// Deploying Azure Functions and Static Sites with GitHub Actions | Cloud with Chris

Chris uses a real side-project — a multi-tenant card inventory app — to demonstrate GitHub Actions from first principles through to a working CI/CD pipeline.

GitHub Actions Fundamentals

The session opens with a clear explanation of what GitHub Actions are and why they are called workflows rather than pipelines: Actions are not limited to CI/CD and can respond to any GitHub event, including issue comments, new contributors, and dependency alerts. Core concepts covered include:

• Continuous Integration (CI): automatically building and testing code on every push or pull request to catch merge conflicts and regressions early
• Continuous Delivery (CD): automatically deploying verified builds to target environments
• Workflow triggers: push vs pull_request events, filtered by branch (master) and path patterns

Deploying Azure Functions (.NET)

Chris walks through the workflow files for his card API microservices, showing:

• Using the Azure Functions GitHub Action with a publish profile stored as a GitHub Secret
• Separate workflow files for PR validation (deploy to catch breakage) vs. master push (deploy to production)
• Downloading the publish profile from the Azure portal and storing it securely in repository secrets

Deploying a Static Frontend

A second workflow demonstrates a multi-job pipeline for the static frontend:

• Job 1: npm install, npm run build with environment-variable configuration (API base URI, search keys), then upload artifact
• Job 2: download artifact, then deploy to Azure using the Azure CLI login action
• Running on Linux (ubuntu-latest) vs. Windows for the .NET jobs

Dependabot and the GitHub Roadmap

The session closes with a live demo of Dependabot automatically raising a pull request for an outdated dependency, triggering the PR workflow, and Chris merging it after watching the deployment succeed in real time. The GitHub public roadmap (github.com/github/roadmap) is shown as a resource for tracking upcoming features — including manual approval gates for production deployments.