June 1, 2026
What Is an Ontology?
An ontology is an organized map of what something is — the names of things, and how they connect.
That is the whole idea. Everything else is detail.
Start with a messy room
Imagine a room full of toys. No labels. No boxes. Someone asks: Where is the red ball? Who plays with it? Where does it go when we are done?
You know the answers because you live here. You carry the map in your head.
An ontology is that map — written down, kept current, and shared.
Not a list. A map.
A spreadsheet might say:
- ▸Red ball
- ▸Blue truck
- ▸47 blocks
That is a list. It tells you what exists. It does not tell you what anything means or relates to.
An ontology says:
- ▸The red ball → lives in → the sports box
- ▸The sports box → sits on → the top shelf
- ▸Maya → plays with → the red ball on Tuesdays
Same objects. But now there are names and connections. You can follow the map instead of guessing.
What that looks like at a company
Companies have the same problem at a larger scale.
- ▸Alice → works on → Project Atlas
- ▸Project Atlas → is owned by → the Infrastructure team
- ▸That hiring decision → supersedes → last quarter's plan
An ontology captures those names and relationships in one structured place. Not as a pile of PDFs. Not as rows in a table with no context. As a graph — things linked to other things, with types and history.
Why agents need it
An AI that only searches documents is like someone rummaging through drawers, hoping the right note turns up. Sometimes it works. Often it does not.
An AI with an ontology can follow the map: find the thing, see what it connects to, and know which connection is authoritative.
That is why Dominir is built around a company ontology — a living map of how your organization actually works, not just what was written down somewhere.
Want to go deeper? Read The Company Ontology — what that map looks like when it is built for production.