Azure OpenAI

Azure OpenAI chat-completions via a per-deployment path, selected with LLM_PROVIDER=azure_openai, used to extract facts from conversations.

Enable it

LLM_PROVIDER=azure_openai
AZURE_OPENAI_API_KEY=<azure-api-key>
AZURE_OPENAI_ENDPOINT=https://<resource>.openai.azure.com
AZURE_OPENAI_DEPLOYMENT=<deployment-name>
AZURE_OPENAI_API_VERSION=2024-06-01

Configuration

Variable	Default	Description
LLM_PROVIDER	openai	Set to azure_openai.
AZURE_OPENAI_API_KEY	(none)	Key sent as the api-key header. Calls fail at runtime if unset.
AZURE_OPENAI_ENDPOINT	(empty)	Resource endpoint, e.g. https://<resource>.openai.azure.com.
AZURE_OPENAI_DEPLOYMENT	(model name)	Deployment name; falls back to the model when unset.
AZURE_OPENAI_API_VERSION	2024-06-01	API version query parameter.

Notes

• Azure authenticates with an api-key header (not a bearer token) and calls chat-completions under /openai/deployments/<deployment>/chat/completions.
• The default model is gpt-4o-mini; the deployment name doubles as the model when AZURE_OPENAI_DEPLOYMENT is unset.