- Choose an appropriate model for each task, including LLMs, small language models, multimodal models, and Foundry Tools.
- Choose Foundry services for generative tasks, grounding, vector search, agent workflows, or multimodal processing.
- Choose an appropriate method for retrieval and indexing.
- Choose memory, tool, and knowledge integration services for agent solutions.
Microsoft Azure AI Apps and Agents Developer Associate (AI-103)
Microsoft Azure AI Apps and Agents Developer Certification Training Course AI-103
Prepare for Exam AI-103: Developing AI Apps and Agents on Azure, the associate credential that replaces AI-102. Accumentum maps the course to Microsoft’s published skills measured as of 16 April 2026.
Course focus: Planning Foundry solutions, generative AI and agents, RAG, computer vision, text analysis, and information extraction. It does not follow the retired AI-102 Cognitive Services outline.
Course Overview
Build, manage, and deploy agents and AI solutions on Microsoft Foundry.
Accumentum’s Microsoft Azure AI Apps and Agents Developer Certification Training Course (AI-103) prepares Azure AI engineers for Exam AI-103: Developing AI Apps and Agents on Azure, which leads to Microsoft Certified: Azure AI Apps and Agents Developer Associate. Exam AI-103 replaces AI-102.
Microsoft’s audience is an Azure AI engineer who builds, manages, and deploys agents and AI solutions that take advantage of Microsoft Foundry. Candidates should have experience developing apps by using Python and be familiar with general AI, generative AI, and Azure services.
This page uses Microsoft’s published skills measured as of 16 April 2026. Accumentum does not invent exam prices or unofficial question counts.
Course Objectives
What you will be able to plan, build, and operate.
01
Plan and manage Foundry solutions
Choose models and Foundry services, design infrastructure and CI/CD, and manage quotas, security, and responsible AI.
02
Generative AI and agents
Build RAG applications, tool-augmented flows, and orchestrated multi-agent solutions with the Foundry SDKs.
03
Vision, text, and speech
Implement image and video generation, multimodal understanding, language analysis, and speech as an agent modality.
04
Information extraction
Ingest and index content, configure semantic, hybrid, and vector search, and extract structured content with Content Understanding.
Who Should Attend
- Azure AI engineers who build, manage, and deploy agents and AI solutions with Microsoft Foundry.
- Developers with Python experience who are moving from prebuilt Azure AI services (the AI-102 era) to Foundry apps and agents.
- People who collaborate with stakeholders, architects, data scientists, DevOps, and cloud security engineers on AI solutions.
- Learners still on AI-102-only Cognitive Services material should use this AI-103 outline instead of treating AI-102 as current.
Prerequisites
- Experience developing apps by using Python.
- Familiarity with general AI, generative AI, and Azure services.
- Comfort with Foundry concepts: models, RAG, agents, and responsible AI instrumentation.
- No Accumentum-invented hour counts. Confirm any Microsoft-stated prerequisites on Microsoft Learn.
Pathway Map
Ten modules mapped to Microsoft AI-103 skills as of 16 April 2026.
Skills measured: plan and manage an Azure AI solution 25 to 30%, implement generative AI and agentic solutions 30 to 35%, computer vision 10 to 15%, text analysis 10 to 15%, and information extraction 10 to 15%.
1
Choose Foundry services for generative AI and agents
2
Set up AI solutions in Foundry
3
Manage, monitor, and secure AI systems
4
Responsible AI across generative and agentic systems
5
Build generative applications with Foundry
6
Build agents with Foundry
7
Optimize and operationalize generative AI systems
8
Implement computer vision solutions
9
Implement text analysis and speech solutions
10
Implement information extraction solutions
Course Content
- Design Azure infrastructure for AI apps and agent-based solutions.
- Choose appropriate deployment options.
- Configure model and agent deployments.
- Integrate Foundry projects with CI/CD pipelines.
- Manage quotas, scaling, rate limits, and cost footprints for model and agent workloads.
- Monitor model performance, drift, safety events, and grounding quality.
- Monitor data ingestion quality, search index health, and relevance performance.
- Configure security, including managed identity, private networking, keyless credentials, and role policies.
- Configure safety filters, guardrails, risk detection, and content moderation.
- Apply responsible AI instrumentation, including evaluators, safety evaluations, and explanation tooling.
- Implement auditing through trace logging, provenance metadata, and approval workflows.
- Govern agent behavior with oversight modes, constraints, and tool-access controls.
- Deploy and consume LLMs, small models, code models, and multimodal models.
- Implement retrieval-augmented generation (RAG) in an application.
- Design workflows, tool-augmented flows, and multistep reasoning pipelines.
- Evaluate models and apps, including fabrications, relevance, quality, and safety, using Foundry SDKs and connectors.
- Define agent roles, goals, conversation tracking, and tool schemas.
- Build agents that integrate retrieval, function-calling, and conversation memory.
- Integrate tools including APIs, knowledge stores, search, content understanding, and custom functions.
- Implement orchestrated multi-agent solutions and autonomous or semiautonomous workflows with safeguards and approval flows.
- Tune generation behavior with prompt engineering and model parameters.
- Implement model reflection, chain-of-thought evaluations, and self-critique loops.
- Set up observability with tracing, token analytics, safety signals, and latency breakdowns.
- Orchestrate multiple models, flows, or hybrid LLM and rules engines.
- Generate images and videos from text prompts and reference media, including inpainting and prompt-driven edits.
- Build multimodal understanding: captions, visual question-answering, alt-text, and Azure Content Understanding.
- Implement video analysis and object or region identification.
- Responsible AI for multimodal content: unsafe visual filters, indirect prompt injection in images, and visual policy rules.
- Extract entities, topics, summaries, and structured JSON using generative prompting and Foundry Tools.
- Detect sentiment, tone, safety issues, and sensitive content, then translate with Azure Translator or LLM-powered flows.
- Speech-to-text and text-to-speech for agentic interactions, including custom speech models.
- Enable multimodal reasoning from audio inputs.
- Ingest and index documents, images, audio, and video.
- Configure semantic, hybrid, and vector search for grounding.
- Implement enrichment skills and RAG ingestion including OCR.
- Extract content with multimodal pipelines and Content Understanding analyzers for structured or markdown outputs.
Course Features
Designed for AI-103 Foundry engineer work.
Skills-measured map
Ten modules aligned to Microsoft AI-103 skills measured as of 16 April 2026.
Agent-first associate
Generative AI and agentic solutions are the largest published domain (30 to 35%), not the prior AI-102 service catalog.
Public exam facts
Pearson VUE. Microsoft publishes a passing score of 700 or greater. Annual renewal on Microsoft Learn. No invented prices.
Prior exam context
AI-102 is the prior Azure AI Engineer Associate exam. This course is AI-103.
Certification Exam
Prepare for the official Microsoft AI-103 exam.
AI-103 Exam Readiness
Accumentum AI-103 training prepares Azure AI engineers for Developing AI Apps and Agents on Azure.
The Microsoft Certified: Azure AI Apps and Agents Developer Associate exam is AI-103. Public facts from Microsoft: the exam is delivered through Pearson VUE, you need a score of 700 or greater to pass, and associate certifications expire annually and can be renewed on Microsoft Learn. Accumentum does not publish unofficial question counts or invented exam prices.
Skills measured as of 16 April 2026: 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%).
Exam AI-103 replaces AI-102. Confirm current scheduling, languages, and interactive-item details on Microsoft Learn and Pearson VUE.
Skills Measured
What Microsoft says AI-103 measures.
Microsoft lists the skills measured for Exam AI-103 as of April 16, 2026. Here are all 5 domains, 14 skill groups and 64 tasks in Microsoft's own wording. Our course maps a lesson or lab to each one.
Choose the appropriate Foundry services for generative AI and agents
Part of Plan and manage an Azure AI solution (25 to 30%)
- Choose an appropriate model for each task, including large language models (LLMs), small language models, multimodal models, and Foundry Tools
- Choose the appropriate Foundry services for generative tasks, grounding, vector search, agent workflows, or multimodal processing
- Choose an appropriate method for retrieval and indexing
- Choose appropriate memory, tool, and knowledge integration services for agent solutions
Set up AI solutions in Foundry
Part of Plan and manage an Azure AI solution (25 to 30%)
- Design Azure infrastructure for AI apps and agent-based solutions
- Choose appropriate deployment options
- Configure model and agent deployments
- Integrate Foundry projects with continuous integration and continuous deployment (CI/CD) pipelines
Manage, monitor, and secure AI systems
Part of Plan and manage an Azure AI solution (25 to 30%)
- Manage quotas, scaling, rate limits, and cost footprints for model and agent workloads
- Monitor model performance, drift, safety events, and grounding quality
- Monitor data ingestion quality, search index health, and relevance performance
- Configure security, including managed identity, private networking, keyless credentials, and role policies
Implement responsible AI across generative AI and agentic systems
Part of Plan and manage an Azure AI solution (25 to 30%)
- Configure safety filters, guardrails, risk detection, and content moderation
- Apply responsible AI instrumentation, including evaluators, safety evaluations, and explanation tooling
- Implement auditing through trace logging, provenance metadata, and approval workflows
- Govern agent behavior with oversight modes, constraints, and tool-access controls
Build generative applications by using Foundry
Part of Implement generative AI and agentic solutions (30 to 35%)
- Deploy and consume LLMs, small models, code models, and multimodal models
- Implement retrieval-augmented generation (RAG) in an application
- Design workflows, tool-augmented flows, and multistep reasoning pipelines
- Evaluate models and apps, including detecting fabrications, relevance, quality, and safety
- Integrate generative workflows into applications by using Foundry SDKs and connectors
- Configure an application to connect to a Foundry project
Build agents by using Foundry
Part of Implement generative AI and agentic solutions (30 to 35%)
- Define agent roles, goals, conversation-tracking approach, and tool schemas
- Build agents that integrate retrieval, function-calling, and conversation memory
- Integrate agent tools, including APIs, knowledge stores, search, content understanding, and custom functions
- Implement orchestrated multi-agent solutions
- Build autonomous or semiautonomous workflows with safeguards and approval flow controls
- Integrate monitoring into deployed agents, evaluate agent behavior, and perform error analysis
Optimize and operationalize generative AI systems
Part of Implement generative AI and agentic solutions (30 to 35%)
- Tune generation behavior, such as prompt engineering and adjusting model parameters
- Implement model reflection, chain-of-thought evaluations, and self-critique loops
- Set up observability by implementing tracing, token analytics, safety signals, and latency breakdowns
- Orchestrate multiple models, flows, or hybrid LLM and rules engines
Design and implement image- and video-generation solutions
Part of Implement computer vision solutions (10 to 15%)
- Implement a solution that generates images from text prompts and reference media
- Implement a solution that generates videos from text prompts and reference media
- Configure image-editing workflows, including inpainting, mask‑based edits, and prompt‑driven modifications
- Implement workflows to edit generated videos
- Select and apply appropriate generation and editing controls provided by the platform
Design and implement multimodal understanding workflows
Part of Implement computer vision solutions (10 to 15%)
- Build a solution that analyzes visual context by using multimodal models
- Configure apps to produce concise or detailed captions for single or multiple images
- Implement a solution that enables question‑answering grounded in visual evidence
- Configure generation of alt‑text and extended image descriptions aligned to accessibility guidelines
- Implement visual understanding by configuring Azure Content Understanding in Foundry Tools to extract visual characteristics
- Implement video analysis workflows to process and interpret video segments
- Configure single‑task and pro‑mode Content Understanding pipelines
- Implement solutions that identify objects, components, or regions within images or video
Implement responsible AI for multimodal content
Part of Implement computer vision solutions (10 to 15%)
- Implement filters to classify unsafe or disallowed visual content
- Detect and mitigate indirect prompt injection by using embedded text in images
- Enforce visual policy rules, such as applying watermarks, flagging prohibited symbols, upholding brand usage requirements, and detecting potentially inappropriate content
Apply language model text analysis
Part of Implement text analysis solutions (10 to 15%)
- Implement solutions to extract entities, topics, summaries, and structured JSON outputs by using generative prompting and Foundry Tools
- Configure detection of sentiment, tone, safety issues, and sensitive content
- Build solutions that translate text by using Azure Translator in Foundry Tools or LLM‑powered translation flows
- Customize language model outputs for domain tasks, such as compliance summarization and domain extraction
Implement speech solutions
Part of Implement text analysis solutions (10 to 15%)
- Implement workflows to convert speech to text and text to speech for agentic interactions
- Integrate speech as an agent modality, including custom speech models
- Enable multimodal reasoning from audio inputs
- Translate speech into other languages by using language models and Foundry Tools
Build retrieval and grounding pipelines
Part of Implement information extraction solutions (10 to 15%)
- Ingest and index content, such as documents, images, audio, and video
- Configure semantic search, hybrid search, and vector search for grounding
- Implement enrichment by using custom or built-in skills for text, images, and layout
- Configure RAG ingestion flow, including documents and using optical character recognition (OCR)
- Connect retrieval pipelines directly to workflows and agent tools
Extract content from documents
Part of Implement information extraction solutions (10 to 15%)
- Extract information by using multimodal pipelines that combine OCR, layout analysis, and field extraction
- Produce clean, grounded representations to use with agents and RAG by using Content Understanding
- Implement analyzers for generating structured or markdown outputs for downstream reasoning by using Content Understanding
Study Plan
How we suggest you work through AI-103.
We start with planning and managing an Azure AI solution. Picking models, deploying them in Microsoft Foundry, and setting up security, monitoring and responsible AI controls come first, because every later lab depends on them.
Generative AI and agents carry the most weight at 30 to 35 percent. We spend the longest block building chat apps, retrieval augmented generation, prompt flows and agents that call tools, then evaluating and tuning what we built.
Next come computer vision, text analysis and information extraction. Each is 10 to 15 percent, so we cover them in focused labs that show where a prebuilt service saves you time over a custom model.
The final week is review. We walk the full skills outline again, rebuild your weakest labs and use the exam sandbox so test day holds no surprises.
Exam Policies
Scoring, renewal and test day for AI-103.
AI-103 leads to Microsoft Certified: Azure AI Apps and Agents Developer Associate. Microsoft requires a score of 700 or greater to pass.
Microsoft notes that the bullets under each skill show how a skill is assessed, and related topics may also be covered. Most questions cover features that are generally available, but the exam can include Preview features if they are commonly used.
Before test day you can try the exam sandbox to see the exam interface. If you use assistive devices or need extra time, Microsoft lets you request an accommodation. If the exam is not offered in your preferred language, you can request an extra 30 minutes.
Microsoft associate certifications expire annually. You renew by passing a free online assessment on Microsoft Learn, and we show you how to keep AI-103 current.
Career Outlook
Where AI-103 can take you.
AI-103 fits developers who ship AI features, not just prototypes. Titles that line up with it include AI engineer, Azure AI developer, machine learning engineer, applied AI developer and solutions engineer. The work is planning and securing Azure AI resources, building generative AI apps and agents on Microsoft Foundry, and adding vision, language and document extraction where a project needs them.
Per the U.S. Bureau of Labor Statistics, software developers, quality assurance analysts and testers earned a median annual wage of $134,040 in May 2025. That figure is a national midpoint. Where you live, how long you have been in the field and who you work for all move the number. Source: BLS Occupational Outlook Handbook.
Funding and PathPay
Ways to pay for AI-103 training.
Accumentum is an Army CA and Air Force COOL Approved Vendor. If you're serving, start with your education center and read our Army CA and Air Force COOL pages for the steps. We'll help you line up the course with your request.
Paying on your own? AI-103 prep is eligible for PathPay, our monthly payment plan. Terms of 3, 4 and 6 months carry no added fee. 9, 12 and 18 month terms add 5%, 10% and 15%. Course access opens when your first payment clears.
Our funding page has every other option, and you can always talk with us or email info@accumentum.net before you enroll.
Frequently Asked Questions
AI-103 Azure AI Apps and Agents Developer training with Accumentum: scope, audience, skills measured, and public exam facts.
No. This is Exam AI-103: Developing AI Apps and Agents on Azure, for Microsoft Certified: Azure AI Apps and Agents Developer Associate. AI-102 is the prior exam.
Azure AI engineers who build, manage, and deploy agents and AI solutions with Microsoft Foundry, with Python development experience.
Plan and secure Foundry solutions, build RAG apps and multi-agent systems, and implement vision, text, speech, and information extraction on the published AI-103 outline.
Ten modules mapped to Microsoft skills measured as of 16 April 2026, from service choice and responsible AI through agents, vision, text, and extraction.
As of 16 April 2026: plan and manage 25 to 30%, generative AI and agents 30 to 35%, vision 10 to 15%, text analysis 10 to 15%, and information extraction 10 to 15%.
Pearson VUE. Microsoft publishes a passing score of 700 or greater. Annual renewal on Microsoft Learn. No invented prices or unofficial question counts.
Use Accumentum registration for AI-103 training dates and enrollment support.
Microsoft requires a score of 700 or greater to pass.
Microsoft lists these 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%
- Implement information extraction solutions, 10 to 15%
Implement generative AI and agentic solutions is the largest at 30 to 35%. We give it the most lab time.
Microsoft lists 64 tasks across 14 skill groups as of April 16, 2026. Our course covers all of them.
Microsoft associate certifications expire annually. You renew by passing a free online assessment on Microsoft Learn, and we show you how to keep AI-103 current.
Yes. Microsoft offers an exam sandbox that shows you the exam interface and question types.
Accumentum is an Army CA and Air Force COOL Approved Vendor. Start with your education center, and we'll help you with the rest.
Microsoft says most questions cover generally available features, but Preview features can appear if they are commonly used.
Yes. AI-103 prep is eligible for PathPay. Terms of 3, 4 and 6 months add no fee, and longer terms add 5% to 15%.
Enrollment
Enroll in Microsoft Azure AI Apps and Agents Developer Certification Training (AI-103) with Accumentum.
Enroll in Accumentum’s AI-103 training to prepare for Exam AI-103. For course dates and enrollment support, visit Accumentum’s registration page linked below.
Prepare for AI-103 Foundry apps and agents work.
Associate-level Azure AI engineering on Microsoft Foundry. Not AI-102.