You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: learn-pr/wwl/design-ai-agents-business-solutions/includes/1-introduction.md
+1-1Lines changed: 1 addition & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -1,4 +1,4 @@
1
-
Artificial Intelligence (AI) continues to transform business operations, enabling organizations to deliver enhanced customer experiences, streamline workflows, and drive innovation. Within the Microsoft ecosystem, Dynamics 365 and Copilot offer robust platforms for integrating AI-powered agents and solutions across sales, service, and contact center environments.
1
+
Artificial Intelligence (AI) continues to transform business operations, enabling organizations to deliver enhanced customer experiences, streamline workflows, and drive innovation. Within the Microsoft ecosystem, Dynamics 365 and Copilot offer robust platforms for integrating AI-powered agents and solutions across sales, service, and contact center environments.
2
2
3
3
This module provides an overview of foundational concepts and practical approaches for designing AI and agents tailored to business needs. Learners will explore the principles of Responsible AI, ensuring that solutions are developed and deployed in ways that are safe, trustworthy, and aligned with organizational values. Special attention is given to Microsoft's Responsible AI guidelines, which establish a framework for fairness, reliability, privacy, inclusiveness, transparency, and accountability throughout the AI lifecycle.
By the end of this unit, learners will be able to:
4
-
- Identify the categories of Azure AI Foundry tools available for agent development and orchestration.
5
-
- Match business or technical requirements to the correct tool in the Foundry tool catalog.
6
-
- Recommend appropriate Foundry tools for building, grounding, extending, or operationalizing AI agents.
7
-
- Evaluate constraints such as data access, API type, compute needs, and integration patterns.
4
+
- Identify the categories of Azure AI Foundry tools available for agent development and orchestration
5
+
- Match business or technical requirements to the correct tool in the Foundry tool catalog
6
+
- Recommend appropriate Foundry tools for building, grounding, extending, or operationalizing AI agents
7
+
- Evaluate constraints such as data access, API type, compute needs, and integration patterns
8
8
9
-
## 1. Introduction
9
+
## Introduction
10
10
11
11
Azure AI Foundry provides a **catalog of tools** that agents can use to perform tasks such as retrieving data, calling APIs, grounding responses, orchestrating workflows, and triggering actions across applications.
12
12
13
13
When designing an AI agent, selecting the correct tool is essential. The goal is to choose tools that:
14
14
15
-
-**Meet the requirement with minimal complexity.**
15
+
-**Meet the requirement with minimal complexity**
16
16
17
-
-**Ensure security and compliance.**
17
+
-**Ensure security and compliance**
18
18
19
-
-**Leverage existing enterprise systems.**
19
+
-**Leverage existing enterprise systems**
20
20
21
-
-**Reduce integration overhead.**
21
+
-**Reduce integration overhead**
22
22
23
-
-**Support accurate, grounded outputs.**
23
+
-**Support accurate, grounded outputs**
24
24
25
-
## 2. Categories of Foundry tools
25
+
## Categories of Foundry tools
26
26
27
27
_Based on Azure AI Foundry tool catalog organization_.
28
28
29
29
Azure AI Foundry tools fall into several functional categories commonly used when designing AI agents.
30
30
31
-
### 2.1 Retrieval and grounding tools
31
+
### Retrieval and grounding tools
32
32
33
-
Used when the agent must **access enterprise knowledge** or **retrieve relevant documents**.Typical capabilities include:
33
+
Used when the agent must **access enterprise knowledge** or **retrieve relevant documents**.Typical capabilities include:
34
34
35
-
- Vector search.
35
+
- Vector search
36
36
37
-
- Hybrid (keyword + semantic) search.
37
+
- Hybrid (keyword + semantic) search
38
38
39
-
- Indexing structured or unstructured sources.
39
+
- Indexing structured or unstructured sources
40
40
41
-
- Querying enterprise knowledge bases.
41
+
- Querying enterprise knowledge bases
42
42
43
-
### 2.2 Data and application connectors
43
+
### Data and application connectors
44
44
45
45
Used when the agent must interact with business applications or databases:
46
46
47
-
- CRM systems.
47
+
- CRM systems
48
48
49
-
- ERP or financial systems.
49
+
- ERP or financial systems
50
50
51
-
- Line-of-business appsofbusiness apps.
51
+
- Line-of-business apps
52
52
53
-
- SQL databases or Cosmos DB.
53
+
- SQL databases or Cosmos DB
54
54
55
-
- REST or Graph API endpoints.
55
+
- REST or Graph API endpoints
56
56
57
-
### 2.3 Workflow and action tools
57
+
### Workflow and action tools
58
58
59
59
Used to trigger **automated business actions**, such as:
60
60
61
-
- Creating records.
61
+
- Creating records
62
62
63
-
- Updating cases.
63
+
- Updating cases
64
64
65
-
- Sending notifications.
65
+
- Sending notifications
66
66
67
-
- Triggering Power Automate flows.
67
+
- Triggering Power Automate flows
68
68
69
-
- Calling custom API operations.
69
+
- Calling custom API operations
70
70
71
-
### 2.4 Reasoning, planning, and execution tools
71
+
### Reasoning, planning, and execution tools
72
72
73
73
Used when the agent must:
74
74
75
-
- Evaluate conditions.
75
+
- Evaluate conditions
76
76
77
-
- Break tasks into steps.
77
+
- Break tasks into steps
78
78
79
-
- Select the right action.
79
+
- Select the right action
80
80
81
-
- Handle branching logic.
81
+
- Handle branching logic
82
82
83
-
### 2.5 Specialized tools
83
+
### Specialized tools
84
84
85
-
Used for purpose specific capabilities:specific capabilities:
85
+
Used for purpose-specific capabilities:
86
86
87
-
- Document summarization.
87
+
- Document summarization
88
88
89
-
- Classification.
89
+
- Classification
90
90
91
-
- Custom ML model execution.
91
+
- Custom ML model execution
92
92
93
-
- Safe completion and validation toolscompletion and validation tools.
93
+
- Safe completion and validation tools
94
94
95
-
## 3. Proposing tools for requirements
95
+
## Proposing tools for requirements
96
96
97
97
Below are examples of how to map business requirements to the correct Foundry tools.
98
98
99
-
### 3.1 Requirement type: Retrieve policies, guidelines, or knowledge
99
+
### Requirement type: Retrieve policies, guidelines, or knowledge
Generative capabilities in Power Apps now allow makers—especially code-first developers—to create model-driven app pages using natural language. Combined with an **agent feed**, organizations can deliver dynamic, personalized, and adaptive app experiences that respond in real-time to business data, user actions, and AI-generated insights. This unit introduces how generative pages work, how developers can architect "code-first" enhancements, and how agent feeds add intelligence to business applications.
4
4
5
5
## Learning objectives
6
6
7
7
By the end of this unit, learners will be able to:
8
-
- Describe the purpose of **generative pages** in model-driven apps.
9
-
- Propose when to use **code-first extensions** with generative pages.
10
-
- Explain the role and value of an **agent feed** in responsive app experiences.
8
+
9
+
- Describe the purpose of **generative pages** in model-driven apps
10
+
- Propose when to use **code-first extensions** with generative pages
11
+
- Explain the role and value of an **agent feed** in responsive app experiences
11
12
- Recommend appropriate use cases for generative pages and agent driven functionality
12
13
13
-
## 1. Understanding generative pages in model-driven apps
14
+
## Understanding generative pages in model-driven apps
14
15
15
16
Generative pages allow app makers to **describe a requirement in natural language**, and Power Apps automatically creates a page layout, data experiences, and UI structure. This accelerates solution development by removing repetitive scaffolding steps.
16
17
17
-
**How it works**
18
+
### How it works
18
19
19
20
Maker enters natural language prompt: *"Create a customer overview page showing recent orders, open cases, and a satisfaction score."*
20
21
21
-
- Power Apps analyzes available Dataverse data.
22
+
- Power Apps analyzes available Dataverse data
22
23
23
-
- A generative engine creates a page layout, bindings, and forms.
24
+
- A generative engine creates a page layout, bindings, and forms
24
25
25
-
- Maker optionally adjusts the page using codefirst or low-code features.
26
+
- Maker optionally adjusts the page using code-first or low-code features
26
27
27
-
- The system generates content aligned with enterprise-grade security and governance.
28
+
- The system generates content aligned with enterprise-grade security and governance
28
29
29
-
## 2. Code-first extensions for generative pages
30
+
## Code-first extensions for generative pages
30
31
31
-
Even though generative pages reduce design time, many enterprise apps require deeper customization. Codefirst developers enhance generated pages using:
32
+
Even though generative pages reduce design time, many enterprise apps require deeper customization. Code-first developers enhance generated pages using:
32
33
33
34
-**JavaScript event handlers**
34
35
@@ -38,66 +39,66 @@ Even though generative pages reduce design time, many enterprise apps require de
38
39
39
40
-**Reusable components and services**
40
41
41
-
-**Security aware data pipelines aware data pipelines**
42
+
-**Security-aware data pipelines**
42
43
43
-
-**Why use code-first with generative pages?**
44
+
### Why use code-first with generative pages?
44
45
45
-
|**Need**|**Code-first BenefitFirst Benefit**|
46
+
|**Need**|**Code-first benefit**|
46
47
|---|---|
47
48
| Complex business rules | Implement logic not expressible through prompts alone |
Use these links for more detail and hands on practice:
100
101
101
-
-**Introducing the new Power Apps generative experiences**<https://www.microsoft.com/power-platform/blog/power-apps/introducing-the-new-power-apps-generative-power-meets-enterprise-grade-trust/>.
102
+
-**Introducing the new Power Apps generative experiences**<https://www.microsoft.com/power-platform/blog/power-apps/introducing-the-new-power-apps-generative-power-meets-enterprise-grade-trust/>
0 commit comments