Evaluate grounding data — accuracy, relevance, timeliness, availability, permissions
Cloud Adoption Framework — AI adoption process aligns with organizational maturity
Responsible AI — fairness, reliability, privacy, inclusiveness, transparency, accountability
ROI analysis — cost of build vs buy vs extend; TCO including inference and operations
Prebuilt Agents
Copilot Studio templates, M365 Copilot agents — faster time-to-value, less customization.
Custom Agents
Microsoft Foundry, custom models, MCP extensibility — full control, higher investment.
Prompt Engineering & Models
Prompt library — reusable, versioned prompts with governance
SLMs (Small Language Models) — cost-efficient for narrow domain tasks
Model router — route requests to appropriate model based on complexity/cost
AI Center of Excellence — governance body for standards and best practices
Exam Focus Areas
Grounding data quality directly impacts agent accuracy — plan data pipelines first
Extend prebuilt agents before building from scratch when requirements align
CAF AI adoption process provides structured migration path
Practice This Domain
Test your understanding with free practice questions at /certifications/microsoft/ab-100/practice — focus on: Assess agent use in task automation, data analytics, and decision-making, Review data for grounding (accuracy, relevance, timeliness, availability), Cloud Adoption Framework AI adoption process.
Read AB-100 notes without distractions
Open Foci to run a focused study block while you review domains, tables, and exam tips.
