engines are entity-recognition systems. When they generate an answer, they are assembling information about entities — not about pages or keywords. The entity graph is the data structure that allows an AI system to understand what your business is, what it does, and how its parts relate to each other.
Identity resolution
When an AI system encounters the name 'Jonomor' on multiple surfaces — the main website, a product domain, a LinkedIn profile, a published article — the entity graph allows it to determine that all of these refer to the same entity. Without the graph, the AI has to guess. With the graph, the connection is declared explicitly.
Relationship traversal
If an AI system is asked about XRNotify, the entity graph allows it to discover that XRNotify is part of the Jonomor ecosystem, built by Ali Morgan, and related to three other products. Each hop through the graph adds context that the AI can use to generate a more complete, accurate answer.
Category association
The entity graph encodes what each entity does — not just what it is called. Jonomor is typed as an Organization in the systems architecture domain. XRNotify is typed as a SoftwareApplication in the XRPL monitoring domain. These category associations determine which AI queries each entity is relevant to.
Authority compounding
Every node in the entity graph that is well-defined and consistently referenced adds authority to every other node. The more surfaces where the graph is consistent, the higher the AI system's confidence in the entire network. This is why cross-domain reinforcement — product sites referencing the parent organization — matters so much.
How to Define Your Own Entity Graph
Defining an entity graph does not require technical expertise. It requires clarity about what your business is, what it contains, and who operates it. The technical implementation — JSON-LD schema — comes after the definition work is done.
Start by answering four questions for every distinct thing in your business: What is its canonical name (the exact name that will be used everywhere, with no variations)? What is its type (Organization, Person, SoftwareApplication, CreativeWork, or another Schema.org type)? What are its relationships to other entities (founder, hasPart, isPartOf, author, publisher)? And what is its single-sentence description?
Document these answers in an entity registry. The registry becomes the governance document — nothing gets published, no schema gets written, and no external listing gets created without consulting the registry for the canonical forms. The Jonomor entity registry is a published example of this governance pattern.
Once the registry exists, the next step is implementing the entity definitions in JSON-LD structured data. The JSON-LD Schema for AI Visibility guide covers the complete implementation process. And the AI Visibility Audit provides a 50-point scoring system to evaluate how well the entity graph is implemented across entity clarity, structured data, and citation surfaces.