💡 Who is this for? This guide is for Account Admins who need to connect a custom Azure OpenAI resource to power wxrks Augmented Actions within the wxrks platform.
Introduction
If you'd rather use your own Azure OpenAI resource instead of wxrks's default AI integration, you'll need to retrieve three values from the Azure Portal — the Resource Name, Deployment Name, and an API Key — before you can point wxrks Augmented Actions at it. In this article, we'll guide you through the process step by step.
Prerequisites
An active Azure account.
A resource and deployed model using Azure OpenAI.
Step-by-Step Configuration
1. Log in to your Azure Portal:
Begin by navigating to the Azure Portal at https://portal.azure.com. Enter your credentials to access your dashboard.
2. Locate your Azure AI Services resource:
From the left-hand menu, click on All resources.
Locate the Azure AI services resource you created earlier. This can be identified by the resource's name.
3. Retrieve the Resource Name:
Click on the AI resource to open its overview page.
Find the name of your Azure AI services resource at the top of the page. This is also the first part of the base URL, i.e.,
https://{resourceName}.openai.azure.com.
4. Locate the Deployment Name:
While still in the AI resource's overview page, navigate to the Deployments section. Here, you can identify the name of the model you deployed on the Azure OpenAI Service.
5. Retrieve the API Key:
From the AI resource's overview page, navigate to the Keys section.
Copy the primary key (KEY 1). It is the value you will paste into the API Key field in wxrks, and it authenticates requests made from wxrks to Azure OpenAI Service.
6. Configure wxrks Augmented Actions with Azure OpenAI Details:
Once you have the Resource Name, Deployment Name, and API Key above, please follow the steps in Enabling AI Integration in wxrks through OpenAI or Microsoft Azure to enter them into wxrks.
wxrks always connects to {resource name}.openai.azure.com. Custom domains, Azure API Management and an API version field are not supported. wxrks tests the credentials when you click Save: if the test fails, the row isn't saved and you see Could not validate followed by the provider, the model and the provider's error message.
Note: configuring a custom AI provider requires an eligible wxrks plan and Account Admin permission.
Related articles: wxrks Augmented Actions — how to enable and use wxrks's LLM-powered translation and editing actions; Configure Custom AI Providers with Model Management — every provider you can connect; and Connect your own AI engine for automatic post-editing (Context Sensitive) and quality evaluation (Smells) — route a specific operation to your own engine.