AI Security and Governance

Few trainers in AI governance also build and operate the systems being governed, but that is the position of Dr. Alfred Ang, a full-stack AI developer, Agentic AI solution architect and corporate trainer who serves as Chief Technology Officer of Tertiary Infotech Academy Pte Ltd in Singapore, the organisation behind Tertiary Courses. Dr. Ang designs AI systems end to end, and governance controls are built into many of them. A self-service HR portal on n8n routes leave applications through human-in-the-loop manager approval and includes an OpenAI-powered HR chatbot protected by input and output guardrails. An ERP prototype requires human approval of purchase orders through email, Asana or Telegram before agents act, using signed webhooks and scheduled tasks to keep actions traceable. Dr. Ang also architected the self-hosted, AI-powered learning and training management system used for course operations, with multi-role access, one-time-password login and AI agent automation, and ai-mms, the course-registration platform that runs the Tertiary Courses sites in Singapore, Malaysia and Ghana. Smaller projects, such as an IT service desk with a credential-leak guard and strict Content Security Policy, reflect the same security-first habits. With more than 300 public GitHub repositories spanning front end, back end, databases, cloud deployment and mobile apps, Dr. Ang brings first-hand knowledge of where AI risk actually appears in production. Dr. Ang has also had to decide which AI features to release, which to restrict and who signs off on changes, giving direct experience of the trade-offs that governance committees face.

In AI Security and Governance, that practitioner view turns abstract frameworks into concrete decisions. The course opens with the foundations of AI security governance and the business risks that arise when chatbots answer customers, models score decisions and agents take actions on company systems. Participants then build an AI governance framework, defining policies, roles and risk ownership so that every AI system has an accountable owner. The third topic walks through governance controls across the AI lifecycle, from data sourcing and model selection to testing, deployment, monitoring and retirement. The final topic focuses on security controls for agentic AI, including tool permissions, guardrails, approval gates and logging, and closes with risk assessment and a practical implementation roadmap. Dr. Ang illustrates each control with examples from real builds, such as why a chatbot needs both input and output guardrails or why an agent that spends money must wait for a human approval. The teaching style is discussion-rich and scenario-based, suited to managers, risk and compliance staff, and IT leaders who need to translate policy into practice. Learners work through realistic cases and leave with a draft governance framework and roadmap they can adapt for their own organisation. Learners also discuss how to communicate AI risk to senior management and boards in plain language, how to keep an AI system inventory current and how to review controls as regulations and technologies evolve. For examples of governed AI systems in practice, view Dr. Ang's portfolio at https://github.com/alfredang.