Uploaded October 2026 | Updated November 2026, 7 hours ago
Adding more agents does not automatically create a better system. Multi-agent solutions need clear responsibilities, controlled execution, recoverable state, and carefully designed decision boundaries.
In this session, we will extend the agent built in Episode 1 into a coordinated incident-management workflow. Specialist agents will handle triage, knowledge retrieval, impact assessment, and remediation planning. We will use Microsoft Agent Framework to demonstrate sequential, concurrent, and handoff orchestration patterns, and discuss where each pattern is appropriate.
The workflow will introduce typed state, checkpointing, failure recovery, and human-in-the-loop approval before any sensitive remediation action can continue. We will also separate deterministic application logic from agentic reasoning, showing which decisions should be handled by an AI agent and which should remain enforced by conventional code.
Attendees will see how a multi-agent workflow behaves when interrupted, how it resumes from saved state, and how human approval can be incorporated without losing execution context.
By the end of the session, attendees will be able to design controlled multi-agent workflows that combine AI judgment with reliable business-process execution. Key topics: Multi-agent orchestration patterns Sequential, concurrent, and handoff workflows Typed workflow state Checkpointing and recovery Human-in-the-loop approvals Agentic versus deterministic steps Failure handling and idempotency.
0:00:00 Introduction and Housekeeping
0:00:54 Series Overview and Episode Introduction
0:02:38 Speaker Introduction
0:03:46 Project Atlas: Incident Management Architecture
0:07:04 Workflow Walkthrough: From Intake to Remediation
0:10:43 Live Demo: Incident Triaging and Approval Process
0:18:47 Recoverable Approval Workflow Design
0:20:34 Orchestration Patterns: Sequential, Concurrent, and Handoff
0:22:12 Typed Workflow State and Cosmos DB Integration
0:24:35 Live Demo: Handoff Orchestration and Decision Gates
0:30:26 Checkpointing, Persistence, and Failure Recovery
0:32:08 Deterministic Policy Gates: Code vs. AI Reasoning
0:35:32 Human-in-the-Loop: Approval Execution
0:39:18 Idempotency and Side-Effect Management
0:43:42 Observability and Traceability in Foundry
0:47:28 Recap of Infrastructure and Identity
0:50:22 Q&A and Closing Remarks
[eventID:27508]
Adding more agents does not automatically create a better system. Multi-agent solutions need clear responsibilities, controlled execution, recoverable state, and carefully designed decision boundaries.
In this session, we will extend the agent built in Episode 1 into a coordinated incident-management workflow. Specialist agents will handle triage, knowledge retrieval, impact assessment, and remediation planning. We will use Microsoft Agent Framework to demonstrate sequential, concurrent, and handoff orchestration patterns, and discuss where each pattern is appropriate.
The workflow will introduce typed state, checkpointing, failure recovery, and human-in-the-loop approval before any sensitive remediation action can continue. We will also separate deterministic application logic from agentic reasoning, showing which decisions should be handled by an AI agent and which should remain enforced by conventional code.
Attendees will see how a multi-agent workflow behaves when interrupted, how it resumes from saved state, and how human approval can be incorporated without losing execution context.
By the end of the session, attendees will be able to design controlled multi-agent workflows that combine AI judgment with reliable business-process execution. Key topics: Multi-agent orchestration patterns Sequential, concurrent, and handoff workflows Typed workflow state Checkpointing and recovery Human-in-the-loop approvals Agentic versus deterministic steps Failure handling and idempotency.
0:00:00 Introduction and Housekeeping
0:00:54 Series Overview and Episode Introduction
0:02:38 Speaker Introduction
0:03:46 Project Atlas: Incident Management Architecture
0:07:04 Workflow Walkthrough: From Intake to Remediation
0:10:43 Live Demo: Incident Triaging and Approval Process
0:18:47 Recoverable Approval Workflow Design
0:20:34 Orchestration Patterns: Sequential, Concurrent, and Handoff
0:22:12 Typed Workflow State and Cosmos DB Integration
0:24:35 Live Demo: Handoff Orchestration and Decision Gates
0:30:26 Checkpointing, Persistence, and Failure Recovery
0:32:08 Deterministic Policy Gates: Code vs. AI Reasoning
0:35:32 Human-in-the-Loop: Approval Execution
0:39:18 Idempotency and Side-Effect Management
0:43:42 Observability and Traceability in Foundry
0:47:28 Recap of Infrastructure and Identity
0:50:22 Q&A and Closing Remarks
[eventID:27508]
![Introducción a ALM para Power Platform con Azure DevOps
En esta sesión conoceremos los fundamentos de la gestión del ciclo de vida de aplicaciones en Microsoft Power Platform. Explicaremos cómo Azure DevOps permite implementar control de versiones, integración continua y despliegues automatizados entre entornos de desarrollo, pruebas y producción.
🔗 Aprende más accediendo a los recursos:
https://aka.ms/PowerPlatformParaAzureDevOps
https://aka.ms/PowerPlatformBuildToolsMKT
📌 Este evento es parte de una serie, aprende más aquí: https://aka.ms/PowerPlatformALM/y
0:05 Introducción y bienvenida
2:48 Introducción a los conceptos de ALM
7:38 Herramientas de ALM en Power Platform
8:13 ¿Qué es un Pipeline?
10:56 Diferencias: Pipelines nativos vs Azure DevOps
13:21 Entornos y soluciones (conceptos clave)
21:28 Requisitos previos y roles de seguridad
24:32 Integración: Power Platform y Azure DevOps
29:36 Creación de la aplicación en Azure (intermediario)
38:15 Demo: Configuración en Azure
40:19 Demo: Configuración en Power Platform
43:24 Demo: Configuración del proyecto en Azure DevOps
44:48 Configuración de Service Connections
47:45 Uso de Federated Credentials
54:20 Preguntas y cierre
#MicrosoftReactor #LearnConnectBuild #Devops, #PowerPlatform, #ALM
[eventID:27567] Introducción a ALM para Power Platform con Azure DevOps](https://i.ytimg.com/vi/wr7gihCMJSM/mqdefault.jpg)
![Durable Execution for .NET Workflows with Dapr, Aspire and Diagrid
Long-running processes fail partway through all the time, whether from a pod restart, an API timeout or a deploy that lands mid-execution.
Durable execution means your code survives those failures and resumes exactly where it stopped.
This session builds a durable workflow in .NET with Dapr Workflow, runs and debugs it locally with Aspire and the Dapr Dev Dashboard, then moves it to Diagrids managed durable execution engine with one change to the AppHost. The Diagrid workflow visualizer is used to inspect workflow executions and manage workflows.
[eventID:27765] Durable Execution for .NET Workflows with Dapr, Aspire and Diagrid](https://i.ytimg.com/vi/wz8eMXesARI/mqdefault.jpg)
![Model Mondays - Spotlight On Claude Models
Update: A single well-crafted prompt only gets an agent so far. This session shows how to take agents further with tool integrations available for Claude in Microsoft Foundry hosted on Azure. Well build workflows from scratch, covering the setup choices that matter most and how to get them right from the start.
- Explore the Resources https://github.com/microsoft/model-mondays
- Read the Newsletter https://aka.ms/model-mondays/newsletter
- Continue the conversation on the Discord https://aka.ms/model-mondays/discord
00:00 Introduction and Welcome
01:51 Weekly Microsoft Foundry News Highlights
08:33 Claude Models in Microsoft Foundry
15:48 Building with Claude: Model Selection, Hosting, and Scaling
21:31 Demo: Deploying Claude in Microsoft Foundry
25:55 New Agentic Capabilities and Tools
33:15 Demo: Building a Field Technician Agent
41:53 Agent Governance and Production Best Practices
49:26 Long-Running Agents and the Planner-Generator-Evaluator Pattern
54:27 Optimizing Token Costs and Closing Resources
[eventID:27542] Model Mondays - Spotlight On Claude Models](https://i.ytimg.com/vi/x5qDSukgTi4/mqdefault.jpg)
![Construyendo servidores y clientes MCP
Aprender cómo implementar servidores y clientes MCP, exponer herramientas y recursos, y conectarlos con aplicaciones de IA.
🔗 Aprende más accediendo a los recursos:
https://aka.ms/MCPrepo
https://aka.ms/IntroMCP
https://aka.ms/MCPtools
https://aka.ms/MCPagents
📌 Este evento es parte de una serie, aprende más aquí: https://aka.ms/mcpafondo
00:00 Introducción y objetivos de la sesión
02:40 Visión general de los SDKs de MCP
05:05 Construyendo servidores MCP con C#
06:25 Paquetes de NuGet esenciales
10:05 Atributos clave: Tools, Resources y Prompts
13:50 Importancia del atributo Description
17:10 Ejemplo práctico: Fabi Robotics
22:15 Inyección de dependencias en ASP.NET Core
33:05 Uso del inspector para validar servidores
36:05 Configuración del host en ASP.NET Core
41:20 Construcción de clientes MCP
42:05 Diferencias entre transportes locales y remotos
47:20 Consumo de herramientas, recursos y prompts
51:15 Integración con LLMs y Tool Calling
59:20 Demostración final y conclusión
#MicrosoftReactor #LearnConnectBuild #MCP #Agents
[eventID:27600] Construyendo servidores y clientes MCP](https://i.ytimg.com/vi/xT3VAGlHc4E/mqdefault.jpg)
![Fabric MythBusters: Is It Really Too Expensive and Complicated?
Fabric gets plenty of criticism around cost, complexity, and whether you really need the whole platform.
Instead of dodging those questions, we’re putting them on the table. Practitioners will test the claims, talk through the tradeoffs, and share where Fabric makes sense and where it may not.
[eventID:27759] Fabric MythBusters: Is It Really Too Expensive and Complicated?](https://i.ytimg.com/vi/xtjZ5Sm4eVc/mqdefault.jpg)
![Frontier Transformation: Reimagining Business with the New Microsoft 365 Copilot
Move beyond AI-assisted productivity to an AI-first enterprise, where the New Microsoft 365 Copilot with Copilot Cowork, Code, Autopilot and intelligent agents work alongside people to transform business processes, accelerate innovation, and unlock measurable business value, powered by organizational intelligence and built on a foundation of trust.
[eventID:27796] Frontier Transformation: Reimagining Business with the New Microsoft 365 Copilot](https://i.ytimg.com/vi/y4jloGqdrrQ/mqdefault.jpg)
![DP-700 in One Hour: What to Know, What to Study, Where to Focus
Get the big picture on DP-700 without trying to learn everything at once. Well break down the Fabric Data Engineer exam, highlight the skills and concepts to focus on, and point you toward the best ways to practice and prepare. Leave with a practical roadmap for getting exam-ready.
📌 This event is part of a series. Learn more here:
https://aka.ms/DataDays/FabConRecap/y
https://aka.ms/DataDays/IntoCerts/y
https://aka.ms/AIatWork/UnderstandIt/y
https://aka.ms/AIatWork/BuildIt/y
https://aka.ms/AIatWork/ShowIt/y
[eventID:27695] DP-700 in One Hour: What to Know, What to Study, Where to Focus](https://i.ytimg.com/vi/yE6Lw5Cg6FU/mqdefault.jpg)
![Fabric IQ na prática: de uma pergunta de negócio a uma resposta fundamentada nos seus dados
Chega de explicar. Vamos colocar o Fabric IQ para trabalhar com uma pergunta de negócio real.
Veremos como o contexto muda a resposta, o que significa uma resposta realmente fundamentada nos seus dados e onde o profissional de dados ainda precisa conferir o trabalho.
Para mais informações sobre o Data Days e tudo o que ele tem a oferecer, acesse https://aka.ms/DataDays
📌 Este evento faz parte de uma série. Saiba mais aqui: https://aka.ms/AIatWork/BuildIt/y
#MIcrosoftReactor #LearnConnectBuild #Fabric #DataDays #AI
[eventID:27732] Fabric IQ na prática: de uma pergunta de negócio a uma resposta fundamentada nos seus dados](https://i.ytimg.com/vi/yVVKrSLkSkw/mqdefault.jpg)
![AI とともにリリースする: 自信を持ってレビュー、セキュリティ確保、デプロイを行う
セッション概要
ソフトウェアを構築することは、開発プロセスの一部にすぎません。安全で信頼性の高いアプリケーションを大規模に提供するには、開発ライフサイクル全体を通じて品質とセキュリティを確保することが重要です。
このセッションでは、コードレビューやセキュリティ検証から CI/CD、自動マージまで、実際のチームのワークフローに AI を取り入れる方法を学びます。ライブデモを通じて、AI が開発の各段階で品質向上、セキュリティ強化、一貫性の維持にどのように役立つかを確認し、チームが AI 搭載およびエージェント駆動型システムを自信を持ってリリースできるようにする方法を紹介します。
学習内容
* AI がコードレビューと開発品質を向上させる方法
* AI をセキュリティおよびコンプライアンスのワークフローに組み込む方法
* 最新の CI/CD プラクティスを使用してデプロイプロセスを自動化する方法
* AI 搭載アプリケーションを大規模に自信を持ってリリースする方法
使用するテクノロジー
* GitHub Copilot Code Review
* Agent Merge
* GitHub Actions (CI/CD)
* AI 支援によるセキュリティレビュー
* エンドツーエンドの開発ライフサイクル自動化
対象者
* コード品質とデプロイを担当する開発者
* CI/CD ワークフローを管理する DevOps およびプラットフォームエンジニア
* AI 搭載アプリケーションを構築するエンジニアリングチーム
* AI を活用してソフトウェアを自信を持ってレビューし、セキュリティを確保し、リリースしたい方
0:00:00 - オープニングとイントロダクション
0:03:52 - ソフトウェア開発ライフサイクルとAIの活用
0:06:13 - AI導入における信頼とセキュリティの課題
0:08:14 - ソフトウェアサプライチェーンのリスク
0:10:00 - 依存関係アップデートのベストプラクティス
0:13:37 - デモシナリオの概要
0:18:07 - デモ:AIエージェントによる開発フローの開始
0:22:18 - GitHub Copilotを使ったIssue作成と実装
0:33:04 - AI支援によるコードレビューとAgent Merge
0:45:32 - Dependabotとセキュリティ機能の確認
0:49:05 - デプロイと最終確認
0:52:36 - セキュリティ対策のまとめと質疑応答
[eventID:27492] AI とともにリリースする: 自信を持ってレビュー、セキュリティ確保、デプロイを行う](https://i.ytimg.com/vi/yXzbxCaiCAs/mqdefault.jpg)
![Build Like a World Champion: The Decisions Behind Great Dataviz
Great dataviz is not just about making something beautiful. It’s about knowing what matters, what to leave out, and how to guide someone through the story. Watch and learn as our experts share their top tips.
📌 This event is part of a series. Learn more here:
https://aka.ms/DataDays/FabConRecap/y
https://aka.ms/DataDays/IntoCerts/y
https://aka.ms/AIatWork/UnderstandIt/y
https://aka.ms/AIatWork/BuildIt/y
https://aka.ms/AIatWork/ShowIt/y
[eventID:27697] Build Like a World Champion: The Decisions Behind Great Dataviz](https://i.ytimg.com/vi/ypwPVyfk5Og/mqdefault.jpg)
![Power BI Dataviz World Champs: The Road to FabCon Starts Here
The search for our next Power BI Dataviz World Champion begins during Data Days.
Join us to learn how the competition works, what it takes to qualify, and how four finalists will earn their place in the live championship finale at FabCon.
📌 This event is part of a series. Learn more here: https://aka.ms/AIatWork/BuildIt/y
For more information on Data Days and all it has to offer, visit https://aka.ms/DataDays
0:00:05 Housekeeping and Code of Conduct
0:00:53 Welcome to the speakers, Lakshmi and Aanor
0:01:24 Introduction to the Power BI Dataviz World Championship
0:03:02 Fabric Data Days and community resources
0:05:17 Contest overview and Hall of Champions
0:07:25 How to participate
0:08:37 Key competition dates
0:09:25 Judging criteria
0:10:34 Getting started: Sample file and building process
0:13:13 How to submit your entry
0:14:41 Power BI best practices: Layouts and the Bento Box structure
0:17:39 Visual design: Padding and colors (the 60-30-10 rule)
0:22:13 Accessibility in reports
0:24:58 Tips for a winning entry and game plan
0:34:04 Unofficial rules for beginners
0:40:07 AI at Work series calendar
0:42:09 Breakdown of the sample data set and task requirements
0:51:32 Deep dive into final judging criteria
#MicrosoftReactor #LearnConnectBuild #Fabric #DataDays #DataViz #FabCon
[eventID:27711] Power BI Dataviz World Champs: The Road to FabCon Starts Here](https://i.ytimg.com/vi/yvjJ2qgp5pE/mqdefault.jpg)