Framework
Pillar 2: Ecosystem Architecture
Why Ecosystem Architecture matters
Ecosystem-oriented architecture (EOA) inverts the old ‘one-app’ mindset. Architects instead build a living cloud ecosystem - a collection of reusable, highly scalable services that evolve to meet today’s needs and embrace future innovation. EOA speeds AI deployment because data is already consolidated, indexed, governed and secured, with application-lifecycle management in place. Undertaking a transition to EOA also reduces technical debt, retires legacy licensing, and lowers data-governance risk.
Core Platform Services
Core Platform Services include infrastructure, security, governance, management, and monitoring services essential for a cloud ecosystem. These services support the technical foundation required for AI strategy, ensuring data security and governance.
Data Distribution
Data Distribution involves consolidating data in storage services like OneLake, enabling downstream distribution for AI, search, APIs, and analytical workloads. It ensures that data is accessible and usable by AI services, providing the essential fuel for AI models.
Integration
Integration includes various technology-specific services like OneLake shortcuts, Dataverse virtual tables, API management, as well as batch, event, and logic driven integration. It ensures seamless data flow and avoids the creation of data silos, facilitating the scaling of AI.
Business Applications
Business Applications encompass core business systems like ERP, CRM, and HRMS, as well as the broader application portfolio. These applications are crucial touchpoints for users and both principal user experiences and principal generators of data for AI.
AI Development Tools
AI Development Tools are specific tools used by developers to build AI capabilities. They include large language models, as well as tools like GitHub Copilot, Azure AI Foundry, or Copilot Studio enabling efficient AI workload development.