I teach what I've actually lived.

The AI risk landscape and what makes it different from traditional model risk, the global governance frameworks (MAS AIRG, NIST AI RMF, ISO 42001, EU AI Act), and how AI risk management works as a system: identifying the AI, assessing materiality, keeping the inventory honest, and applying the eleven lifecycle controls with proportionality and relevance. Activities run throughout, from risk vignettes to materiality scoring to choosing your own governance framework. It scales from a two-hour leadership briefing to a three-and-a-half-hour AIRG primer to a full-day deep dive, tailored by sector and audience.

Audience: Boards, senior leaders, risk, compliance, tech, legal, and model risk functions

Written for the people who do not build the AI but must challenge and assure it. It takes the second and third line through the AIRG from their seat: holding independent oversight of identification, inventory and materiality; providing effective challenge and independent validation; running the monitoring that catches drift; and owning the vendor checkpoint. It shows how the FEAT principles, the MAS Project MindForge AI Risk Management Toolkit and the AIRG connect as one lineage rather than three frameworks to memorise, and how to map MindForge to the AIRG control by control. Participants leave with a MindForge-to-AIRG crosswalk, from the person who aligned the two, and with the judgement to tell a real control from governance theatre.

Audience: Risk, compliance, and model validation (second line), and internal audit (third line); governance and assurance staff

Participants go under the hood of the AI systems they are asked to govern - across machine learning, generative, and agentic AI - working with live models and demos to see how these systems are built, how they fail, and how they are tested. They run fairness, explainability, and evaluation techniques for themselves, and probe systems for hallucination, data leakage, and adversarial failure. The emphasis is governance, not engineering: sharper challengers and validators who know what evidence to demand. Anchored in the AIRG lifecycle controls; laptops required.

Audience: Risk, compliance, audit, and technology teams; model validators; governance staff who want hands-on

A working session that runs on your own framework and policies, whether they already exist or are still in progress. We walk through materiality, inventory, controls, and oversight, pressure-test the choices your team has made, and ask the questions a supervisor would ask, surfacing where the framework holds and where it thins. The format is a sounding board: your team proposes, and we stress-test and sharpen together.

Audience: Risk, compliance, and governance teams

For the senior leaders who answer for AI and have to resource it, not run the models themselves. It covers what the AIRG asks of leadership in practice: setting AI risk appetite, owning accountability, keeping enough understanding to challenge what you are told, and funding the controls that actually matter. It treats AI risk as an extension of the model, third-party and technology risk the institution already manages, scaled to how much harm a failure would do, and shows how capability and risk rise together with generative AI and agents. Grounded in developing the AIRG and supervising AI risk at the MAS.

Audience: Senior management, business and function heads, and C-suite who own and resource the AI risk programme

For boards and directors who must oversee, decide, and answer for AI without running the models themselves. Covers what AI risk management asks of leadership in terms of oversight, accountability, and adequate understanding, the risk-based and proportionate philosophy and what it means for structures, how capability and risk rise together with generative AI and agents, and the global regulatory trends worth watching. Delivered to insurer and bank boards and leadership teams.

Audience: Boards, directors, and board-level risk committees

A regulator-facing programme on understanding, assessing, and supervising AI across the lifecycle. Covers the AI risk taxonomy, the global regulatory landscape, supervisory approaches, and the distinctive challenge of overseeing AI lifecycles - drawn from developing Singapore's AIRG and supervising AI risk at the MAS.

Audience: Central banks, financial regulators, supervisors, public authorities

Structured programmes for professionals who need to use AI effectively in their specific domain - without becoming data scientists. Delivered as a 1-day primer or extended multi-week programme, with tracks for market research, leadership, and other domain applications. Focuses on building durable AI intuition and practical workflows, not tool familiarity that dates quickly.

Audience: Market researchers, L&D leaders, domain professionals new to AI

How AI and AI agents actually apply to investing and markets - from data and forecasting to portfolio construction, risk, and agentic workflows. Draws on Gary's peer-reviewed finance research (ACL, ICAIF, IEEE, ACM), the "AI Agents for Investing" curriculum with working notebooks, and a course designed for fund managers. Contextualised for asset management, including fiduciary duty and proportionate governance.

Audience: Portfolio managers, analysts, quants, asset & wealth managers, fintech teams

Covers agentic AI fundamentals, how to translate existing professional skills into effective AI workflows, and hands-on practice using participants' own real deliverables. Customisable to any domain - policy writing, legal, market research, financial analysis, risk and governance, and more. The focus is on doing, not watching.

Audience: Policy writers, legal, compliance, market researchers, financial analysts, course creators, risk teams

An honest, anti-hype grounding in generative AI. Covers how GenAI actually works, its real strengths and failure modes (hallucination, sycophancy, shortcut, slop), context and task design beyond "prompting", RAG and grounding, and a practical "will AI actually help here?" decision lens. The antidote to prompt-and-pray courses.

Audience: Professionals new to GenAI, teams adopting it across functions

Modular technical track covering ML and deep learning fundamentals, computer vision, multimodal ML, explainability, and finance-specific applications including credit scoring, fraud detection, and time series forecasting. Any module can be taught standalone or combined into a broader programme.

Audience: Data scientists, ML engineers, quantitative professionals, financial services technologists

A modular workshop on prudential regulation, contextualised by jurisdiction. Centres on Basel II/III for banks - the three pillars, capital quality and buffers, leverage, and the post-2008 liquidity standards (LCR, NSFR), with Singapore's implementation (MAS Notices 637/649, the D-SIB framework) as a case study - and extends to prudential and governance requirements for other financial institutions where relevant.

Audience: Bank and insurer boards and risk teams, supervisors, visiting executive cohorts

An integrated view of the risks financial institutions run and how to manage them - market, credit, liquidity, model, cyber, and geopolitical risk - plus business continuity and resilience. Grounded in Gary's experience overseeing risk management of Singapore's foreign reserves and supervising model risk.

Audience: Risk teams, institutional finance professionals, boards, executive cohorts

An orientation to Singapore's financial system and regulatory framework - the MAS as integrated regulator, monetary policy, the banking and asset-management landscape (including family offices, REITs, and cross-border fund structures), and how Singapore introduces and enforces regulation. Delivered to senior executive cohorts from major overseas banks.

Audience: Overseas bank and regulator cohorts, visiting executives, professionals new to Singapore