Building AI applications and agents in Azure has never been more accessible. The AI-103 training guides you through the Azure AI Foundry platform to develop, deploy and operate generative and conventional AI solutions in production environments.
In four days, you work on AI agent orchestration, GPT and Azure OpenAI model integration, Retrieval Augmented Generation (RAG) with Azure AI Search, and Azure cognitive AI services: language, vision, document intelligence, speech. The training is delivered in Geneva and Lausanne by a Microsoft Certified Trainer.
Module 1 : Plan and prepare to develop AI solutions on Azure
Module 2 : Select, deploy, and evaluate Microsoft Foundry models
Module 3 : Develop a generative AI chat app with Microsoft Foundry
Module 4 : Develop generative AI apps that use tools
Module 5 : Optimize generative AI model performance with Microsoft Foundry
Module 6 : Implement a responsible generative AI solution in Microsoft Foundry
Module 7 : Develop AI agents with Microsoft Foundry and Visual Studio Code
Module 8 : Integrate custom tools into your agent
Module 9 : Integrate MCP Tools with Azure AI Agents
Module 10 : Build knowledge-enhanced AI agents with Foundry IQ
Module 11 : Integrate your agent with Microsoft 365
Module 12 : Build agent-driven workflows using Microsoft Foundry
Module 13 : Develop an AI agent with Microsoft Agent Framework
Module 14 : Orchestrate a multi-agent solution using the Microsoft Agent Framework
Module 15 : Discover Azure AI Agents with A2A
Module 16 : Analyze text with Azure Language in Foundry Tools
Module 17 : Develop a text analysis agent with the Azure Language MCP server
Module 18 : Develop a speech-capable generative AI application
Module 19 : Create speech-enabled apps with Azure Speech in Microsoft Foundry Tools
Module 20 : Develop a speech agent with the Azure Speech MCP server
Module 21 : Develop an Azure Speech Voice Live Agent in Microsoft Foundry
Module 22 : Translate text and speech with Microsoft Foundry Tools
Module 23 : Develop a vision-enabled generative AI application
Module 24 : Generate images with AI
Module 25 : Generate videos with Microsoft Foundry
Module 26 : Analyze images with Content Understanding
Module 27 : Create a multimodal analysis solution with Azure Content Understanding
Module 28 : Create an Azure Content Understanding client application
Module 29 : Extract data with Azure Document Intelligence
Module 30 : Create a knowledge mining solution with Azure AI Search
This course prepares you to the AI-103: Developing AI Apps and Agents on Azure exam.
Azure AI Foundry, formerly Azure AI Studio, is the unified environment that brings together AI application development in the Microsoft ecosystem. The Develop AI apps and agents on Azure (AI-103) training covers project creation, model deployment, response evaluation and production rollout. You work with the essential tools: the model catalog to compare GPT, Phi, Mistral and open source models using benchmarks, the endpoints and SDKs used to consume those models from your code, and the playground to try out your scenarios interactively before industrialising them.
Agents sit at the centre of the program, accounting for close to a third of the skills measured on the exam. You learn to design agents able to call external tools (function calling), reason over multistep plans and query both structured and unstructured data sources. The course covers the Foundry Agent service, the built-in tools (code interpreter, web search, file search), agent development from Visual Studio Code, and the architecture patterns behind production agentic solutions: planner and executor, multi-agent collaboration, tool routing.
Retrieval Augmented Generation has become the standard pattern for grounding generative models on enterprise data. The AI-103 training has you build a full pipeline: document ingestion, chunking, embedding generation with Azure OpenAI, vector indexing in Azure AI Search, hybrid search combining vector similarity and BM25, result reranking, then final answer generation. Advanced techniques are covered as well: query rewriting, multi-stage retrieval, contextual compression. This is the foundation of document assistants and internal answer engines.
Beyond generative AI, the training covers the specialised services most real projects rely on: Azure AI Language for entity extraction, sentiment analysis, classification and question answering, Azure AI Vision for OCR, object detection and image analysis, Azure AI Document Intelligence for invoices, forms and structured documents, Azure AI Speech for speech recognition, synthesis and translation. The goal is to combine these building blocks with generative models to build multimodal applications.
The program covers the security and responsible AI requirements expected in the enterprise. You configure Azure AI Content Safety to filter harmful content, set up prompt and response monitoring, manage managed identities and private networking, and apply the Microsoft Responsible AI Standard principles. Prompt injection, jailbreak, data leakage and model bias are addressed along with their countermeasures, as is auditability through trace logging and provenance metadata.
The Develop AI apps and agents on Azure (AI-103) training targets developers, data engineers and solution architects who build AI applications in Azure. Python development experience is expected, since most of the examples and SDKs in the program rely on that language. Knowledge of Azure cloud fundamentals and AI concepts, at AZ-900 and AI-900 level, makes the four days significantly easier to follow.
The course prepares you for the AI-103: Developing AI Apps and Agents on Azure exam, which leads to the Microsoft Certified: Azure AI Apps and Agents Developer Associate certification. The exam measures five domains: plan and manage an Azure AI solution (25 to 30 %), implement generative AI and agentic solutions (30 to 35 %), implement computer vision solutions (10 to 15 %), implement text analysis solutions (10 to 15 %) and implement information extraction solutions (10 to 15 %). A score of 700 is required to pass.
What is the difference between AI-103 and AI-102?
AI-102 was the previous course and exam, Designing and Implementing a Microsoft Azure AI Solution, leading to the Azure AI Engineer Associate certification, which Microsoft has now retired. AI-103 replaces it with a program refocused on Microsoft Foundry, agents and generative AI, and a new associated certification.
Do I need to know Python to attend?
Yes. The exercises and SDKs used in the program are based on Python. Familiarity with REST API calls, JSON and notebooks is a further advantage.
Does the course cover Azure OpenAI?
Yes, OpenAI models are central to the course: deployment to endpoints, model selection using benchmarks, embeddings for vector search, prompt engineering, fine-tuning, content filters and quota management.
Which roles does this certification target?
Azure AI developer, AI engineer, AI solution architect, Azure OpenAI consultant and ML engineer in a Microsoft environment.
Monday to Friday
8:30 AM to 6:00 PM
Tel. 058 307 73 00
ITTA
Route des jeunes 35
1227 Carouge, Suisse
Monday to Friday, from 8:30 am to 06:00 pm.
