Azure AI

1. Create an Azure account

Before using Azure AI, an account and subscription must be created.

• Go to the Azure portal

• Sign in or create an account

• Ensure you have an active Azure subscription

https://portal.azure.com

---

2. Create Microsoft Foundry Project

Azure organizes AI usage through resources.

• Go to Create resource

• Search for Azure OpenAI or Azure AI Foundry

• Create a new resource

Configuration:

• Resource group: create or select one

• Region:

  • West Europe (recommended)

  • Sweden Central (for latest models like GPT-5.x)

• Name: use a clear naming convention (e.g. ai-studio-prod)

This resource contains:

• Models and deployments

• API access (endpoint + keys)

• Monitoring and usage data

---

3. Deploy a model

Before using Azure AI, a model must be deployed.

• Open your Azure AI resource

• Go to:

Models + Endpoints

• Create a new deployment

• Select a model (GPT-4o transcribe, GPT-5.x)

Important:

• Define a deployment name

• This is what you will use in Any2Info (NOT the model name but it can be the same)

---

4. Retrieve API key and endpoint

After deploying a model, you can retrieve the required credentials.

• Open your Azure AI resource

• Go to Models + Endpoints

• Select your deployed model

From here, you can find:

• API Key

These values are required to connect Azure to the Any2Info platform.

---

5. Cost monitoring

There are two places where you can monitor usage and costs:

Model usage (per deployment)

• Open your Azure AI resource

• Go to:

Models + Endpoints

• Select your deployment

• Go to Metrics

Here you can see:

• Token usage

• Requests

• Performance metrics

Azure cost management

This can be in the azure portal