Skip to content

Azure AI Proxy

Introduction to the Azure AI Proxy

The goal of the Azure OpenAI proxy service is to simplify access to an AI Playground experience, support for Azure OpenAI SDKs, LangChain, and REST endpoints for developer events, workshops, and hackathons. Access is granted using a time bound API Key.

There are four primary use cases for the Azure OpenAI proxy service:

  1. Access to an AI Playground experience for developers to explore the Azure OpenAI chat completion using a time bound event code and different models and parameters.
  2. Access to developer APIs via REST endpoints and the OpenAI SDKs and LangChain. Access to these services is granted using a time bound event code. Initially, the proxy service supports the chat completion, completion, embeddings, and dall-e-3 APIs.
  3. Access to Azure AI Search queries using the Azure OpenAI proxy service. Access to these services is granted using a time bound event code.
  4. You are running a hackathon and users can't provision their own Azure OpenAI resources as they don't have a corporate email address.

OpenAI Proxy Playground

The Azure OpenAI proxy service provides a Playground-like experience for developers to explore the Azure OpenAI chat completion using the time bound event code with different models and parameters.

OpenAI Proxy Playground

Azure AI Proxy Architecture

The Azure AI Proxy consists of the following components:

  1. Self-service event management. Configure and manage events and resources for the events.
  2. Self-service attendee registration. Attendees can register for an event and receive a time bound API Key to access the AI Proxy service.
  3. The AI Proxy service. The AI Proxy service provides access to the Azure AI resources using a time bound API Key.
  4. The AI Playground. The AI Playground provides an AI Playground experience for developers to explore the Azure OpenAI chat completion using the time bound event code with different models and parameters.