Which orders with tire complaints trace back to Supplier A’s September 2025 batches, and what is their total value?
Ontology infrastructure · By UINO
Your business ontology.
Answers you can trace.
Define object types, map your data to them, and follow relationships from a question to its source records.
Product maturity01
RawGrowing in the open.
Current phase
Supplier → Material batch → Product → Order → Evidence
Scroll to follow the connections
The whole map is your ontology. Select an object type to inspect it.
3,000+ customers choose UINO.
Finance, energy, telecoms, industry and the public sector.
A small demo. A complete path.
Watch a question
become an answer.
Ask a business question. Ontomato follows the ontology and replies with a result you can inspect, down to the records behind it.
OntomatoExample workspace
Ontomato
Supplier1
Material batch2
Product2
Order3
Feedback4
Following the ontology
I found 3 orders totaling $99,500 linked to Supplier A’s September batches. They match 4 tire-related feedback records. Here’s the breakdown, with each order counted once.
Supplier impact overview
Supplier A · September 2025 batchesAll tire reports
- Total order amount
- $99,500
- Matching orders
- 3
- Feedback records
- 4
Order amounts
USDAll reportsConfirmed only
Each order is counted once, even if several feedback records refer to it.
Object types describe the path. These counts show the source records represented in this answer.
Supplier1Material batch2Product2Order3Feedback4Source fieldOntology property
orders.productOrder.productIdorders.amountOrder.amountfeedback.orderFeedback.orderIdThe relationship matches orders.product to products.id. Batch supply dates set the scope.
Open a feedback record to read the source passage used in this example.
An interactive product illustration with fictional data and preset questions. The animation is a walkthrough; the totals are calculated from the sample records.
See the ontology at work
One ontology.
For people and agents.
Filter, inspect, and calculate in a familiar spreadsheet. Or let an agent request the result. One ontology keeps their business definitions consistent.
One question. Two ways of working.
What is the value of tire-issue orders linked to Supplier A’s September 2025 batches?
E9ƒx
=SUBTOTAL(109,E2:E7)| 1 | Order⌄ | Product | Batch | Feedback⌄ | Amount · USD |
|---|---|---|---|---|---|
| 2 | SO-1042 | PR-CITY | MB-A0901 | Tire · confirmed | 48,000 |
| 3 | SO-1048 | PR-ROAD | MB-A0902 | Tire · investigating | 32,500 |
| 4 | SO-1051 | PR-CITY | MB-A0901 | Tire · confirmed | 19,000 |
| 9 | Filtered total | 99,500 | |||
Review the subtotal3 matching orders
- 01Open the sheet
- 02Filter with meaning
- 03Select the amounts
- 04Review the total
query_ontology
Ask for the result and its source references.
Following the ontology…
Resolve the scope, apply the definition, and sum each order once.
Result returnedJSON · with evidence
{
"orders": 3,
"total_amount": 99500,
"currency": "USD",
"order_ids": ["SO-1042","SO-1048","SO-1051"],
"evidence_ids": ["FB-701","FB-702","FB-704","FB-708"],
"definition": "all_tire_reports"
}Illustrative JSON · Same calculated resultA tool call in. Structured data out.Example resolved
- 01Send request
- 02Resolve context
- 03Apply rules
- 04Return JSON
Shared business context
Example resolvedOne ontology underneath.
Supplier→
Material batch→
Product→
Order→
Feedback
- Source mapping
orders.product Order.productId- Question scope
- Supplier A · Batch supplied in September 2025
- Business rule
tire_issue = topic = "tire"- Calculation
- Sum order amounts. Count each Order.id once.
Change the definition. Watch the result change.
Reported tire issues, including those still under investigation.
5 related object types4 matching feedback records
Different workflows. Shared definitions. The same answer.
Inspect the evidenceIllustrative workflows with fictional records. The spreadsheet and agent use the same local calculation. No Excel connection, live AI, or external API.
Inspect the types, mappings, and source records
Object type
Order
A recorded sale with a customer, a product, and an amount. Its identity keeps one order distinct from the feedback about it.
The type defines the structure. Individual records supply the values.
PropertyValue type
- id
- Identifier
- customerId
- Reference
- productId
- Reference
- amount
- Amount · USD
One definition, shared across the example.
Local sample sourceexampleData.orders
6 source records| Source field | Ontology property | Sample value |
|---|---|---|
orders.id | Order.id | SO-1042 |
orders.customer | Order.customerId | horizon |
orders.product | Order.productId | PR-CITY |
orders.amount | Order.amount | 48000 |
Named relationshipOrder refers to Product
orders.product = products.idField names differ. The business meaning stays explicit.
Interactive example · Fictional records
Which orders with tire complaints trace back to Supplier A’s September 2025 batches, and what is their total value?
Includes reported tire issues, even when the cause is still being investigated.
Change the definition. Watch the answer change.
Total order amount$99,500
3 matching orders
| Customer / order | Amount (USD) |
|---|---|
| Horizon CyclesSO-1042 | 48,000 |
| NorthlineSO-1048 | 32,500 |
| Atlas MobilitySO-1051 | 19,000 |
Each order is counted once, even when it has more than one feedback record.
This example follows a fixed plan across local sample records. Inspect the scope, calculation, and source evidence.
- A
Find the scope
Supplier A → batches supplied in September 2025 → products → orders → tire feedback.
- B
Select the facts
Keep one row per order. Read its customer and recorded order amount. Multiple feedback records do not multiply the amount.
- C
Sort and aggregate
Sort order amounts from highest to lowest, then sum the distinct matching orders.
The date filter belongs to the material batch’s supply date, not the order date.
Open an order to follow its source records. These are the fictional records used to calculate this example.
This walkthrough uses a fixed question and local sample data. Change the definition to recalculate the result; no live database or AI service is connected.
Object types give data structure
First, define
your world.
An ontology defines the types of things in your business, the properties they carry, and how they relate.
Supplier, Product, and Order are object types. Each defines a shared set of properties for your data and documents.
Follow its relationshipsRelationships give it context
Then, define
how it connects.
Named relationships give AI a path through your business. Follow a supplier’s material batches into products, orders, and customer feedback.
Our example question
Which orders with tire complaints trace back to Supplier A’s September batches?
Start with a real business object.
Supplier A is an object with an identity. Its relationships tell us which material batches it supplied.
Rules give it meaning
Your business rules
shape the answer.
A shared definition changes which records belong in an answer. The example’s current definition gives this result; narrow it to confirmed defects to see what changes.
Current definition: All tire reports
$99,5003 matching orders
Supplier A · September 2025 batch supply dates · Each order counted once
Try changing the definition
What makes an answer useful
The answer is
only the beginning.
A total is useful when you can explain its scope, check its calculation, and find the records behind it.
The right scope
“September batches” filters the batch supply date. The relationship then finds the products and orders in scope.
Inspect the planA stable identity
Two feedback records can refer to one order. The example counts that order’s amount once.
Check the calculationThe original records
Follow a result through its batch, product, order, and feedback. See why each record belongs in the answer.
Read the source evidenceThe context inside your contracts.
Model the contract as an object. Use its dates and terms to narrow the search, then return to the source passages for evidence.
“Which supplier contracts expire in Q1 2027, have payment terms longer than 60 days, and contain no liquidated-damages clause?”
Document workflow preview · Coming soon.Made to be yours
Build with your
own ontology.
The open-source project connects ontology modeling, query planning and source evidence. Explore the source and follow the repository guide to self-host.
Planned model and data integrations Coming soon
Published sourceApache-2.0
Object types & relationships01
Source mappings02
Questions & source evidence03
Agent skills04
REST API & MCP05
Keep exploring
A question is only the beginning.
Product
Understand the model, its mappings, and the direction of the product.
Resources
Practical guides to types, relationships, and answers with evidence.
Developers
Run the supply-chain calculation with readable source records.
Aohan / OntomatoApache-2.0
Open the source. Build with Ontomato.
Read the model, explore the implementation and follow the published self-hosting guide.
Taking shape, in the open.
Source code, the license and contribution guide are published. Explore the website guides still being prepared.
Start somewhere useful
Start with a question.
See how it is answered.
Try the supply-chain example now. Explore its object types, inspect the field mappings, and change a definition to see the result update.
Try the working exampleNo signup or API key needed for this example.
- Try today
- A local example with type definitions, field mappings, a rule you can change, and inspectable source records.
- Release status
- The repository and self-hosting guide are available. Some website guides remain previews; deployment validation and live-data connections are still in progress.
- Source & license
- The project source uses Apache-2.0. The license text and NOTICE are published in the repository.