Announcing Deployment capabilities (preview) in Azure Copilot

Marvin_Morales

Microsoft

Nov 19, 2025

What's new?

Today, we are excited to announce the Deployment Agent for Azure Copilot, designed to make infrastructure planning and deployment faster, smarter, and easier. The Deployment Agent acts as your dedicated cloud architecture assistant, leveraging the Azure Well-Architected Framework to help you deploy workloads aligned with best practices.

With multi-turn conversations, the Deployment Agent helps you:

Clarify requirements and translate goals into actionable workload deployment plans.
Get tailored recommendations for optimal resource configurations.
Follow step-by-step guidance to deploy production-ready environments—whether you’re building analytics pipelines, web apps, or complex multi-tier architectures.

By leveraging the Deployment Agent, you can reduce manual effort, minimize errors, and accelerate time-to-value—turning tasks that once took hours into minutes.

How it works?

To start using agents, your global administrator must request access to the agents preview at the tenant level in the Azure Copilot admin center.

This confirms your interest in the preview and allows us to enable access. Once approved, users will see the Agent mode toggle in Azure Copilot chat and can then start using Copilot agents. Capacity is limited, so sign up early for the best chance to participate.

Additionally, if you are interested in helping shape the future of agentic cloud ops and the role Copilot will play in it, please join our customer feedback program by filling up this form.

Describe your goal

Once agent access is granted, navigate to the Azure Portal, select Copilot, and activate Agent mode. Begin by entering your requirements as a natural language prompt—for example, “Help me host a Python application on Azure following production-grade best practices.” The agent interprets your intent and guides you through each subsequent step.
Get a best-practice workload plan
The agent automatically generates a comprehensive infrastructure plan based on the Azure Well-Architected pillars, detailing the proposed architecture, resource selections, and the rationale behind each choice. You have the opportunity to review and refine the plan before proceeding, ensuring it aligns with your specific requirements and best practices.
Instantly generate code
After you approve the plan, a single click generates modular Terraform templates tailored to the target architecture. You can open them in VS Code for Web for immediate review and customization or initiate a draft pull request to your GitHub repository to enable collaborative workflows and streamline CI/CD deployment.
Choose your deployment path

- GitHub Pull Request: Review the generated code in your repository, merge when ready, and trigger deployment through your existing CI/CD pipeline.
- VS Code for Web: Instantly edit and deploy directly from your browser—ideal for rapid iterations and immediate feedback.

Deploy with confidence
Every plan and template is aligned with Azure Well-Architected Framework best practices, so you can move from idea to deployment in minutes, not hours.

Getting Started with Prompts

Before trying prompts, make sure you activate Copilot Agent mode.

Here are some example prompts you can try:

"Deploy a GPT-4 based language model for summarization tasks, exposing a REST API secured via Azure AD and autoscale compute resources as demand grows."

"Launch a multilingual chatbot service using Azure OpenAI Service, integrate logging with Azure Monitor, and use Azure Key Vault for API credential management."

"Set up a scalable GenAI document search API using embeddings, index content in Azure Cognitive Search, and cache recent queries with Azure Redis Cache."

"Deploy a Python Flask app with a PostgreSQL backend, secure secrets in Key Vault, and enable monitoring with Application Insights."