Develop AI apps and agents on Azure (AI103)

Overview

AI-powered applications are rapidly evolving beyond traditional models into intelligent, agent-driven systems that can reason, plan, and act. This course equips developers and AI engineers with the skills to design, build, and deploy production-ready AI applications and agents using Microsoft Azure and Microsoft Foundry.

Learners will explore how to create generative AI applications, orchestrate intelligent agents, and integrate tools and knowledge sources into cohesive, agentic solutions. The course also introduces multimodal AI capabilities, enabling systems to process and reason across text, images, and other data formats. By combining modern AI patterns with scalable Azure services, this course prepares learners to deliver robust, real-world AI solutions.

Prerequisites

Before attending this course, learners should have:

• Experience programming in Python
• Familiarity with REST APIs and SDK-based development
• Foundational knowledge of building and deploying applications on Azure

Target Audience

This course is designed for professionals who are responsible for building and deploying AI-powered applications on Azure. Typical roles include:

• Software developers working with cloud-based applications
• AI engineers developing intelligent and generative AI systems
• Technical professionals implementing AI solutions within enterprise environments

Learners should be comfortable working with code and APIs, and interested in building modern AI-driven applications using Microsoft technologies.

Objectives

By the end of this course, learners will be able to:

• Develop generative AI applications using Microsoft Foundry on Azure
• Design and implement AI agents capable of task execution and orchestration
• Integrate external tools and knowledge sources into agent-based solutions
• Apply multimodal AI techniques to process diverse data types
• Build natural language processing solutions for conversational and text-based scenarios
• Extract, analyse, and reason over visual and complex content
• Design scalable, production-ready AI systems using Azure services

Outline

Module 1: introduction to AI applications and agents on Azure

This module introduces the evolution of AI applications, focusing on the transition from traditional models to generative and agent-based systems.

Topics include:

• Overview of AI application architectures
• Introduction to generative AI concepts
• Understanding agent-based systems and their role in modern AI
• Overview of Microsoft Foundry capabilities
• Azure services for AI development

Learning outcomes:

• Understand the key components of AI-powered applications
• Identify when to use generative AI versus agent-based approaches
• Describe the role of Microsoft Foundry in AI development

Module 2: developing generative AI applications

This module focuses on building applications powered by generative models using Azure and Microsoft Foundry.

Topics include:

• Working with foundation models
• Prompt engineering techniques
• Designing application workflows with generative AI
• Managing inputs, outputs, and context
• Evaluating and refining model responses

Learning outcomes:

• Build and test generative AI applications
• Apply prompt engineering to improve output quality
• Design workflows that incorporate generative models effectively

Module 3: building AI agents on Azure

This module explores how to design and implement AI agents that can perform tasks, make decisions, and interact with users and systems.

Topics include:

• Agent architecture and design patterns
• Task planning and execution
• Managing agent state and memory
• Event-driven and autonomous agent behaviours
• Debugging and monitoring agent performance

Learning outcomes:

• Design and implement AI agents for real-world scenarios
• Enable agents to plan and execute tasks
• Monitor and optimise agent behaviour

Module 4: integrating tools and knowledge into agentic solutions

This module focuses on enhancing AI agents by connecting them to external tools, APIs, and knowledge sources.

Topics include:

• Tool integration patterns for agents
• Connecting agents to APIs and external services
• Knowledge grounding and retrieval techniques
• Implementing retrieval-augmented generation (RAG)
• Managing data sources and context injection

Learning outcomes:

• Integrate tools and APIs into agent workflows
• Enable agents to access and use external knowledge
• Improve response accuracy through grounded AI techniques

Module 5: developing natural language AI solutions

This module covers techniques for building applications that understand and generate human language.

Topics include:

• Natural language processing fundamentals
• Conversational AI design
• Text analysis and classification
• Language generation and summarisation
• Building chat-based interfaces

Learning outcomes:

• Develop applications that process and generate natural language
• Design conversational experiences
• Apply NLP techniques to real-world use cases

Module 6: multimodal AI and complex content understanding

This module introduces multimodal AI, enabling applications to process and reason across different types of data.

Topics include:

• Working with image and text inputs
• Multimodal model capabilities
• Extracting insights from visual data
• Combining modalities in a single workflow
• Handling complex and unstructured content

Learning outcomes:

• Build applications that process both text and images
• Extract meaningful insights from visual data
• Design workflows that combine multiple data types

Module 7: building scalable AI solutions with Microsoft Foundry
This module focuses on deploying and scaling AI applications and agents in production environments.

Topics include:

• Application deployment strategies on Azure
• Scaling AI workloads
• Performance optimisation
• Monitoring and logging
• Security and responsible AI considerations

Learning outcomes:

• Deploy AI applications to production environments
• Optimise performance and scalability
• Implement monitoring and governance practices

Module 8: designing production-ready AI systems

This module brings together all course concepts to design robust, enterprise-ready AI solutions.

Topics include:

• End-to-end solution design
• Architectural best practices
• Managing lifecycle and updates
• Testing and validation strategies
• Real-world use case scenarios

Learning outcomes:

• Design complete AI systems from concept to deployment
• Apply best practices for reliability and maintainability
• Deliver scalable, production-ready AI solutions

Key benefits

• Official Microsoft-authored content aligned to current Azure AI capabilities
• Focus on modern generative AI and agent-based development patterns
• Hands-on approach to building real-world AI applications
• Coverage of multimodal AI and knowledge-integrated solutions
• Prepares learners to design and deploy production-ready AI systems
• Natural progression from AI-102 with updated, future-focused content