
Design and prototype an AI agent that receives a business request, reasons over policy/knowledge, calls approved tools or workflow steps, and pauses for human approval before sensitive actions.
Enrollment for this capstone is closed right now. Get an email the moment it reopens.
Design and prototype an AI agent that receives a business request, reasons over policy/knowledge, calls approved tools or workflow steps, and pauses for human approval before sensitive actions. The capstone is designed around a build-and-validate workflow so the learner produces working evidence, not only conceptual recommendations.
SMEs want automation without giving an AI unrestricted authority. The practical challenge is separating reasoning from deterministic execution and introducing approval where financial, customer, or operational risk is material.
Your analysis must consider:
Demonstrates your ability to:
Agent instructions
Design bounded AI agents that use approved tools, retain task state, and complete workflows without uncontrolled autonomy.
Tool calling
Design bounded AI agents that use approved tools, retain task state, and complete workflows without uncontrolled autonomy.
Workflow orchestration
Design bounded AI agents that use approved tools, retain task state, and complete workflows without uncontrolled autonomy.
State handling
Design bounded AI agents that use approved tools, retain task state, and complete workflows without uncontrolled autonomy.
Human approval
Supply sDesign checkpoints where people verify uncertain, high-impact, or sensitive AI outputs before action.ecurity mapping
Working Operations Agent
A prototype agent for one bounded workflow such as onboarding, order exception handling, or service request triage.
Tool & Approval Map
Explicit mapping of what the agent may read, draft, recommend, or execute.
Evaluation & Safety Report
Task-success tests, failure cases, approval triggers, and monitoring metrics.
Industry-Recognized Certificate
Stand out with a verified certificate from NetZeroX AI.
Capstone Kick-off
Map the workflow and agent authority.
Build & Analysis
Implement agent, tools, and deterministic steps.
Practitioner Review
Test normal, ambiguous, and failure scenarios.
Final Submission & Demo
Demo the agent and submit permission, safety, and monitoring artifacts.
Completion standard: submit the working artifact, evaluation evidence, limitations, and a short final presentation. The certificate should reflect demonstrated capability rather than content consumption alone.
Analyse. Compare. Recommend. Lead with Confidence.
Build in-demand commercial and analytical skills. Solve real energy & AI challenges.
Enrollment for this capstone is closed right now. Get an email the moment it reopens.
Build a stronger portfolio with our other practitioner-led capstone across Energy and AI designed for real-world impact.