
The NexusAI ontology
A verified model of how your business works, so AI reasons over facts and rules instead of guessing.
What is the NexusAI ontology?
The ontology is a model of your business that AI can read. It maps customers, orders, assets, contracts and processes into one semantic layer with four parts: objects, the links between them, the rules that apply, and the actions that can be taken.
Agents reason over these verified facts and links, and every action they can take is defined on the object itself. That is how answers stay explainable and actions stay inside your policies.
Objects
The things your business runs on: Customer, Order, Invoice, Product, Network site.
Links
How those things relate: a customer places an order, an order is billed by an invoice.
Rules
What is allowed and who decides, for example credit limits and approval thresholds.
Actions
What can be done to each object, such as sending a payment reminder or expediting an order.
A verified model of how your business works
Pick an object to see its links, the rules that apply and the actions an agent may take.
Customer
Segmentlifetime valuechurn riskconsent
Linked toholdsAccountsignsContractplacesOrderraisesTicket
Actions agents can take- Offer a retention plan
- Escalate to the account manager
Account
Balancestatusproduct holdings
Linked toholdsCustomersettlesInvoice
Actions agents can take- Flag unusual activity
- Recommend the next product
Contract
TermSLArenewal datevalue
Linked tosignsCustomercoversNetwork site
Actions agents can take- Prepare the renewal
- Alert on SLA breach
Invoice
Amountdue dateageingdisputes
Linked tobilled byOrdersettlesAccount
Actions agents can take- Send a payment reminder
- Resolve a deduction
Network site
Locationhealthtrafficrevenue
Linked toaffectsTicketcoversContract
Actions agents can take- Run a remote reset
- Dispatch a field engineer
Order
Itemsvaluechannelfulfilment status
Linked toplacesCustomerbilled byInvoicecontainsProduct
Actions agents can take- Reroute fulfilment
- Expedite a delayed order
Ticket
Issueprioritysentimentowner
Linked toraisesCustomeraffectsNetwork site
Actions agents can take- Auto-resolve
- Assign to the right team
Product
SKUpricestock levelmargin
Linked tocontainsOrdersourced fromSupplier
Actions agents can take- Reorder stock
- Adjust the price
Supplier
Reliabilityrisk scorelead time
Linked tosourced fromProduct
Actions agents can take- Switch to a backup supplier
- Renegotiate terms
Select any object in the graph
Guides from the founders
What is an enterprise ontology? A plain-English guide explains the four parts with examples, and Why enterprises need a second brain, not a bigger AI model explains why an ontology matters more than model size.
Questions leaders ask
What is an enterprise ontology?
A shared model of how a business works: the objects it runs on (customers, orders, invoices, assets), the links between them, the rules that apply and the actions that can be taken. AI that reads it understands your business in your own terms.
How is an ontology different from a knowledge graph?
The knowledge graph stores objects and the links between them. The ontology adds meaning, rules and actions on top, so an agent knows not only what is connected but what it is allowed to do.
Do we need clean data before building an ontology?
No. Mapping rules match records on real keys, translate codes once and trace every field back to its source, so the model is built from the data you already have.
Does the ontology copy our data?
No. It is built on the data plane, which reads your systems where they live. The ontology describes the data; the data stays in the source systems.
Why does an ontology make AI more reliable?
Agents answer from verified objects and links instead of guessing, follow the rules attached to each object, and can only take the actions defined on it. Every answer can be traced back to its sources.
Get started
See your business as an ontology
Turn enterprise systems into an intelligent operating layer that understands your business and executes operations autonomously.