agao LogoDocs
System Settings

Embeddings Configuration

Configure Azure OpenAI embeddings service for vector search, document retrieval, and semantic operations.

Overview

Configure Azure OpenAI embeddings to enable advanced AI features like document similarity search, semantic content retrieval, and intelligent data analysis. Embeddings are required for vector search capabilities and smart document processing.

Azure OpenAI Embeddings Setup

Configure your Azure OpenAI embeddings service for vector operations and semantic search.

Required Information:

  • Credential Name: Display name for identification (e.g., "Azure OpenAI Embeddings")
  • API Endpoint: Your Azure OpenAI service endpoint (e.g., https://your-resource.openai.azure.com/)
  • Instance Name: Your Azure resource name
  • Deployment Name: Name of your embeddings model deployment (eg. "text-embedding-3-small")
  • API Key: Your Azure OpenAI API key
  • API Version: Azure OpenAI API version

Setup Instructions:

  1. Create an Azure OpenAI resource in your Azure subscription
  2. Deploy an embeddings model (recommended: text-embedding-ada-002)
  3. Obtain your API endpoint, resource name, and API key from the Azure portal
  4. Configure the embeddings credentials in your platform settings
  5. Save the configuration to enable embedding features

Supported Embedding Models:

  • text-embedding-3-small

Testing Your Configuration

After setting up embeddings credentials, verify the configuration works correctly:

How to Test

  1. Save Configuration: Complete all required fields and save your embeddings credentials
  2. Check Status: The system will automatically test the connection and show the status
  3. Verify Features: Test document upload and search functionality to confirm embeddings are working

Troubleshooting

Common Issues:

  • Invalid API Key: Ensure your Azure OpenAI API key is correct and has embeddings permissions
  • Wrong Endpoint: Verify the API endpoint matches your Azure OpenAI resource URL
  • Model Deployment: Confirm your embeddings model is deployed and the deployment name is correct
  • Network Access: Ensure your platform can reach the Azure OpenAI endpoint

Configuration Errors:

  • Double-check all field values match your Azure OpenAI resource settings
  • Verify the deployment name exactly matches what's configured in Azure
  • Test the endpoint URL and API key using Azure's test interface first
  • Check Azure OpenAI service status and quota limits