
SPECIALIST PROGRAM 02 | THE BRAIN
Agentic AI
Engineering
Design AI systems that can reason across work, use tools and know when to escalate.
This pathway is designed for learners who want to move beyond isolated prompts and build agentic systems around real objectives, tools, context, exceptions and human control.
PROGRAM PURPOSE
Why It Matters
Useful agentic systems must do more than generate fluent output. They must understand objectives, choose actions, interact with tools, manage context, handle uncertainty and escalate when the system should not decide alone.
Planned Competency Areas
Problem and Workflow Definition
Translate a real objective into steps, constraints, decision points and measurable outcomes.
Agent Design
Structure roles, goals, tools, memory, context and coordination across the workflow.
Tool Use and Integration
Connect the agentic system to approved data, applications and operational actions.
Evaluation and Reliability
Test whether the system behaves consistently, uses evidence and responds appropriately to failure.
Human-in-the-Loop Control
Define where the system acts, asks, pauses or escalates.
Security and Observability
Monitor behaviour, permissions, exceptions and operational risk.
Expected Evidence

Agentic Workflow
A working prototype built around a defined objective and set of approved tools.
Exception Design
Documented handling of incomplete information, unusual cases and escalation boundaries.
Evaluation Record
Evidence of test cases, failure analysis and improvement.
Technical Explanation
A teach-back and faculty-reviewed explanation of architecture, limitations and controls.
WITT Relevance
Employer Input
Participating employers may contribute workflow patterns, role requirements, exception types and approved case-study contexts.
Academic Review
Commercial relevance does not override academic independence, fairness, confidentiality or responsible-use requirements.
Candidate Access
Eligible graduates may be considered for relevant WITT opportunities based on evidence, interest, geography and employer criteria.
