Fundamentals of Cloud Computing and OpenAI with Microsoft Azure

This course introduces modern cloud computing, cloud architecture, and artificial intelligence (AI)-enabled software development, with a focus on designing and building secure, scalable, cloud-native applications using Microsoft Azure. The concepts and architectural patterns learned apply broadly across cloud providers and technology stacks. Students learn when and how to apply traditional, cloud-native, and serverless architectures; design and implement representational state transfer application programming interfaces (RESTful APIs); select between relational and NoSQL data stores; and evolve monolithic systems into macroservice and microservice architectures. The course emphasizes defense-in-depth security practices, including network segmentation with virtual networks (VNETs) and secure service-to-service communication. AI-assisted software development is incorporated throughout the course, with students using tools such as GitHub Copilot to write, test, review, and refactor production-quality code while validating AI-generated outputs and applying responsible AI practices. The course examines how model context protocol (MCP) servers enhance context-aware AI-assisted development to improve agent efficiency and accuracy. Pragmatic strategies for applying multi-agent workflows to enhance developer productivity are also covered. Students are introduced to models available through Azure and Microsoft Foundry, including OpenAI, Anthropic, open-source, and other foundation models. The course covers the fundamentals of AI development for software-as-a-service (SaaS) applications, beginning with a conceptual understanding of the generative pre-trained transformer (GPT) architecture, then progressing to system and user prompts, few-shot design and refinement, fine-tuning, and retrieval-augmented generation (RAG), with emphasis on trade-offs involving response quality, domain alignment, latency, and cost. AI safety is a core theme of the course, including identifying and mitigating prompt injection attacks and designing multi-layered architectures for AI guardrails using heuristic analysis, risk scoring, and Azure AI Content Safety capabilities, including prompt shields. Students gain hands-on experience designing and deploying secure, scalable, geo-redundant, and cost-effective infrastructure using infrastructure as code with the Bicep language, and work with a range of Azure platform as a service (PaaS) offerings including API Management, Azure SQL, Cosmos DB, Azure Storage, Azure AI Search, Azure App Services, Azure Container Apps, and Azure Service Bus. Students learn how to enable SaaS applications to utilize authentication and authorization using OAuth and Microsoft Entra ID and apply production-readiness and software development and information technology operations (DevOps) practices such as continuous information/continuous delivery (CI/CD) pipelines, monitoring and alerting, rollback strategies, and always-up deployment patterns. By the end of the course, students are prepared to design, build, and operate cloud-native and AI-enabled SaaS applications suitable for real-world enterprise environments.