|
| 1 | +The Azure Copilot Deployment agent helps you design, configure, and deploy Azure infrastructure using best practices and automation. It streamlines everything from deciding which Azure services to use, to generating Infrastructure-as-Code templates, to integrating with CI/CD pipelines to deploy those resources. |
| 2 | + |
| 3 | +The Deployment agent interprets your requirements and creates a detailed deployment plan. The Deployment agent references the Azure Well-Architected Framework (WAF) and Azure architecture best practices to ensure the solution it devises is robust, secure, and cost-effective. The agent supports an interactive, multi-turn dialog. The agent may ask to clarify questions about your needs including "What kind of redundancy do you need?" or "Is this a production or dev environment?" and then incorporate your answers into the plan. |
| 4 | + |
| 5 | +Once you specify the requirements, the Deployment agent outputs a comprehensive infrastructure design. This includes recommendations on which Azure services and resources to use, how to configure them, and how to secure and monitor them. |
| 6 | + |
| 7 | +The Deployment agent can translate the finalized plan into Infrastructure-as-Code templates. Currently the agent focuses on creating Terraform configurations for these artifacts. Azure Copilot then generates the Terraform code needed to stand up all the specified Azure resources including virtual networks, VMs, or App Services, databases, Key Vaults, and monitoring setup, that are specified in the design. This code generation saves a huge amount of time for operators, as writing Infrastructure as Code manually for complex environments can be tedious and error-prone. |
| 8 | + |
| 9 | +Once the templates are generated, the agent facilitates review and deployment of the infrastructure. This includes the following functionality: |
| 10 | + |
| 11 | +- The agent can open the generated files in an embedded editor or Visual Studio Code for the Web, so you can inspect or tweak the Infrastructure as Code before applying it. |
| 12 | +- The agent can integrate with GitHub by automatically creating a pull request in your repository with the new Terraform files. |
| 13 | +- You can have Copilot execute the deployment directly. The agent provides step-by-step instructions to run the Terraform configuration. The agent can provide guidance on setting up continuous deployment for these resources, for example advising how to connect the Terraform configuration to an Azure DevOps pipeline for repeatable deployments. |
| 14 | + |
| 15 | +The Deployment agent uses Azure's services for validation. It cross-checks with services including the Azure Pricing Calculator to estimate costs of the proposed architecture. The agent may also use Azure Resource Graph or reference architectures from the Azure Architecture Center to validate that the design meets best practice guidelines. |
| 16 | + |
| 17 | +The Deployment agent is useful whenever you need to set up new infrastructure or modify existing deployments in Azure. |
| 18 | + |
| 19 | +Example scenarios for the Deployment agent include: |
| 20 | + |
| 21 | +- **Greenfield environment provisioning:** You can prompt Azure Copilot with the following: _"I want to deploy a highly available web application with a front-end, an API layer, a database, and monitoring."_ The Deployment agent designs a multi-tier architecture including Azure Front Door or Azure Application Gateway, an App Service for the web front-end, an AKS, or App Service for the API, Azure SQL or Cosmos DB for the database, and integrate Azure Monitor/App Insights for observability and configure the appropriate secure connectivity. The agent will generate the Terraform configuration to create all these resources. You can then review this plan and deploy it. |
| 22 | +- **Infrastructure-as-Code automation:** Organizations practicing Infrastructure-as-Code can use the deployment agent to codify their existing architecture in Terraform. Instead of writing the Terraform configuration from scratch, they can describe the setup to Copilot. The Azure Copilot Deployment agent can ensure things like tags, naming conventions, and resource group structure are consistent and compliant with relevant policy. The deployment agent can serve as an Infrastructure as Code templating tool that encapsulates Azure's evolving best practices. |
| 23 | +- **Complex multi-service deployments:** The Deployment agent allows you to orchestrate complex or multi-step deployments. For example, you can use the Deployment agent to help deploying a data analytics pipeline that includes Event Hubs, Databricks, Data Lake Storage, and Power BI integration. By describing this to the Azure Copilot Deployment agent in a prompt, the agent would lay out an architecture complete with networking and governance. It would output Infrastructure as Code and provide guidance on how to configure each piece. |
| 24 | +- **Integration with DevOps workflows:** The Deployment agent can integrate into enterprise change management systems. For example, if you want to add a new subnet or deploy additional instances, you can prompt the Deployment agent to adjust the infrastructure. The Azure Copilot Deployment agent provides the Terraform delta or a new configuration, which can then be added to a GitHub repository through a pull request. By using the existing workflow processes, agent-proposed changes can go through the compliance checks that standard changes go through. The advantage to Ops teams is that the heavy lifting of writing infrastructure code is handled by the agent. |
| 25 | + |
| 26 | +By automating the design and code generation, the Deployment agent reduces human errors and increases continuous deployment velocity. |
0 commit comments