feat: add Azure AI multi-provider support (TTS, STT, Embeddings, Realtime)

Enables Azure AI services across all model layers so users with Azure
credits can consolidate billing on a single provider.

- Voice (TTS): AzureSpeechTTSConfiguration via azure_speech provider
- Transcriber (STT): AzureSpeechSTTConfiguration via azure_speech provider
- Embedding: AzureOpenAIEmbeddingsConfiguration via azure provider
- Realtime: AzureRealtimeLLMConfiguration via azure_realtime provider

New files:
- api/services/pipecat/realtime/azure_realtime.py
- api/services/gen_ai/embedding/azure_openai_service.py
- api/tests/test_azure_speech_service_factory.py

The UI picks up all four providers automatically from the schema —
no frontend changes required.

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

api/services/gen_ai/embedding/__init__.py

@@ -1,9 +1,12 @@
 """Embedding services for document processing and retrieval."""
 
+from .azure_openai_service import AzureEmbeddingAPIKeyNotConfiguredError, AzureOpenAIEmbeddingService
 from .base import BaseEmbeddingService
 from .openai_service import EmbeddingAPIKeyNotConfiguredError, OpenAIEmbeddingService
 
 __all__ = [
+    "AzureEmbeddingAPIKeyNotConfiguredError",
+    "AzureOpenAIEmbeddingService",
     "BaseEmbeddingService",
     "EmbeddingAPIKeyNotConfiguredError",
     "OpenAIEmbeddingService",

api/services/gen_ai/embedding/azure_openai_service.py

@@ -0,0 +1,119 @@
+"""Azure OpenAI embedding service.
+
+Uses the Azure OpenAI REST API for text embeddings, compatible with
+text-embedding-3-small, text-embedding-3-large, and text-embedding-ada-002
+deployments.
+"""
+
+from typing import Any, Dict, List, Optional
+
+from loguru import logger
+from openai import AsyncAzureOpenAI
+
+from api.db.db_client import DBClient
+from api.utils.url_security import validate_user_configured_service_url
+
+from .base import BaseEmbeddingService
+
+DEFAULT_MODEL_ID = "text-embedding-3-small"
+EMBEDDING_DIMENSION = 1536
+
+
+class AzureEmbeddingAPIKeyNotConfiguredError(Exception):
+    """Raised when Azure OpenAI credentials are not configured for embeddings."""
+
+    def __init__(self):
+        super().__init__(
+            "Azure OpenAI endpoint or API key not configured. Please set your "
+            "endpoint and API key in Model Configurations > Embedding to use "
+            "document processing."
+        )
+
+
+class AzureOpenAIEmbeddingService(BaseEmbeddingService):
+    """Embedding service using Azure OpenAI text-embedding deployments."""
+
+    def __init__(
+        self,
+        db_client: DBClient,
+        api_key: Optional[str] = None,
+        endpoint: Optional[str] = None,
+        model_id: str = DEFAULT_MODEL_ID,
+        api_version: str = "2024-02-15-preview",
+    ):
+        """Initialize the Azure OpenAI embedding service.
+
+        Args:
+            db_client: Database client for vector similarity search.
+            api_key: Azure OpenAI API key.
+            endpoint: Azure OpenAI resource endpoint (e.g. https://<resource>.openai.azure.com).
+            model_id: Deployment name, used as both the deployment and model identifier.
+            api_version: Azure OpenAI API version.
+        """
+        self.db = db_client
+        self.model_id = model_id
+
+        self._configured = bool(api_key and endpoint)
+        if self._configured:
+            validate_user_configured_service_url(endpoint, field_name="endpoint")
+            self.client = AsyncAzureOpenAI(
+                api_key=api_key,
+                azure_endpoint=endpoint,
+                api_version=api_version,
+            )
+            logger.info(
+                f"Azure OpenAI embedding service initialized with deployment: {model_id}"
+            )
+        else:
+            self.client = None
+            logger.warning(
+                "Azure OpenAI embedding service initialized without credentials. "
+                "Operations will fail until endpoint and API key are configured."
+            )
+
+    def get_model_id(self) -> str:
+        return self.model_id
+
+    def get_embedding_dimension(self) -> int:
+        return EMBEDDING_DIMENSION
+
+    def _ensure_configured(self):
+        if not self._configured or self.client is None:
+            raise AzureEmbeddingAPIKeyNotConfiguredError()
+
+    async def embed_texts(self, texts: List[str]) -> List[List[float]]:
+        """Embed a batch of texts using Azure OpenAI API."""
+        self._ensure_configured()
+        try:
+            response = await self.client.embeddings.create(
+                input=texts,
+                model=self.model_id,
+            )
+            return [item.embedding for item in response.data]
+        except Exception as e:
+            logger.error(f"Error generating Azure OpenAI embeddings: {e}")
+            raise
+
+    async def embed_query(self, query: str) -> List[float]:
+        """Embed a single query text using Azure OpenAI API."""
+        self._ensure_configured()
+        embeddings = await self.embed_texts([query])
+        return embeddings[0]
+
+    async def search_similar_chunks(
+        self,
+        query: str,
+        organization_id: int,
+        limit: int = 5,
+        document_uuids: Optional[List[str]] = None,
+    ) -> List[Dict[str, Any]]:
+        """Search for similar chunks using vector similarity."""
+        self._ensure_configured()
+        query_embedding = await self.embed_query(query)
+        return await self.db.search_similar_chunks(
+            query_embedding=query_embedding,
+            organization_id=organization_id,
+            limit=limit,
+            document_uuids=document_uuids,
+            embedding_model=self.model_id,
+        )