examOS.
Exam CatalogueStudy PlansRoadmapsBlogs
Login

ExamOS

Credits PolicyReferral PolicyQuality StandardsPricingPrivacy PolicyTerms of UseContact UsReport a Bug

Follow us

Disclaimer: ExamOS is an independent platform, not affiliated with any certification provider, and does not use or distribute exam dumps.

Share your feedback

Checking sign-in status...

examOS.Study Plan
Disclaimer: ExamOS is an independent platform, not affiliated with any certification provider, and does not use or distribute exam dumps.
← Back to Exam Details

Study Plan

Microsoft Certified Azure AI Apps and Agents Developer Associate (AI-103) – Study Plan

A condensed 6-week plan for AI-103. Build, manage, and deploy generative AI and agentic solutions using Microsoft Foundry, Azure AI services, and responsible AI practices.

MicrosoftAI-103Passing score: 700 / 1000Azure AI engineers with Python development experience and familiarity with Azure AI services, building generative AI and agentic applications on Microsoft Foundry18-Jul-20262 views
Start date: _______________Target exam date: _______________
6 WeeksDuration
~45 hrsTotal Study Time
5 DomainsExam Coverage

Stay consistent by setting a target date for this certification.

Set target

How to use this plan

  1. 1Write code. This is a developer exam. Every week includes Python coding activities using the Foundry SDK. Reading documentation without writing code will not prepare you for implementation questions.
  2. 2Build your judgment. For every service and pattern you encounter, ask: when is this the right choice, and what would change my answer if the scenario requirements were different?
  3. 3Practice with ExamOS. Use the quiz modes in the order specified in this Study Plan.
  4. 4Think in terms of the full solution. AI-103 questions frequently describe an end-to-end scenario: a model deployment, a retrieval pipeline, an agent with tools, and monitoring. Train yourself to connect these components into a coherent architecture.
Rookie ModeChallenger ModeLegend Mode

Week-by-Week Breakdown


W1

Week 1

Self-Assessment and Foundry Overview

This week is about mapping your current knowledge against the five exam domains, setting up your development environment, and getting oriented with Microsoft Foundry. You are not expected to master anything yet. You are building a baseline.

Topics

  • AI-103 exam guide and domain mapping
  • Microsoft Foundry platform architecture
  • Azure AI services landscape (updated for Foundry)
  • Python SDK setup and environment configuration
  • Responsible AI principles on Azure

Activities

  • Read the official AI-103 exam guide and identify weak domains.
W2

Week 2

Plan and Manage an Azure AI Solution (25-30%)

This is the second-largest domain and the governance foundation for everything else. It covers how to choose the right Foundry services for a task, how to set up infrastructure, how to monitor and secure AI systems, and how to implement responsible AI across generative and agentic solutions. If you are weak in this domain, every other domain will be harder to answer correctly.

Topics

  • Model selection: LLMs, small language models, multimodal models
  • Foundry services for generative tasks, grounding, agents
  • Retrieval and indexing method selection
  • Agent memory, tool, and knowledge integration services
  • Azure infrastructure design for AI apps
  • Model and agent deployment configuration
W3

Week 3

Build Generative AI Applications (Domain 2, Part 1)

This is the first half of the largest domain. It covers how to deploy and consume foundation models, implement RAG, design tool-augmented workflows, and evaluate model outputs. This week focuses on the generative application layer: taking a model and turning it into a working application that produces reliable, grounded outputs.

Topics

  • Deploy and consume LLMs, small models, code models, multimodal models
  • RAG implementation: retrieval, augmentation, generation
  • Retrieval pipeline design: vector search, hybrid search, indexing
  • Tool-augmented flows and multistep reasoning pipelines
  • Model evaluation: fabrications, relevance, quality, safety
  • Foundry SDKs and connectors for application integration
W4

Week 4

Build Agents and Optimize Generative AI (Domain 2, Part 2)

This is the second half of the largest domain and the most technically complex week. It covers agent design, multi-agent orchestration, autonomous workflows, and the optimization techniques (prompt engineering, chain-of-thought, observability) that make generative AI systems production-ready. Agents are the defining feature of AI-103 compared to AI-102.

Topics

  • Agent roles, goals, conversation tracking, tool schemas
  • Agent retrieval, function-calling, conversation memory
  • Agent tool integration: APIs, knowledge stores, search, Content Understanding
  • Multi-agent orchestration patterns
  • Autonomous and semiautonomous workflows with safeguards
  • Agent monitoring, behavior evaluation, and error analysis
W5

Week 5

Computer Vision, Text Analysis, and Information Extraction (Domains 3-5)

Domains 3, 4, and 5 are each 10-15% and cover specialized AI capabilities. They are studied together because they share a common pattern: each domain uses Foundry Tools and Azure AI services to process a specific modality (images/video, text/speech, documents). Candidates with prior Azure AI experience will recognize many of these services from AI-102, but the integration with Foundry and generative workflows is new.

Topics

  • Image and video generation from text prompts
  • Multimodal understanding: captions, visual QA, object detection
  • Content Understanding for visual and document analysis
  • Responsible AI for multimodal content: unsafe content filtering, prompt injection defense
  • Language model text analysis: entities, sentiment, summarization, structured output
  • Speech solutions: STT, TTS, speech as agent modality, custom speech models
W6

Week 6

Exam Simulation and Booking

Your final push. This week is full exam simulation mode: filling in remaining gaps and building the confidence to book the exam.

Topics

  • Full syllabus review across all five domains
  • Time management (120 minutes for 40-60 questions)
  • Cross-domain scenario reasoning
  • Weak area targeted review

Activities

  • Take the Microsoft exam sandbox to familiarize yourself with the exam interface.
  • Take at least 3 ExamOS Legend mode full quizzes (80% hard questions, 120-minute timer).

Daily Study Routine

Suggested 2–3 Hour Day

TimeActivity
15 minLook over yesterday's wrong answers. For each one, identify the concept you missed.
30 minRead Microsoft Learn documentation for this week's topic
30 minHands-on coding (Foundry SDK, model deployment, agent building)
30 minTake an ExamOS quiz (Challenger or Legend mode, depending on the week)
15 minLog missed concepts and review them the next morning

Stay consistent by setting a target date for this certification.

Set target
  • Set up an Azure account with Foundry access and explore the Foundry portal.
  • Install the Azure AI Foundry SDK for Python and verify your environment.
  • Deploy a sample model in Foundry and invoke it using the Python SDK.
  • Review Microsoft's Responsible AI principles and the six pillars.
  • Take the ExamOS Rookie mode quiz (30 questions). Note any domain below 60%.
  • Goal:Know where your gaps are across all five domains and have a working Foundry development environment.
    Rookie Mode
    Rookie Mode
  • CI/CD integration for Foundry projects
  • Quotas, scaling, rate limits, and cost management
  • Monitoring: model performance, drift, safety, grounding quality
  • Security: managed identity, private networking, keyless credentials, RBAC
  • Safety filters, guardrails, content moderation
  • Responsible AI instrumentation: evaluators, safety evaluations
  • Auditing: trace logging, provenance metadata, approval workflows
  • Agent governance: oversight modes, constraints, tool-access controls
  • Activities

    • Design Foundry infrastructure for a described AI application scenario, including networking and security.
    • Configure a model deployment with appropriate scaling, quota settings, and managed identity.
    • Set up safety filters and content moderation for a generative AI application.
    • Implement trace logging and configure an evaluator for a sample model.
    • Design an agent governance strategy with tool-access controls and approval workflows.
    • Take the ExamOS Challenger quiz. Repeat until you pass 2 times in a row.
    Goal:Design and secure Foundry-based AI solutions with appropriate monitoring and governance.
    Challenger Mode
    Challenger Mode
  • Foundry project configuration and connection
  • Activities

    • Deploy a model via Foundry and consume it using the Python SDK with streaming responses.
    • Implement a basic RAG application: ingest documents, generate embeddings, query with vector search.
    • Build a tool-augmented flow that calls an external API based on model output.
    • Evaluate model outputs using Foundry evaluators for relevance and groundedness.
    • Take the ExamOS Challenger quiz. Repeat until you pass 2 times in a row.
    Goal:Build and evaluate generative AI applications with RAG and tool integration.
    Challenger Mode
    Challenger Mode
  • Prompt engineering and model parameter tuning
  • Model reflection, chain-of-thought, self-critique loops
  • Observability: tracing, token analytics, safety signals, latency
  • Multi-model orchestration and hybrid LLM/rules engines
  • Activities

    • Build a Foundry agent with function-calling, conversation memory, and a defined tool schema.
    • Implement a multi-agent solution where an orchestrator agent delegates to specialist agents.
    • Configure monitoring for agent behavior and implement error analysis for failed tool calls.
    • Implement a chain-of-thought evaluation loop for a generative workflow.
    • Take the ExamOS Challenger quiz. Repeat until you pass 2 times in a row.
    Goal:Build and monitor production agents with appropriate safeguards and optimization.
    Challenger Mode
    Challenger Mode
  • Translation using Azure Translator and LLM-powered flows
  • Retrieval and grounding pipelines: ingestion, indexing, semantic/hybrid/vector search
  • Document extraction: multimodal pipelines, OCR, layout analysis, field extraction
  • Content Understanding for structured and markdown outputs
  • Activities

    • Implement image generation from text prompts using Foundry and configure generation controls.
    • Build a multimodal understanding solution that answers questions about uploaded images.
    • Implement a speech-to-text workflow integrated with an agent for voice-driven interactions.
    • Build a document extraction pipeline using Content Understanding with OCR and layout analysis.
    • Implement a retrieval pipeline with hybrid search (semantic plus keyword) for RAG grounding.
    • Configure responsible AI filters to classify unsafe visual content and detect prompt injection in images.
    • Take the ExamOS Challenger quiz. Repeat until you pass 2 times in a row.
    Goal:Implement vision, text, speech, and document extraction solutions integrated with Foundry.
    Challenger Mode
    Challenger Mode
  • Simulate real exam conditions at least once: no phone, no breaks, 120 minutes, same time of day you plan to actually sit the exam.
  • For every Legend question you miss, review the explanation and write one sentence identifying the Foundry concept or Azure service behind the correct answer.
  • Once you are consistently hitting 80% or above on Legend mode across two or more sessions, book your exam. You are ready.
  • Goal:Consistent above 80% on Legend mode across two or more sessions.
    Legend Mode
    Legend Mode

    Overview

    Exam AI-103 validates your ability to build, manage, and deploy generative AI and agentic solutions on Microsoft Azure using Microsoft Foundry. It replaces the previous AI-102 exam with a fundamentally different scope: where AI-102 focused on traditional Azure AI services (Computer Vision, Language, Speech), AI-103 is centered on generative AI, agents, RAG pipelines, and the Microsoft Foundry platform. AI-102 officially retired on June 30, 2026, so AI-103 is now the only current path for this credential area.

    This is a developer exam. You are expected to write Python code, use Foundry SDKs, and build applications that integrate models, agents, retrieval pipelines, and specialized AI tools. The exam tests whether you can design and implement production-grade AI solutions, not just understand what the services do.

    Domain Weight
    Plan and manage an Azure AI solution 25-30%
    Implement generative AI and agentic solutions 30-35%
    Implement computer vision solutions 10-15%
    Implement text analysis solutions 10-15%
    Implement information extraction solutions 10-15%

    Domain 2 (generative AI and agents) is the largest at 30-35% and covers three major areas: building generative applications, building agents, and optimizing/operationalizing generative AI systems. This is the core of the exam and the area where most candidates need the most preparation.

    A note on Microsoft Foundry: Foundry is the central platform for this exam. It is the unified environment for deploying models, building agents, managing projects, and integrating AI capabilities into applications. You will encounter "Foundry projects," "Foundry SDKs," "Foundry Tools," and "Foundry connectors" throughout every domain. If you are coming from AI-102, think of Foundry as the layer that unifies what were previously separate Azure AI services into a cohesive development platform.

    Recommended experience: Python development experience. Familiarity with Azure AI services and the Azure portal. Understanding of generative AI concepts (LLMs, prompt engineering, RAG). No formal prerequisites, but hands-on experience building AI applications is strongly recommended.

    A note on responsible AI: Responsible AI is not a separate domain in AI-103. It is woven into Domains 1, 2, and 3. Safety filters, guardrails, content moderation, auditing, and agent governance appear throughout the exam. Do not treat responsible AI as an afterthought. It is a thread that runs through every scenario.

    Frequently Asked Questions

    What is Microsoft Foundry?

    Microsoft Foundry is the unified platform for building AI applications on Azure. It brings together model deployment, agent development, retrieval pipelines, monitoring, and responsible AI tools into a single environment. Think of it as the layer that unifies what were previously separate Azure AI services (Azure OpenAI, Azure AI Search, Azure AI Vision, etc.) into a cohesive development platform with shared projects, SDKs, and governance.

    How is AI-103 different from AI-102?

    AI-102 (Azure AI Engineer Associate) focused on traditional Azure AI services: Computer Vision, Language, Speech, Document Intelligence, and Azure OpenAI as separate services. AI-103 is fundamentally different:

    • Foundry-centric: Everything is accessed through Microsoft Foundry, not separate service endpoints.
    • Generative AI and agents: 30-35% of the exam covers generative applications, RAG, agents, and multi-agent orchestration. AI-102 had minimal generative AI content.
    • Agent-focused: Building, monitoring, and governing agents is a core skill. AI-102 had no agent content.
    • Image/video generation: AI-103 tests image and video generation from prompts, not just analysis.
    • Content Understanding: A new Foundry Tool tested across vision and document domains.
    • Responsible AI for generative systems: Safety filters, guardrails, and agent governance are integrated throughout, not limited to traditional content moderation.

    AI-102 officially retired on June 30, 2026. If you are coming from AI-102 experience, expect approximately 40-50% new content focused on Foundry, generative AI, and agents.

    Do I need Python experience?

    Yes. This is a developer exam. The audience profile explicitly states you should have "experience developing apps by using Python." You will use the Foundry SDK, implement RAG pipelines, build agents with function-calling, and integrate retrieval workflows, all in Python. If you are not comfortable writing Python, the implementation questions will be difficult.

    How many practice questions should I aim for?

    At least 600 unique questions across all domains. Focus on questions that explain why each answer is right and why the others are wrong. For AI-103, the explanations matter more than the questions because the exam tests integration patterns, not isolated service features.

    When should I book the exam?

    Once you are consistently hitting 80% or above on Legend mode across two or more sessions, and you feel confident implementing Foundry-based solutions end to end, you are ready. Trust your preparation.

    How long is the exam?

    120 minutes for 40-60 questions. That averages to roughly 2-3 minutes per question. Some questions may take longer if they involve reading code snippets or scenario descriptions. Budget your time for the full 120 minutes.

    What is the passing score?

    700 out of 1000. Microsoft uses scaled scoring across all certification exams. The raw number of correct answers needed varies slightly between exam versions. Do not try to calculate how many questions you can afford to miss.

    I am a non-native English speaker. Can I get extra time?

    Yes. If the exam is not available in your preferred language, you can request an additional 30 minutes to complete the exam. Request this accommodation through the exam scheduling process. If the exam is available in your preferred language, you can still take it in English with the extra time if English is not your primary language.

    Can I use reference materials during the exam?

    No. The AI-103 exam is closed-book. You cannot access documentation, websites, or any external resources during the test. All questions are designed to be answerable from your knowledge of Microsoft Foundry, Azure AI services, and generative AI patterns.

    Does the exam cover Preview features?

    Possibly. Microsoft states that "most questions cover features that are general availability (GA). The exam may contain questions on Preview features if those features are commonly used." Focus on GA features but be aware of widely adopted Preview capabilities.

    What is the most important domain to study?

    Domain 2 (Implement generative AI and agentic solutions, 30-35%) is the largest and most heavily weighted domain. But Domain 1 (Plan and manage, 25-30%) is the second largest and provides the governance and infrastructure foundation that Domain 2 questions assume. Together they represent 55-65% of the exam. If you are strong in both, you have a significant margin even if you lose marks on the specialized domains.

    What are the most common ways people fail?

    • Treating AI-103 like a harder AI-102. If you study for AI-103 by reviewing the same Azure AI services you used for AI-102, you will miss the Foundry, generative AI, and agent content that makes up 30-35% of the exam.
    • Not understanding RAG at implementation depth. The exam tests whether you can implement a RAG pipeline, not just explain what RAG is. You need to know about embedding models, vector search configuration, chunking strategies, and how retrieval connects to generation.
    • Ignoring agent governance. Agents are tested heavily, and so is their governance: oversight modes, tool-access controls, approval workflows, and monitoring. Building an agent without governance is not enough.
    • Confusing Foundry Tools with standalone Azure services. Azure Translator, Azure Content Understanding, and Azure AI Speech are now accessed as Foundry Tools. The exam tests this integration, not just the standalone service capabilities.
    • Skipping responsible AI for generative systems. Responsible AI is not just content moderation. It includes safety evaluations, trace logging, provenance metadata, and agent behavior governance. These appear in questions across Domains 1, 2, and 3.
    • Not practicing with Python. Reading about the Foundry SDK is not the same as using it. If you have not imported the SDK, configured a client, deployed a model, or built an agent in code, you will struggle with implementation questions.

    What are the three biggest preparation mistakes?

    1. Studying services in isolation. AI-103 tests how services work together in Foundry: a model deployment connected to a retrieval pipeline connected to an agent connected to monitoring. If you study each service separately, you will not be prepared for integration scenarios.
    2. Not building agents. Agents are the defining feature of AI-103. If you have not built at least one agent with function-calling, conversation memory, and tool integration, you are unprepared for 30-35% of the exam.
    3. Treating responsible AI as a separate topic. Responsible AI is integrated into every domain. Safety filters, guardrails, content moderation, and agent governance appear throughout the exam, not in a single section. Study responsible AI as a thread that runs through every solution you design.

    Is there a renewal option?

    Yes. Microsoft associate, expert, and specialty certifications expire annually. You can renew by passing a free online assessment on Microsoft Learn, typically 25-30 questions, available starting 6 months before your expiration date. No exam fee is required for renewal. If you don't pass on the first attempt, you can retake after 24 hours; a second and third retake each require a 14-day wait. Retakes are capped at 5 attempts per rolling 12-month period, and each attempt requires a new (free) voucher. Check the Microsoft Learn certification renewal page for current details.


    Quick-Reference: What Is on the Exam (Skills measured as of April 16, 2026)

    For the full, always-current list, visit the official AI-103 exam page.

    Domain Key Skills
    Plan and manage an Azure AI solution (25-30%) Model selection (LLMs, small language models, multimodal models, Foundry Tools), Foundry service selection for generative tasks, grounding, vector search, agent workflows, multimodal processing, retrieval and indexing methods, agent memory and knowledge integration, Azure infrastructure design, deployment options, CI/CD integration, quotas and cost management, model performance monitoring, drift detection, safety monitoring, managed identity, private networking, keyless credentials, RBAC, safety filters, guardrails, content moderation, responsible AI evaluators, trace logging, provenance metadata, approval workflows, agent governance (oversight modes, constraints, tool-access controls)
    Implement generative AI and agentic solutions (30-35%) Deploy and consume LLMs, small models, code models, multimodal models, RAG implementation, retrieval pipeline design, tool-augmented flows, multistep reasoning, model evaluation (fabrications, relevance, quality, safety), Foundry SDKs and connectors, Foundry project configuration, agent roles and goals, conversation tracking, tool schemas, function-calling, agent memory, agent tool integration (APIs, knowledge stores, search, Content Understanding), multi-agent orchestration, autonomous workflows with safeguards, agent monitoring and error analysis, prompt engineering, model parameter tuning, model reflection, chain-of-thought, self-critique loops, observability (tracing, token analytics, safety signals, latency), multi-model orchestration, hybrid LLM and rules engines
    Implement computer vision solutions (10-15%) Image generation from text prompts and reference media, video generation from text prompts, image editing workflows (inpainting, mask-based edits, prompt-driven modifications), video editing workflows, generation and editing controls, multimodal understanding (visual context analysis, image captions, visual QA), alt-text and accessibility descriptions, Content Understanding for visual characteristics, video analysis workflows, single-task and pro-mode Content Understanding pipelines, object and region identification, unsafe visual content classification, indirect prompt injection detection in images, visual policy enforcement (watermarks, brand usage, prohibited symbols)
    Implement text analysis solutions (10-15%) Entity extraction, topic extraction, summarization, structured JSON output using generative prompting and Foundry Tools, sentiment and tone detection, safety and sensitive content detection, text translation (Azure Translator, LLM-powered translation), domain-specific language model customization, speech-to-text and text-to-speech for agentic interactions, speech as agent modality, custom speech models, multimodal reasoning from audio inputs, speech translation
    Implement information extraction solutions (10-15%) Content ingestion and indexing (documents, images, audio, video), semantic search, hybrid search, vector search for grounding, enrichment using custom or built-in skills, RAG ingestion flow with OCR, retrieval pipeline integration with workflows and agent tools, multimodal document extraction (OCR, layout analysis, field extraction), Content Understanding for clean grounded representations, analyzers for structured or markdown outputs for downstream reasoning