Loading demo links...
What to study for each D-AAI-FN-A-00 domain
Concepts of Agentic AI and Agent Architectures
This domain covers what an AI agent is, levels of autonomy, and the perception-action loop that drives agent behaviour. It also covers reactive, deliberative, and hybrid architectures, planning, reasoning patterns, memory, tool use, and multi-agent systems with orchestration.
Key skills
Agent definitions and autonomy, reactive and deliberative architectures, BDI and hybrid designs, planning and task decomposition, tool use and function calling, memory types, ReAct and reflection, multi-agent coordination
Agents' Perceptions and Interactions
This domain covers how agents sense and interpret their environments, including sensors, state representations, and environment types such as observable, stochastic, and dynamic settings. It also covers actuators, agent communication protocols, grounding, multimodal perception, and human-agent interaction including human-in-the-loop design.
Key skills
Perception and state representation, environment classification, sensors and actuators, communication protocols, grounding and multimodal inputs, human-in-the-loop interaction and confirmation practices
Learning in Agents
This domain covers how agents improve from experience, including reinforcement learning concepts such as rewards, policies, and value functions. It also covers the exploration-exploitation trade-off, human feedback methods such as RLHF, imitation learning, feedback loops, online adaptation, and concept drift.
Key skills
Reinforcement learning fundamentals, rewards and policies, exploration versus exploitation, RLHF and feedback loops, imitation learning, continual adaptation, drift detection and agent evaluation
Ethics, Safety, and Compliance
This domain covers the responsible deployment of autonomous agents, including guardrails, least-privilege tool access, prompt injection defences, and human oversight for high-impact decisions. It also covers bias and fairness, transparency, privacy, audit logging, accountability, red teaming, and compliance documentation.
Key skills
Guardrails and safety controls, least privilege and argument validation, prompt injection mitigation, human oversight and escalation, fairness and bias testing, transparency, privacy, auditability and accountability