Setup
Connect KnowOS to Claude, ChatGPT or any MCP client.
One endpoint and the same contract for all of them. Switch models without touching the integration.
Under the hood · For developers
How KnowOS is built.
Every product value comes in with its source and date, is modeled with business names and is served via MCP and API, with permissions enforced before every answer.
Architecture
Five layers, from source system to answer.
01 · Sources
Integrafy connectors, files and APIs.
- ERP
- PLM
- eCommerce
- CSV and PDF
- REST API
02 · Curation
Every value with date, source and status. Only what changes gets reindexed.
- Source
- Date
- Status
- Versioning
03 · Product model
Objects, attributes, relationships and actions with business names.
- PIM
- MDM
- DAM
- Relationships
04 · Permissions
Enforced before answering, by role, row, column and action.
- Role
- Rows
- Columns
- Actions
- Log
05 · Delivery
Hybrid retrieval and output to your tools.
- MCP
- REST
- GraphQL
- Agents
MCP server
Your catalog, inside your AI.
Tools
Reading needs no approval. Changing anything always does.
API
REST and GraphQL generated from the product model.
Permissions
Per token: role, rows, columns and actions.
Technical questions
What APIs does KnowOS offer?
REST and GraphQL APIs generated from the product model, plus an MCP server for AI clients.
Do I need a data team?
No. Ingestion, modeling and permissions are configured without code during onboarding.
Does it write to my ERP?
Not in the first phase. After that, only approved actions, executed through Connect or Flow.
Is there an on-premise option?
On request. By default, a dedicated EU server in your name.
Connect your first source and start asking your catalog.
In 30 minutes we work through a real business question using your own data. You leave with a defined scope and a fixed price.