Azure AI Foundry connector for model inference
Azure AI Foundry is the default AI inference provider for the Demiton platform. The adapter invokes Azure AI Foundry’s deployment API for the Studio chat interface, document intelligence, and workflow LLM steps.
Authentication: Platform-managed (Entra ID managed identity)
System type: AZURE_FOUNDRY
API reference: Azure AI Foundry
Run Demiton inference on your Azure AI Foundry deployment — book a 30-minute call.
Integration model
Section titled “Integration model”This connector is configured and managed by Demiton as part of your deployment - you do not supply credentials or manage the connection yourself.
The adapter is currently in Beta. The CHECK_HEALTH command is implemented for connectivity verification. Full deployment management integration is in progress.
Connection setup
Section titled “Connection setup”No tenant credentials are required or accepted. This connector is configured and managed by Demiton as part of your deployment.
Security
Section titled “Security”Access type: Platform-managed - inference only
Demiton calls the Azure AI Foundry inference API under its managed identity. Document content and query context is sent to the model for the duration of inference only - it is not stored by the AI provider beyond the request lifetime.
No customer data is used for model training. All inference runs under Demiton’s managed identity, not under individual user credentials.
Next steps
Section titled “Next steps”- Connecting a System - step-by-step wizard for setting up the connector
- Civil Memory - where connector data lands and is retrieved from
- MCP Server - cognitive capabilities power the AI layer behind MCP queries
- Connect your operation - talk to us about scoping a Connected deployment
Talk to the team
Tell us what you run - your ERP, field, payroll and document systems - and we will show you the registers Demiton would keep current from them, on your data, in a 30-minute call.