Developing AI Apps and Agents on Azure (AI-103)

Duration: 4 days

This hands-on course prepares developers for AI-103 by covering how to design, implement, secure, monitor, and optimize AI applications and agentic solutions on Azure.

Who Should Attend

• Software developers building AI-enabled cloud applications
• Azure engineers implementing generative AI and agent workflows
• Solution architects designing enterprise AI platforms
• Teams preparing for the AI-103 certification exam

Prerequisites

• Solid programming experience with C# or Python
• Basic Azure experience (resource groups, identity, networking)
• Familiarity with REST APIs, JSON, and authentication concepts
• Basic understanding of large language models and AI fundamentals

Learning Goals

• Choose the appropriate Azure AI Foundry services and models for specific business scenarios
• Build generative AI and agentic solutions with grounding and tool integration
• Implement responsible AI practices, safety controls, and governance for production systems
• Develop computer vision solutions for image analysis and custom model use cases
• Build text analysis and conversational language solutions using Azure AI services
• Extract structured data from documents with Azure AI Document Intelligence
• Monitor, secure, and optimize AI workloads for cost, scale, and reliability

Course Content

1. Plan and Manage Azure AI Solutions
   • Azure AI Foundry architecture and project setup
   • Model selection for LLM, SLM, multimodal, and agent scenarios
   • Deployment options, quotas, scaling, and cost planning
   • Security baseline: managed identity, private networking, RBAC
2. Implement Generative AI and Agentic Solutions
   • Prompt orchestration and retrieval-augmented generation (RAG)
   • Vector stores, indexing strategies, and grounding quality
   • Agent workflows with memory, tools, and knowledge integration
   • Evaluation, tracing, and safety controls for agent behavior
3. Implement Computer Vision Solutions
   • Image analysis and OCR capabilities
   • Content moderation and visual reasoning scenarios
   • Integrating vision features into web and API applications
4. Implement Text Analysis Solutions
   • Named entity recognition, sentiment, and key phrase extraction
   • Conversational language understanding and classification
   • Translation and language detection for multilingual workflows
5. Implement Information Extraction Solutions
   • Document Intelligence models for forms, receipts, and invoices
   • Custom extraction models and validation workflows
   • Post-processing extracted data for downstream systems

Hands-on Labs

• Provision an Azure AI Foundry project and deploy model endpoints
• Build a RAG-enabled assistant using Azure AI Search and grounded prompts
• Implement an agent workflow with tool calling and approval checkpoints
• Add image analysis and OCR to a sample application
• Build a text analysis pipeline for sentiment and entity extraction
• Extract structured data from real-world documents and validate outputs
• Configure monitoring dashboards, safety evaluation, and cost controls

Outcomes

After this course, participants can deliver secure and scalable Azure AI solutions, implement agentic patterns end-to-end, and confidently prepare for the AI-103 exam domains.

Development Environment

• Windows, macOS, or Linux
• Visual Studio 2022/2026 or Visual Studio Code
• .NET 10 SDK (or Python 3.10+ for selected labs)
• Azure CLI and access to an Azure subscription
• Azure AI Foundry project with required service permissions