About the Workshop

This full-day, hands-on workshop is for those who want to go beyond basic chat applications and learn how to build, extend, orchestrate, and deploy enterprise-ready AI agents on Azure. Participants will work with Azure AI Foundry, Foundry Agent Service, Foundry IQ, and the Microsoft Agent Framework to create agents, integrate built-in and external tools such as File Search, Code Interpreter, Bing Grounding, custom functions, and MCP servers, and design multi-agent workflows using visual orchestration, code-first patterns, connected agents, and A2A. Across four hands-on labs, learners will build working agents, connect tools and knowledge sources, explore when to use Foundry Agent Service versus Microsoft Agent Framework, and apply production best practices for monitoring, security, deployment, governance, scaling, and cost management. By the end, they will have a practical understanding of how to choose the right Azure-based agent architecture and build real-world agent systems confidently.

Prerequisites

Azure subscription - Create a free account if needed
Python 3.10+ installed
VS Code with the Python extension
Azure CLI installed and signed in (az login)
Azure Developer CLI (azd) installed
Docker Desktop installed (for hosted agent lab)
Node.js 18+ installed (for MCP server lab)

Workshop Modules

What are AI Agents?
- From chatbots to autonomous agents — the evolution
- Key characteristics: reasoning, tool use, memory, planning
- Agent design patterns: ReAct, tool-augmented generation, multi-agent
The Azure AI Agent Platform
- Azure AI Foundry — portal, projects, and resource model
- Model catalog — GPT-4o, Llama, DeepSeek, Cohere, and more
- Two pillars for building agents: Foundry Agent Service & Microsoft Agent Framework
Key Protocols & Standards
- MCP (Model Context Protocol) — open standard for tool integration
- A2A (Agent-to-Agent) — standardized inter-agent communication
- How MCP and A2A fit into the Azure AI ecosystem

Set up the development environment
Create an Azure AI Foundry project
Build and run a prompt agent using the Python SDK
Interact with the agent through threads and conversations

Lecture: The Agent Tool Ecosystem

Built-in Tools Overview : Code Interpreter, File Search, Bing Grounding, Azure Functions, OpenAPI, MCP Servers, and Memory
Model Context Protocol (MCP) — Deep Dive
- What is MCP? Open standard for tool integration with LLMs
- MCP in Foundry Agent Service: connect remote MCP servers as tools
- MCP tool catalog: Azure DevOps, GitHub, and custom servers
- Authentication, approval workflows, and security best practices
- Exposing your own services as MCP servers via Azure Functions
Foundry IQ — Enterprise Knowledge Layer
- What is Foundry IQ? Managed knowledge layer for enterprise data
- Knowledge bases and knowledge sources (Azure, SharePoint, OneLake, Web)
- Agentic retrieval engine — multi-query, iterative search with reasoning
- Connecting Foundry IQ knowledge bases to agents
- Foundry IQ vs. File Search vs. Azure AI Search — when to use which
Function Calling — Define custom functions the agent can invoke
Tool selection best practices — When to use which tool

Lab Steps

Code Interpreter & File Search
- Attach Code Interpreter to your agent for data analysis
- Upload documents and query them with File Search
Connect an MCP Server
- Connect the GitHub MCP server to your agent
- Configure server_label, server_url, and authentication headers
- Ask the agent to query GitHub repositories
- Review MCP tool call approval and auditing
Foundry IQ Knowledge Base
- Create a Foundry IQ knowledge base in the portal
- Connect knowledge sources (sample documents)
- Attach the knowledge base to your agent
- Test agentic retrieval with citation-backed answers
Bing Grounding + Custom Functions
- Create an agent with Bing Grounding for real-time web search
- Define and attach a custom function

What is Microsoft Agent Framework?
Agent Framework Architecture: Agents, Workflows, Model Clients, Session, Context Providers, Middleware
Tools & MCP in Agent Framework
A2A Protocol in Agent Framework
Hosting Options: A2A Protocol, OpenAI-Compatible Endpoints, Azure Functions (Durable), AG-UI Protocol
Agent Framework vs. Foundry Agent Service — Choosing the Right Approach

Build an agent using the Microsoft Agent Framework SDK
Integrate MCP tools (local and remote)
Connect to a remote A2A agent
Compare the Agent Framework experience with Foundry Agent Service

Why Multi-Agent?
- Single agent limitations
- Specialization, delegation, and collaboration patterns
- When to use multi-agent vs. a single agent with many tools
Foundry Workflows — Visual Multi-Agent Orchestration
- What are workflows? Declarative, UI-based agent orchestration
- Workflow patterns: Sequential, Group Chat and Human in the Loop
- Creating workflows in the Foundry portal
- Workflow YAML editing in VS Code
- Workflow versioning, change logs, and visual monitoring
Agent Framework Workflows — Code-First Orchestration
- Graph-based workflow architecture
- Key features: type safety, conditional routing, parallel processing, checkpointing
- Orchestration patterns: sequential, concurrent, hand-off, magentic
- Human-in-the-loop with request/response patterns
- Comparing Foundry Workflows vs. Agent Framework Workflow
- Connected Agents in Foundry
- Architecture: main agent + connected agents
- How connected agents communicate and aggregate responses
- The ConnectedAgentTool API
A2A Protocol for Cross-System Agent Communication
- A2A protocol specification: agent cards, message-based communication, tasks
- A2A in Foundry Agent Service — connecting external agent endpoints
- A2A in Agent Framework — A2AAgent for cross-framework interop
- A2A vs. Connected Agents — choosing the right approach
Hosted Agents
- Containerized agents with Agent Framework or LangGraph
- Docker + Azure Developer CLI (azd) deployment
- When to choose hosted agents vs. prompt agents

Build a Foundry Workflow with multiple agents (visual orchestration)
Build a code-first workflow with Microsoft Agent Framework
Connect agents across systems using the A2A protocol

Observability & Monitoring
- End-to-end tracing with Application Insights
- Monitoring agent decisions, tool calls, and token usage
- Agent Framework telemetry and middleware-based logging
- Setting up alerts for failures and performance degradation
Security & Identity
- Microsoft Entra authentication and RBAC
- Content filters and safety guardrails
- MCP security best practices: allow-lists, approval workflows, auditing
- A2A authentication with AuthInterceptor
- Virtual network isolation and data protection
- Managed credentials and On-Behalf-Of (OBO) authentication
Deployment & Scaling
- Agent versioning and stable endpoints
- Hosted agents: Docker + azd deployment to Foundry
- Agent Framework hosting: Azure Functions (Durable), Container Apps
- Publishing to Microsoft Teams and Microsoft 365 Copilot
- Entra Agent Registry for enterprise discovery
Foundry IQ in Production
- Connecting multiple knowledge sources at scale
- Permission-aware retrieval and data governance
- Foundry IQ vs. Fabric IQ vs. Work IQ — choosing the right IQ layer
Cost Management
- Understanding token consumption and tool call costs
- MCP tool call billing and Bing transaction costs
- Optimizing agent instructions for efficiency
- Monitoring usage with Azure Cost Management
Choosing Your Architecture
What's Next?
- Explore the Foundry Model Catalog
- Build production workflows with Agent Framework
- Dive into Foundry IQ
- Build and register your own MCP servers
- Join the Azure AI community

Instructor

Manoranjan Rajguru

AI Architect

Workshop Details

brahmaid

It is needed for personalizing the website.

Expiry: Session

Type: HTTP

csrftoken

This cookie is used to prevent Cross-site request forgery (often abbreviated as CSRF) attacks of the website

Expiry: Session

Type: HTTPS

Identityid

Preserves the login/logout state of users across the whole site.

Expiry: Session

Type: HTTPS

sessionid

Preserves users' states across page requests.

Expiry: Session

Type: HTTPS

g_state

Google One-Tap login adds this g_state cookie to set the user status on how they interact with the One-Tap modal.

Expiry: 365 days

Type: HTTP

MUID

Used by Microsoft Clarity, to store and track visits across websites.

Expiry: 1 Year

Type: HTTP

_clck

Used by Microsoft Clarity, Persists the Clarity User ID and preferences, unique to that site, on the browser. This ensures that behavior in subsequent visits to the same site will be attributed to the same user ID.

Expiry: 1 year

Type: HTTP

_clsk

Used by Microsoft Clarity, Connects multiple page views by a user into a single Clarity session recording.

Expiry: 1 Day

Type: HTTP

SRM_I

Collects user data is specifically adapted to the user or device. The user can also be followed outside of the loaded website, creating a picture of the visitor's behavior.

Expiry: 2 years

Type: HTTP

SM

Use to measure the use of the website for internal analytics

Expiry: 1 years

Type: HTTP

CLID

The cookie is set by embedded Microsoft Clarity scripts. The purpose of this cookie is for heatmap and session recording.

Expiry: 1 year

Type: HTTP

SRM_B

Collected user data is specifically adapted to the user or device. The user can also be followed outside of the loaded website, creating a picture of the visitor's behavior.

Expiry: 2 months

Type: HTTP

_gid

This cookie is installed by Google Analytics. The cookie is used to store information of how visitors use a website and helps in creating an analytics report of how the website is doing. The data collected includes the number of visitors, the source where they have come from, and the pages visited in an anonymous form.

Expiry: 399 days

Type: HTTP

_ga_#

Used by Google Analytics, to store and count pageviews.

Expiry: 399 Days

Type: HTTP

_gat_#

Used by Google Analytics to collect data on the number of times a user has visited the website as well as dates for the first and most recent visit.

Expiry: 1 day

Type: HTTP

collect

Used to send data to Google Analytics about the visitor's device and behavior. Tracks the visitor across devices and marketing channels.

Expiry: Session

Type: PIXEL

AEC

cookies ensure that requests within a browsing session are made by the user, and not by other sites.

Expiry: 6 months

Type: HTTP

G_ENABLED_IDPS

use the cookie when customers want to make a referral from their gmail contacts; it helps auth the gmail account.

Expiry: 2 years

Type: HTTP

test_cookie

This cookie is set by DoubleClick (which is owned by Google) to determine if the website visitor's browser supports cookies.

Expiry: 1 year

Type: HTTP

_we_us

this is used to send push notification using webengage.

Expiry: 1 year

Type: HTTP

WebKlipperAuth

used by webenage to track auth of webenagage.

Expiry: session

Type: HTTP

ln_or

Linkedin sets this cookie to registers statistical data on users' behavior on the website for internal analytics.

Expiry: 1 day

Type: HTTP

JSESSIONID

Use to maintain an anonymous user session by the server.

Expiry: 1 year

Type: HTTP

li_rm

Used as part of the LinkedIn Remember Me feature and is set when a user clicks Remember Me on the device to make it easier for him or her to sign in to that device.

Expiry: 1 year

Type: HTTP

AnalyticsSyncHistory

Used to store information about the time a sync with the lms_analytics cookie took place for users in the Designated Countries.

Expiry: 6 months

Type: HTTP

lms_analytics

Used to store information about the time a sync with the AnalyticsSyncHistory cookie took place for users in the Designated Countries.

Expiry: 6 months

Type: HTTP

liap

Cookie used for Sign-in with Linkedin and/or to allow for the Linkedin follow feature.

Expiry: 6 months

Type: HTTP

visit

allow for the Linkedin follow feature.

Expiry: 1 year

Type: HTTP

li_at

often used to identify you, including your name, interests, and previous activity.

Expiry: 2 months

Type: HTTP

s_plt

Tracks the time that the previous page took to load

Expiry: Session

Type: HTTP

lang

Used to remember a user's language setting to ensure LinkedIn.com displays in the language selected by the user in their settings

Expiry: Session

Type: HTTP

s_tp

Tracks percent of page viewed

Expiry: Session

Type: HTTP

AMCV_14215E3D5995C57C0A495C55%40AdobeOrg

Indicates the start of a session for Adobe Experience Cloud

Expiry: Session

Type: HTTP

s_pltp

Provides page name value (URL) for use by Adobe Analytics

Expiry: Session

Type: HTTP

s_tslv

Used to retain and fetch time since last visit in Adobe Analytics

Expiry: 6 months

Type: HTTP

li_theme

Remembers a user's display preference/theme setting

Expiry: 6 months

Type: HTTP

li_theme_set

Remembers which users have updated their display / theme preferences

Expiry: 6 months

Type: HTTP

We do not use cookies of this type.

_gcl_au

Used by Google Adsense, to store and track conversions.

Expiry: 3 months

Type: HTTP