Accelerate Modernization at Scale: From Legacy to Cloud-Native with AI
Session Date: Microsoft Build May 19-22, 2025
Duration: ~45 minutes
Venue: Microsoft Build 2025 Conference - BRK199
Speakers:
- Mohammad Nofal (Global Black Belt Lead, Microsoft)
- Anoop Iyer (Director, Business Strategy, Microsoft)
- Michael Yen-Chi Ho (Senior Product Manager, Microsoft)
- Bryce Hunt (Founding GTM Engineer, Cognition AI)
- Tinius Alexander Lystad (CTO, Visma AS)
Link: Microsoft Build 2025 Session BRK199
Table of Contents
- Introduction and Session Overview
- The AI-Driven Modernization Imperative
- Cloud-Native Architecture Foundations
- App Modernization Guidance Framework
- AI-Powered Modernization Tools Demo
- Cognition Devin AI Agent Integration
- Enterprise Case Study: Visma’s Modernization
- Session Wrap-up and Resources
1. Introduction and Session Overview
Timeframe: 00:00:00
Duration: 5m 30s
Speaker: Mohammad Nofal
Mohammad Nofal opened the session by introducing the comprehensive speaker lineup and setting the session agenda. The session was structured to provide both theoretical framework and practical demonstrations of AI-powered modernization approaches.
The session included audience polling that revealed a diverse mix of developers (.NET and Java), architects, and decision-makers, setting the stage for content relevant to all roles in the modernization journey.
Key Session Structure:
- Need for modernization (Mohammad Nofal)
- How to modernize framework (Anoop Iyer)
- Java application demo (Michael Yen-Chi Ho)
- Devin AI and .NET demo (Bryce Hunt)
- Enterprise scale case study (Tinius Alexander Lystad)
- Wrap-up and resources (Mohammad Nofal)
2. The AI-Driven Modernization Imperative
Timeframe: 00:03:30 - 00:12:45 (9m 15s)
Speaker: Mohammad Nofal
2.1. The Scale Challenge
Timeframe: 00:03:30 - 00:05:00 (1m 30s)
Mohammad emphasized the fundamental message that “every app will be reinvented with AI, and new apps will be built with the aid of generative AI.” The critical word “every” signifies the scale challenge organizations face. Based on Microsoft’s experience helping thousands of customers on their AI journey, application modernization is identified as the most important phase in scaling AI adoption.
2.2. Microsoft’s Three-Bucket Approach
Timeframe: 00:05:00 - 00:07:30 (2m 30s)
Microsoft’s approach to accelerating AI journeys is organized into three distinct buckets:
Bucket 1: AI-Assisted Software Development Lifecycle
- Inner Loop: Developer and IDE integration with GitHub Copilot and modernization agents
- Outer Loop: Software Engineering Agent (SWE agent) for broader automation
- Third Loop: SRE agent for operational excellence and reducing SRE toil
Bucket 2: Application Deployment Platforms
- Infrastructure abstraction and operational burden offloading
- High-velocity user demand fulfillment
- Comprehensive Azure service ecosystem
Bucket 3: AI Development and Operational Practices
- AI templates for quick starts
- Framework guidance (Cloud Adoption Framework, Well-Architected Framework)
- Application modernization framework
- Operational excellence features
2.3. Azure Application Ecosystem
Timeframe: 00:07:30 - 00:12:45 (5m 15s)
Mohammad provided a comprehensive overview of the Azure ecosystem, emphasizing that with 50 million professional developers using the Visual Studio family, Microsoft has created a complete platform for any application, architecture, or language.
Developer Tooling Layer:
- Visual Studio family and Visual Studio Code
- GitHub (world’s most popular source code repository)
- Code modernization and remediation tooling
Application Services Layer:
- Azure Kubernetes Service (AKS): Full Kubernetes API access with managed cluster operations
- Azure Red Hat OpenShift: Fully managed OpenShift for on-premises OpenShift workload migration
- Azure Container Apps: Fully managed container service exposing Azure APIs
- Azure App Service: Mature platform service with complete operational management
- Azure Functions: Event-driven and serverless workloads
Data and Integration Layer:
- Managed database services (SQL Server, MySQL, PostgreSQL)
- Cosmos DB for NoSQL requirements
- Azure Managed Redis for caching
- Integration services for AI applications (chatbots, RAG applications, agents)
Security, AI, and Observability Layer:
- Security and AI services
- Third loop feedback systems (Azure Monitor, Managed Prometheus, Managed Grafana)
3. Understanding Application Modernization
Timeframe: 00:12:45 - 00:18:30 (5m 45s)
Speaker: Mohammad Nofal
3.1. Definition and Core Concepts
Timeframe: 00:12:45 - 00:14:00 (1m 15s)
Application modernization is defined as a technical and operational transformation of existing applications, processes, and data management practices to leverage cloud-native technologies. This transformation enables organizations to adopt new technologies, particularly generative AI, and infuse AI capabilities into their applications.
3.2. Modernization Categories
Timeframe: 00:14:00 - 00:17:00 (3m 00s)
The modernization approach is organized into four distinct categories, based on Gartner’s Six R framework (excluding rehost, retain, and retire):
Code and Language Platform Modernization:
- Platform/framework upgrades (Java 8 → Java 21, .NET Framework → .NET 8)
- Dependency upgrades (NuGet packages, Maven dependencies)
- Application refactoring for cloud compatibility (local file systems → Azure Blob Storage)
Replatforming to Azure:
- On-premises containerization and Azure Kubernetes Service deployment
- Cross-cloud migration (AWS → Azure PaaS)
- Database migration (on-premises MySQL → managed Azure MySQL)
Refactor and Rearchitect to Cloud-Native:
- Monolith to microservices transformation
- Event-driven architecture adoption
- API-first development approach
Process Modernization:
- DevOps process modernization with modern tooling
- Security modernization implementing zero-trust architecture
3.3. Why Modernize: Business Drivers
Timeframe: 00:17:00 - 00:18:30 (1m 30s)
Performance Improvements:
- New frameworks with reduced memory and CPU footprint
- PaaS operational burden offloading
Security Enhancements:
- Latest framework patches and reduced security exposure
- PaaS service security management delegation
Feature Access:
- Modern framework requirements (Quarkus requires Java 17+)
- Improved tooling, compatibility, and community support
Mohammad concluded this section with the memorable quote: “If it’s painful to move, then it’s probably painful to keep.”
4. Cloud-Native Architecture Foundations
Timeframe: 00:18:30 - 00:25:15 (6m 45s)
Speaker: Anoop Iyer
4.1. Four Core Pillars
Timeframe: 00:18:30 - 00:21:30 (3m 00s)
Anoop outlined the foundational pillars of cloud-native applications:
Pillar 1: Containers
- Build once, manage and scale anywhere philosophy
- Backbone of microservices architectures
- Platform-agnostic deployment capabilities
Pillar 2: Serverless Computing
- Infrastructure abstraction enabling developer focus on business logic
- Automatic provisioning, scaling, and cost optimization
- Productivity enhancement through reduced operational overhead
Pillar 3: Data Ecosystem Intelligence
- Comprehensive connectivity to analytics and intelligence services
- Day-one analytics integration requirement
- Multi-database platform support
Pillar 4: APIs
- Developer productivity acceleration through standardized interfaces
- Real-time ecosystem connectivity and data sharing
- Application monetization and amplification opportunities
4.2. Azure Cloud-Native Implementation
Timeframe: 00:21:30 - 00:25:15 (3m 45s)
Container Management:
- AKS: Full Kubernetes control for advanced users
- Container Apps: Fully managed Kubernetes with Azure APIs for simplified deployment
Serverless and Event-Driven Architecture:
- Azure Functions: Event-driven architecture implementation with developer velocity focus
Data Platform Flexibility:
- Comprehensive database ecosystem (PostgreSQL, Cosmos DB, Azure SQL family)
- Integrated analytics and intelligence capabilities
API Management and Security:
- Secure API exposure across organizations and applications
- Monetization capabilities and developer ecosystem integration
Integrated Developer Experience:
- GitHub, Visual Studio, and Defender for DevOps integration
- Complete cloud-native development stack
5. Application Modernization Framework
Timeframe: 00:25:15 - 00:35:45 (10m 30s)
Speaker: Anoop Iyer
5.1. New Framework Introduction
Timeframe: 00:25:15 - 00:27:00 (1m 45s)
Microsoft announced the release of comprehensive App Modernization Guidance at Build 2025. This framework goes beyond simple cloud migration to focus on application transformation for the AI era, providing step-by-step guidance for every stage of the modernization journey.
Key Framework Characteristics:
- AI-led scenario integration throughout the process
- Comprehensive coverage from refactor to rebuild stages
- Post-migration optimization for resiliency, observability, cost, security, and scalability
- Business and cultural transformation focus for maximum impact
5.2. Step-by-Step Modernization Process
Timeframe: 00:27:00 - 00:33:30 (6m 30s)
Phase 1: Discovery and Assessment
- Estate Discovery: Azure Migrate and Dr. Migrate tools for comprehensive application inventory
- Code-Level Assessment: Complexity, urgency, and criticality analysis
- Criticality Matrix Development: Business value, urgency, and complexity evaluation
Phase 2: Planning and Segmentation
- Modernization Waves: Strategic application grouping using Gartner framework
- Cost-Benefit Analysis: Azure Pricing Calculator and Cost Management integration
- ROI Measurement: Quantifiable investment outcomes for business justification
Phase 3: Stakeholder Alignment and Readiness
- Project Planning: Product team approach rather than project team mentality
- Readiness Assessment: People, infrastructure, and skills evaluation
- Quick Wins Identification: High-impact, low-complexity modernization opportunities
Phase 4: Execution and Factory Model
- Replatforming Execution: Initial modernization wave implementation
- Factory Model Development: Infrastructure as Code, DevOps modernization, cloud-native technology adoption
- Refactor and Rearchitecting: Advanced transformation scenarios
Phase 5: Rebuild and Continuous Optimization
- Legacy System Replacement: Cloud-native, AI-ready application development
- Continuous Journey: Monitoring insights, performance optimization, efficiency improvements
5.3. AI-Led Modernization Transformation
Timeframe: 00:33:30 - 00:35:45 (2m 15s)
The framework emphasizes a paradigm shift from traditional “lift, shift, and survive” rehosting to AI-enabled replatforming and refactoring at equivalent velocity. This transformation is enabled by:
GitHub Copilot Modernization Integration:
- .NET and Java modernization and upgrade scenarios
- Autonomous handling of repetitive tasks (dependency analysis, framework upgrades, remediation)
- Developer focus shift to high-value tasks while AI handles repetitive work
- Transparent and efficient process control with full developer oversight
6. AI-Powered Modernization Tools
Timeframe: 00:35:45 - 00:48:30 (12m 45s)
Speaker: Michael Yen-Chi Ho
6.1. GitHub Copilot App Modernization Extension
Timeframe: 00:35:45 - 00:38:00 (2m 15s)
Michael introduced the GitHub Copilot App Modernization Extension, a new tool announced at Build 2025 that integrates AppCAT (Application and Code Assessment) tool capabilities with AI-powered automation.
Key Capabilities:
- Automated Assessment: 3-5 minute analysis replacing days of manual work
- Multi-Platform Support: Java and .NET applications with additional language support planned
- Target Platform Flexibility: App Service, Container Apps, or AKS deployment options
- Comprehensive Analysis: Dependencies, databases, authentication, and configuration assessment
6.2. Live Java Application Demo
Timeframe: 00:38:00 - 00:48:30 (10m 30s)
Michael demonstrated the end-to-end modernization process using Airsonic, a complex Java 8 application:
Assessment Phase (00:38:00 - 00:40:30):
- One-click assessment initiation in VS Code
- Automated dependency analysis and vulnerability scanning
- Framework compatibility evaluation and migration complexity scoring
- Detailed report generation with prioritized action items
Framework Upgrade Phase (00:40:30 - 00:44:00):
- OpenRewrite Integration: Automated Java 8 to Java 21 upgrade
- Custom Formula System: Microsoft-built LLM-powered transformation patterns
- Automated Dependency Resolution: POM file updates and build validation
- Error Handling: Intelligent build error identification and resolution guidance
Code Transformation Phase (00:44:00 - 00:46:30):
- Pre-built Formulas: Logging to console, managed identity migration, file mounting fixes
- Custom Formula Creation: Organization-specific pattern development for reuse
- Transparent Process: Developer oversight and approval for all changes
- Real-time Validation: Maven build integration for continuous validation
Deployment Phase (00:46:30 - 00:48:30):
- GitHub Copilot for Azure Integration: Automated Bicep file generation
- Best Practices Implementation: Azure deployment guidance integration
- End-to-End Automation: From assessment to running application on Azure Container Apps
Demonstrated Results:
- Time Reduction: Weeks/months to hours/days transformation timeline
- Error Transparency: Clear identification of manual intervention requirements
- Build Validation: Continuous testing throughout the transformation process
7. Devin AI Agent for Autonomous Development
Timeframe: 00:48:30 - 00:58:15 (9m 45s)
Speaker: Bryce Hunt
7.1. AI Development Tool Landscape
Timeframe: 00:48:30 - 00:50:30 (2m 00s)
Bryce outlined the evolution of AI development tooling across three distinct levels:
Level 1: Real-Time Assistance
- Synchronous workflow with hands-on-keyboard typing
- GitHub Copilot-style autocomplete and suggestions
- 20-40% productivity improvements
- Tab-completion and real-time code generation
Level 2: IDE-Embedded AI
- Contextual AI companions within development environment
- File-level understanding and code transformation
- Interactive development support with contextual awareness
Level 3: End-to-End Autonomy (Devin)
- Task delegation to AI software engineers
- Complete workflow automation from requirements to deployment
- Asynchronous, parallel task execution
- Cloud-based virtual environment execution
7.2. End-to-End Autonomous Workflows
Timeframe: 00:50:30 - 00:53:00 (2m 30s)
Devin’s Autonomous Capabilities:
- Requirements to Code: PRD interpretation and implementation
- Build and Test Execution: Automated build processes and test running
- Error Resolution: Autonomous error identification and fixing
- Local Development: Local host deployment and manual testing
- Pull Request Management: Automated PR creation and management
Parallel Processing Architecture:
- Cloud-Based Execution: Multiple Devin instances running simultaneously
- Task Distribution: File-level task assignment across multiple agents
- 6X to 12X Productivity Gains: Demonstrated improvements for repetitive migration and upgrade work
7.3. .NET Modernization with Devin
Timeframe: 00:53:00 - 00:58:15 (5m 15s)
Microsoft .NET Upgrade Assistant Integration:
- Native Tool Leverage: Devin utilizes Microsoft’s .NET Upgrade Assistant
- Analysis and Execution: Automated code analysis and partial migration execution
- Planning and Design: Repository analysis with Mermaid diagrams and architectural breakdowns
Live .NET Modernization Demo:
- Repository Analysis: Comprehensive codebase understanding and documentation
- Upgrade Assistant Execution: Automated CLI tool execution with issue identification
- Error Resolution: Autonomous build failure and warning resolution
- Pull Request Creation: Automated PR generation with detailed descriptions
- Feedback Integration: GitHub comment processing and automatic updates
Key Demo Insights:
- Devin Wiki Feature: Popular repository documentation generation for onboarding
- Visual Feedback: Screenshot integration for better understanding
- Collaborative Workflow: Integration with existing GitHub review processes
- Transparent Process: Complete visibility into agent decision-making and actions
8. Enterprise Case Study: Visma’s Scale Modernization
Timeframe: 00:58:15 - 01:07:30 (9m 15s)
Speaker: Tinius Alexander Lystad
8.1. Organization Context and Approach
Timeframe: 00:58:15 - 01:00:30 (2m 15s)
Visma Group Overview:
- Scale: 190 software companies across Europe and Latin America
- Portfolio: 400+ SaaS products serving SMBs and public sector
- Focus: Core business process automation for higher-value task allocation
- AI Strategy: Comprehensive AI adoption across entire development organization
Data-Driven Modernization Selection:
- Technology Assessment: Identification of legacy backend technologies and expensive database systems
- Business Alignment: Product strategy and roadmap evaluation for modernization timing
- Candidate Prioritization: Strategic selection based on technical debt and business value
8.2. Three-Day Workshop Methodology
Timeframe: 01:00:30 - 01:02:00 (1m 30s)
Workshop Structure:
- Collaborative Planning: Joint sessions with development teams
- Process Definition: Tool selection and methodology establishment
- Proof of Concept: Vertical slice modernization validation
- Knowledge Transfer: Team enablement for autonomous execution
Post-Workshop Execution:
- Team Ownership: Full responsibility transfer to development teams
- Parallel Modernization: Code and infrastructure transformation simultaneously
- Technology Migration: Database technology changes and cloud adoption
8.3. Flex HRM Success Story
Timeframe: 01:02:00 - 01:07:30 (5m 30s)
Project Context:
- Application Size: Nearly 3 million lines of code
- Team Size: 30 developers
- Development Timeline: Product development since 2009
- Modernization Goal: .NET Framework to .NET 8 upgrade
Workshop Implementation (Early 2024):
- AI Tool Introduction: GitHub Copilot training from 0% to 100% adoption
- Technology Experimentation: GPT-4 and GPT-4 Turbo evaluation
- Vertical Slice Development: Collaborative small-scale modernization proof
Technical Implementation:
- .NET Upgrade Assistant: Starting point for framework migration
- Compilation and Runtime: Multi-month effort ensuring .NET 8 compatibility
- Production Deployment: Successful .NET 8 deployment achievement
- Architecture Evolution: Application rearchitecting and Azure migration
- Platform Migration: Windows to Linux hosting transition via Azure App Service
Quantifiable Results:
- Developer Adoption: 100% AI tool adoption overnight post-workshop
- Engagement Improvement: Increased development organization engagement
- Effort Reduction: 40% reduction in migration effort through AI assistance
- Cost Savings: €600,000 annual reduction in hosting and licensing costs
- Performance Gains: Significant application performance improvements
Future Modernization Roadmap:
- Database Migration: PostgreSQL adoption for further cost optimization
- Mobile Application: Complete mobile app rebuild
- Testing Enhancement: Increased test coverage implementation
- Microservices Architecture: Monolith decomposition strategy
- AI Agent Integration: Automated security, tech debt, and monitoring issue resolution
Key Success Factors:
- Upskilling Investment: Comprehensive internal training programs
- Tool Mastery: Focus on achieving maximum AI tool effectiveness
- Developer Empowerment: Training programs for optimal tool utilization
9. Implementation Roadmap and Resources
Timeframe: 01:07:30 - 01:10:00 (2m 30s)
Speaker: Mohammad Nofal
9.1. Getting Started Resources
Timeframe: 01:07:30 - 01:09:00 (1m 30s)
Self-Service Resources:
- QR Code Access: Immediate access to comprehensive App Modernization Framework documentation
- Tooling Availability: GitHub Copilot App Modernization Extension in VS Code marketplace
- Assessment Tools: Azure Migrate and Dr. Migrate for application estate discovery
Enterprise Support:
- Microsoft Account Team: Large-scale project consultation and guidance
- Partner Ecosystem: Implementation support and professional services
- Comprehensive Framework: Step-by-step guidance for enterprise-scale modernization
9.2. Key Success Factors
Timeframe: 01:09:00 - 01:10:00 (1m 00s)
Critical Implementation Elements:
- Estate Discovery First: Comprehensive application inventory and assessment
- Strategic Planning: Business value, urgency, and complexity-based prioritization
- Tool Adoption: GitHub Copilot, Devin AI, and Microsoft modernization tooling
- Team Enablement: Comprehensive upskilling and training programs
- Continuous Optimization: Ongoing monitoring, performance tuning, and cost management
Appendix
A1. Technical Architecture Details
Azure Service Ecosystem Deep Dive: The session provided extensive coverage of Azure’s comprehensive application platform ecosystem, highlighting the strategic positioning of various services for different modernization scenarios. The differentiation between AKS (Kubernetes API access) and Container Apps (Azure APIs) represents Microsoft’s approach to providing both control and simplicity options for containerized applications.
AI Integration Patterns: The demonstration of three distinct AI tool integration levels (real-time assistance, IDE-embedded AI, and end-to-end autonomy) illustrates the evolution path for organizations adopting AI-assisted development workflows.
A2. Industry Context and Competitive Landscape
Market Positioning: Microsoft’s emphasis on the 50 million professional developers using Visual Studio family products positions the company strategically in the AI-powered development tools market. The integration with GitHub (world’s largest source code repository) creates a comprehensive ecosystem advantage.
Open Source Integration: The use of OpenRewrite for Java modernization and integration with various database technologies (PostgreSQL, MySQL) demonstrates Microsoft’s commitment to multi-platform and open-source technology support.
A3. Audience Engagement Insights
Developer Community Response: The session’s audience polling revealed a diverse mix of .NET developers, Java developers, architects, and decision-makers, indicating the broad applicability of AI-powered modernization approaches across different technology stacks and organizational roles.
Practical Implementation Focus: The emphasis on live demonstrations and real-world case studies (Airsonic Java application, Visma’s enterprise transformation) provided concrete evidence of tool effectiveness and business value realization.
References
Microsoft Official Resources
- Microsoft App Modernization Guidance
- Comprehensive framework documentation released at Build 2025
- Step-by-step guidance for AI-era application transformation
- Relevant for understanding the strategic approach to modernization planning and execution
- GitHub Copilot App Modernization Extension
- VS Code extension for automated application assessment and modernization
- Integrates AppCAT tool capabilities with AI-powered automation
- Essential for developers beginning AI-assisted modernization projects
- Azure Migrate Documentation
- Application estate discovery and assessment tooling
- Critical for the discovery phase of modernization projects
- Provides the foundation for building criticality matrices and modernization wave planning
- .NET Upgrade Assistant
- Microsoft’s native .NET Framework to .NET modernization tool
- Demonstrated integration with Devin AI agent for autonomous workflows
- Key tool for .NET modernization scenarios
Third-Party Tools and Platforms
- OpenRewrite Framework
- Automated source code refactoring and framework upgrade tooling
- Integrated into GitHub Copilot App Modernization Extension for Java applications
- Important for understanding the technical foundation of automated code transformation
- Cognition AI Devin Platform
- AI software engineering agent for autonomous development workflows
- Demonstrates 6X-12X productivity improvements for repetitive migration tasks
- Relevant for organizations considering AI agent adoption for large-scale modernization
Azure Services Documentation
- Azure Kubernetes Service (AKS)
- Fully managed Kubernetes service with complete API access
- Recommended for applications requiring Kubernetes-level control
- Essential for understanding container orchestration modernization paths
- Azure Container Apps
- Fully managed container service exposing Azure APIs
- Simplified alternative to AKS for applications not requiring Kubernetes API access
- Important for modern application deployment scenarios
- Azure App Service
- Platform-as-a-Service offering for web applications
- Demonstrated in Visma case study for Linux hosting migration
- Critical for understanding PaaS modernization benefits and capabilities
Industry Analysis and Best Practices
- Gartner’s Six R Migration Strategies
- Strategic framework for application migration and modernization
- Referenced throughout the session for categorizing modernization approaches
- Provides industry-standard context for modernization strategy development
- Cloud Native Computing Foundation (CNCF)
- Industry standards and best practices for cloud-native application development
- Aligns with the four pillars of cloud-native architecture presented in the session
- Valuable for understanding industry trends and technological directions
- Microsoft Well-Architected Framework
- Architectural best practices for Azure applications
- Referenced as part of Microsoft’s comprehensive guidance ecosystem
- Essential for ensuring modernized applications meet operational excellence standards
These references provide comprehensive coverage of the tools, technologies, and methodologies discussed in the session, enabling readers to dive deeper into specific areas of interest and begin practical implementation of AI-powered modernization strategies.