Model router
A model router is a component that intelligently directs each request to the most suitable model for cost and quality. AB-100 covers implementing a model router when evaluating solution costs and benefits.
All AB-100 Terms
Agentic AI
Agentic AI is an approach in which AI agents autonomously plan and take multi-step actions toward a goal rather than answering a single prompt.
Multi-agent orchestration
Multi-agent orchestration is the coordination of several specialized agents that collaborate to complete a complex task.
Copilot Studio
Microsoft Copilot Studio is the low-code platform for building custom agents, topics, and prompt actions.
Microsoft Foundry
Microsoft Foundry is Microsoft's platform for building and operating custom models and the Foundry Agents service.
Foundry Tools
Foundry Tools are the Azure AI capabilities — such as vision, search, and document understanding — proposed to meet business requirements.
Model Context Protocol
The Model Context Protocol (MCP) is an open standard that lets agents connect to external tools and data sources in a uniform way.
A2A
A2A (Agent2Agent) is an open protocol for interoperability and communication between independent AI agents.
Model router
A model router is a component that intelligently directs each request to the most suitable model for cost and quality.
ROI analysis
An ROI analysis is an evaluation of the return on investment, including total cost of ownership, for an AI-powered solution.
Cloud Adoption Framework
The Cloud Adoption Framework for Azure is Microsoft's guidance for adopting cloud and AI at scale, including an AI adoption process.
ALM
ALM (application lifecycle management) is the process of managing a solution from development through deployment and maintenance.
Grounding
Grounding is anchoring an AI model's responses in trusted business data so outputs are accurate and relevant.
Prompt manipulation
Prompt manipulation (prompt injection) is an attack that plants adversarial instructions to subvert an agent's intended behavior.
Responsible AI
Responsible AI is Microsoft's standard of principles for trustworthy AI, including fairness, reliability, safety, privacy, security, inclusiveness, transparency, and accountability.
Dynamics 365
Dynamics 365 is Microsoft's suite of business applications for finance, supply chain, sales, and customer service that AB-100 architects extend with AI and Copilot.