Archon
Archon copied to clipboard
Feature Request: Add the ability to use Azure Open AI LLM's
Hello, A lot of users will probably want to deploy this in their organisation which has taken onboard the Azure model to govern and consume there AI units.
We are working on OpenAI API compatibility which should help with this! I do know that the Azure Open AI integration is trickier though since there are extra parameters for the client if I remember correctly... something like the organization ID needs to be passed in when initializing the OpenAI client?
Summary of requirements on the Azure Platform
| Requirement | Details |
|---|---|
| Azure Subscription | Needed for resource allocation |
| Resource Deployment | Model selection, deployment type, region-specific availability |
| Authentication | API Key or Entra ID (Azure AD); assign RBAC roles |
| API Parameters | Endpoint URL, API key/token, Deployment Name, sometimes Organization ID |
| Network & Security | CORS for storage, VNet/private endpoint for advanced security |
| Compliance | Align with standards (GDPR, HIPAA, SOC2, etc.) |
| SDK/Client Libraries | Use official clients for Python, .NET, Java, others |