Study guide
9.2 Azure Vision, Azure AI Search, and Microsoft Foundry
Microsoft Certified: AI Transformation Leader Exam AB-731
Track your progress as you read
Finished sections get a check mark.
1 Introduction & Exam Overview
0.1 Current AB-731 Exam Facts 0.2 Audience Profile and the July 22, 2026 Blueprint 0.3 Independent Study Path for AB-731
2 Generative AI Foundations
1.1 Generative AI Versus Other Types of AI 1.2 Select a Generative AI Solution for a Business Need 1.3 Pretrained Versus Fine-Tuned Models
3 Cost, Risks, and Business Value
2.1 Tokens and ROI Cost Drivers 2.2 Fabrications, Reliability, and Bias 2.3 Scalability, Automation, and When Generative AI Adds Value
4 Prompt Engineering
3.1 Impact of Prompt Engineering 3.2 Prompt Engineering Techniques
5 Grounding, RAG, and Data
4.1 Business Requirements for Grounding Solutions 4.2 Retrieval-Augmented Generation (RAG) 4.3 Data Type, Quality, and Representative Datasets
6 Secure AI and Machine Learning
5.1 Importance of Secure AI 5.2 Application, Data, and Authentication Security for AI 5.3 When Machine Learning Adds Value 5.4 Lifecycle of a Machine Learning Solution
7 Microsoft 365 Copilot and Microsoft Copilot
6.1 Map Business Processes and Use Cases to Copilot 6.2 Differences in Capabilities Between Versions of Copilot 6.3 Microsoft 365 Copilot Chat on Web and Mobile 6.4 Copilot in Microsoft 365 Apps
8 Copilot Studio, Microsoft Graph, and Integrated AI
7.1 Microsoft Copilot Studio Capabilities 7.2 Microsoft Graph Capabilities 7.3 Integrated Microsoft AI, Risk Mitigation, and Safety
9 Agents, Extensibility, and Stack Mapping
8.1 Map Processes to Microsoft AI Apps and Services 8.2 When to Use Researcher or Analyst in Copilot 8.3 Build, Buy, or Extend Copilot
10 Microsoft Foundry and Foundry Tools
9.1 Map Business Processes to Foundry Tools 9.2 Azure Vision, Azure AI Search, and Microsoft Foundry 9.3 Match an AI Model to a Business Need 9.4 Scalability and Security Benefits of Foundry
11 Responsible AI and Governance
10.1 Importance of Responsible AI 10.2 Governance Principles for AI Use 10.3 Establish an AI Council 10.4 Microsoft Responsible AI Standards
12 Adoption Teams, Barriers, and Champions
11.1 Establish an Adoption Team 11.2 Common Barriers to AI Adoption 11.3 Establish an AI Champions Program
13 Impacts, Copilot Licensing, and Foundry Cost Models
12.1 Impacts to Data, Security, Privacy, and Cost 12.2 Copilot License Types 12.3 Foundry Tools Subscription Models
9.2 Azure Vision, Azure AI Search, and Microsoft Foundry
Key Takeaways Explain This Topic AI
Azure Vision in Foundry Tools processes images and returns visual features such as objects, faces, optical character recognition, and captions for inspection and digital asset management.
Image Analysis 4.0 in Azure Vision is deprecated and scheduled to retire on September 25, 2028; Microsoft documents migration paths including Document Intelligence for optical character recognition, Face, Foundry models, and Content Understanding.
Azure AI Search is a managed retrieval service for classic indexes and agentic knowledge bases; it underpins Foundry IQ for permission-aware grounding of agents.
Microsoft Foundry unifies agents, models, and tools under one Azure resource, with a catalog of more than 10,000 models and Foundry Agent Service for prompt or hosted agents.
Foundry Tools is Microsoft's current name for the prebuilt API family formerly labeled Azure AI Services or Cognitive Services; teach the July 22, 2026 blueprint names.
Last updated: September 2026
9.2 Azure Vision, Azure AI Search, and Microsoft Foundry
Quick Answer: Azure Vision in Foundry Tools is for images: inspect, tag, caption, read text, and manage digital assets. Azure AI Search is for retrieval: classic indexes and knowledge bases that ground apps and agents in your content. Microsoft Foundry is the platform: catalog models, host agents, and govern tools under one Azure resource. Teach those July 22, 2026 names even when a Learn page still says Computer Vision or Azure AI Foundry.
Microsoft's AB-731 skills list asks you to identify capabilities of Foundry Tools, including Azure Vision in Foundry Tools, Azure AI Search, and Microsoft Foundry. That is a leader identification job. You do not configure REST clients on this exam. You must still open the current Learn pages, because product names and retirement notes move.
Foundry Tools (Microsoft's current umbrella for the prebuilt API family) include Speech, Translator, Language, Content Understanding, Document Intelligence, Vision, Azure AI Search, Content Safety, Custom Vision, and Immersive Reader. Microsoft lists several retired services you should not start new work on, including Anomaly Detector, Content Moderator, Language Understanding (LUIS), Metrics Advisor, Personalizer, and QnA Maker. This section spends its depth on the three names the blueprint calls out, then shows how the rest of the toolbox supports them.
Azure Vision in Foundry Tools
Microsoft's Vision overview states that Azure Vision gives you algorithms that process images and return information based on the visual features you care about. For a leader, that sentence is the whole product: send an image (or an image URL), get structured facts back, attach those facts to a business process.
Capability What the leader uses it for Not the same as
Image analysis Objects, tags, captions, adult-content flags, and related visual features Copilot describing one screenshot in a chat
Optical character recognition (OCR) Printed or handwritten text on photos, posters, whiteboards, and scanned pages Opening a PDF in Word and asking Copilot to summarize
Face Detect, recognize, or blur faces for identification, access, or privacy Storing badge photos in a shared mailbox
Digital asset management (DAM) Group media by logos, objects, colors, and keywords so a library is searchable A SharePoint folder with no AI tagging
Microsoft currently documents image constraints you should not invent around: JPEG, PNG, GIF, or BMP; file size less than 4 megabytes; dimensions greater than 50 × 50 pixels (Read API up to 10,000 × 10,000). Those limits belong in an operations conversation about camera resolution and batch size.
Retirement you must not hide. Microsoft cautions that Image Analysis 4.0 in Azure Vision is deprecated and will retire on September 25, 2028 , after which calls fail. Existing customers are supported until that date. Microsoft's migration guidance points leaders to alternatives that match the scenario: Document Intelligence for OCR and Read-style extraction, Face for face scenarios, Azure Content Understanding in Foundry Tools, and generative vision models in Microsoft Foundry . On AB-731, still identify Vision as the named inspection and DAM capability. In a real program, ask engineering which Vision API version is in production and what the 2028 plan is.
Vision plus Search is a common composition. Microsoft's DAM note is explicit: tag images with Vision, then make those tags and captions searchable with Azure AI Search. That is how a retailer finds every photo that shows a competitor logo, or how facilities finds every whiteboard photo that mentions a shutdown date.
Azure AI Search
Microsoft's introduction page describes Azure AI Search as a fully managed, cloud-hosted service that connects your data to AI. It unifies access to enterprise and web content so agents and large language models can use context and produce grounded answers. It is available through the Azure portal, REST APIs, and Azure software development kits ( SDKs ). It also underpins Foundry IQ , the managed knowledge layer that turns enterprise content into reusable, permission-aware knowledge bases for agents in the Microsoft Foundry portal.
Leaders need two retrieval modes, not a coding lab.
- Aspect Classic search Agentic retrieval
- Corpus A search index you define Knowledge sources behind a knowledge base
Query One request, ranked documents, no planning loop Optional large language model planning, subqueries, merge
Best for Predictable, low-latency lookup in a known schema Agents that must roam multiple sources and return citations
Status (current Learn) Generally available Generally available, with some capabilities still in preview
What you get when you create a search service, according to Microsoft: two engines (classic and agentic); full-text, vector, hybrid, and multimodal queries; AI enrichment to chunk and vectorize raw content; relevance tuning; Azure scale, security, monitoring, and compliance; integrations with supported data platforms, Azure OpenAI, and Microsoft Foundry.
Security is part of the capability, not an afterthought. Microsoft lists Microsoft Entra ID , Azure Private Link , document-level access control, and role-based access. That is why legal maps a matter corpus here instead of dumping PDFs into a consumer chatbot. Remote SharePoint knowledge sources are documented for user-permission inheritance; other sources may need security filters. Do not promise that every connector inherits every Access Control List ( ACL ) until engineering confirms the source type.
Pricing models exist (Dedicated and a Serverless preview). Full Foundry Tools subscription mechanics are chapter 13. Here, know Search is a billed Azure service you size by capacity and query pattern, not a Copilot license checkbox.
Microsoft Foundry
Microsoft Foundry unifies agents, models, and tools under a single management grouping, with tracing, monitoring, evaluations, and enterprise setup. Microsoft documents streamlined role-based access control (RBAC) , networking, and policies under one Azure resource provider namespace.
Capability What the leader uses it for
Agents Declarative prompt agents in the portal or SDK, or hosted agents that run your team's code in a managed endpoint
Models More than 10,000 models from Microsoft, OpenAI, Anthropic, Meta, and others in the Foundry Models catalog
Tools and knowledge Built-in tools, memory, and retrieval (Foundry Toolbox), including connections to Foundry Tools such as Search and Vision
Observability Trace and evaluate agents and models; monitor with built-in dashboards (some capabilities in preview)
Governance Entra identity, RBAC, content filters, network isolation, Azure Policy, and Foundry Control Plane for fleet visibility
Microsoft's evolution table is the naming decoder you need on exam day:
- Brand: Azure AI Studio / Azure AI Foundry → Microsoft Foundry
- Brand: Azure AI Services → Foundry Tools
Portal: Foundry (classic) hub-based projects still exist; new investment is on Foundry projects in the new portal
Resource model: hub plus Azure OpenAI plus Azure AI Services → a Foundry resource with projects
Coming from Azure OpenAI, Microsoft documents an upgrade path to a Foundry resource that preserves endpoint, API keys, and existing state. That is an IT conversation, not a reason to keep saying Azure AI Foundry on AB-731.
How the three work as one program
A claims organization is the clean composition. Vision (or Document Intelligence) extracts what is on the page or photo. Search indexes the extracted text and the original files. Foundry hosts the model and the agent that answers an adjuster, with Content Safety on the prompt and the completion. Copilot may still draft the customer letter in Word after a human accepts the grounded answer. If you reverse that order—start with Copilot Chat and hope it can see the claims archive—you mapped the wrong family.
Official Microsoft pages to reopen
What is Azure Vision in Foundry Tools?
- Azure Vision documentation
- Migrate from Azure Vision Image Analysis — September 25, 2028 retirement note
- Introduction to Azure AI Search
- Azure AI Search documentation
What is Microsoft Foundry?
What are Foundry Tools?
Loading diagram... Azure Vision, Azure AI Search, and Microsoft Foundry in one claims flow
Illustrative share of leader attention across the three named AB-731 capabilities (teaching split, not Microsoft weights)