From Know-How
to Ontology
Describe your domain in plain language or hand over a policy. The AI Assistant drafts the concepts, writes the rules and runs them on your data where it lives, as readable code you review, own and export.
| concept | rows |
|---|---|
| customer | 12,480 |
| placed | 88,312 |
| contains | 241,907 |
Proven in production
Months of Modelling.
Or One Conversation.
Classic modelling projects need specialists, OWL/RDF tooling or a platform you must load your data into, and a long wait before anything runs. Prometheux starts from what your experts already say and gets to running logic fast.
Say It Once.
Get Running Logic.
The AI Assistant works inside your workspace. Ask it to create an ontology, add entity and relationship types, or turn a rule from a policy into Vadalog. It saves concepts, runs them and shows you the results.
Prompt to Logic
Ask for something like "Supply Chain Risk for analysing supplier data" and the assistant creates it, writes the concepts and runs them.
Concepts From Documents
Upload a PDF, policy, glossary or regulatory text. The assistant chunks and embeds it, then extracts concepts from it.
Schema You Approve
The assistant reads and extends the schema with entity and relationship types, and asks before making changes.
Self-Repairing Rules
It writes, runs and repairs failing Vadalog code in a loop, or explains what is wrong with the data or the rule.
Go from domain knowledge to working business rules and definitions in a session. Let experts model without learning a new language first. Keep humans in charge of every schema change.
From Brief to
Working Rules
Describe the Domain
Tell the assistant what you analyse, or upload the policy, glossary or spec that defines it.
Connect the Data
Point at a warehouse or database, or upload CSV, Excel, Parquet or JSON. No data to sample yet? Generate mock facts.
Review the Draft
Check the proposed entity types, relationships and rules. Approve schema changes before they are applied.
Run and Refine
Run the concepts, inspect results and lineage, and iterate in conversation or in the concept editor.
Built for the
Work You Already Do
Policy to Rules
Turn a regulation or internal policy document into concepts and rules that run against your data, not a checklist in a PDF.
Supplier Risk
Model suppliers, tiers and disruptions from a supplies table in one conversation.
Prototype Before Integration
Generate sample facts to test rules before production sources are connected.
Reuse Across Teams
Start from marketplace templates, then copy or export them between workspaces.
“Generative AI without ontologies is an engine without a road. Prometheux builds the road.”
Marco RomeiDeputy CIO · Banca Sella
Questions,
Answered
Do I need to know OWL, RDF or Vadalog to build an ontology?
No. You describe the domain in plain language and the AI Assistant generates the Vadalog for you. The rules stay readable, so engineers can review and edit them.
Can it build rules and definitions from our documents?
Yes. Upload a PDF, policy, glossary or regulatory text and the assistant chunks and embeds it, then extracts concepts from it into your workspace.
Will the AI change my model without asking?
No. Schema changes such as new entity or relationship types need your confirmation, because they are hard to reverse.
Do we have to load our data into a new platform?
No. Generated rules and definitions are connected to your existing warehouses, databases and files, and run where the data lives.
Can agents and apps use the generated model?
Yes. Saved concepts run through the REST API, Python SDK and MCP, and can back Prometheux apps with tables, charts and cards.
Is the generated logic tied to the assistant?
No. It is ordinary Vadalog connected to your data sources. You can edit it in the concept editor, call it through the APIs or SDK, and export or copy it.
Related Capabilities
Turn Expertise
Into Running Rules
Bring a policy, a glossary or a domain expert. We'll generate working business rules and definitions on your own data.
