AI for Credit Officers : Practical Applications

Welcome and Session Objectives

• Introductions and session objectives
• The "AI drafts, humans decide" principle that underpins the whole session
• Overview of the session: practical workshops on core credit tasks, plus a module on AI in credit decisioning

Module 1: Where AI Fits in the Credit Officer's Workflow

• A practical refresher on prompt engineering essentials, enough for anyone in the room to follow the rest of the session, whatever their starting point
• Mapping the credit officer's workflow, from initial screening through to committee sign-off, and marking where AI can realistically help today, including where the decisioning layer sits within that map
• The golden rule: AI drafts, humans decide
• Data handling and confidentiality considerations when using AI tools on client and counterparty information

Module 2: AI-Assisted Credit Analysis in Practice

• Extracting key financial and qualitative information from a lengthy annual report
• Pulling out and structuring covenant terms from loan documentation for quick review
• Using AI as a first-pass flag for unusual accounting treatment, related-party disclosures, or other areas warranting closer attention, and why this is a starting point for investigation, not a conclusion
• Comparing two borrowers or counterparties side by side consistently
• Live Demo: Extracting covenant terms and flagging a potential red flag from a real annual report extract

Module 3: The Human vs AI Workshop: Where Judgement Still Wins

• In small teams, delegates review a detailed transaction summary, drawn from a realistic corporate credit scenario, and identify the key risks themselves
• Each team's assessment is compared against an AI-generated risk assessment of the same transaction
• Teams present their findings, highlighting where the AI helped, where it missed context or market colour that experienced credit judgement would catch, and where it introduced risks of its own, such as plausible-sounding but ungrounded assertions
• Facilitated group discussion on how to use AI risk assessments responsibly
• Interactive Workshop: Team-based human vs AI risk identification on a transaction case study, with team presentations and group debrief

Module 4: Drafting Credit Committee Papers with AI Support

• Using AI to generate a structured first draft of a credit committee paper from source materials and analyst notes
• What AI drafts well (structure, summarising financials, first-pass narrative) versus what still needs the analyst's own judgement (risk weighting, recommendation, sign-off)
• A practical checklist for reviewing and validating an AI-assisted draft before it goes to committee
• Governance considerations: version control, disclosure of AI use, and maintaining a clear audit trail
• Workshop: In teams, delegates work through a full AI-generated committee paper draft, correcting and strengthening it before presenting their final version and reasoning to the group

Module 5: AI in the Credit Decisioning Layer

• Where AI-based scoring and PD models sit in the credit process, relative to the document-based AI tools covered earlier
• A plain-English overview of how these models are typically built and what data they draw on, pitched for credit officers rather than model builders
• Explainability: why "the model said no" is not an acceptable answer, and what a reasonable explanation looks like in practice
• Model risk and governance basics: validation, monitoring for drift, and the credit officer's role in escalating concerns about model behaviour
• Where human override sits in a model-assisted decisioning process, and how to document that judgement
• Group Discussion: Given a sample AI-generated credit decision and its stated explanation, delegates discuss whether the explanation is sufficient and what they would ask before accepting it

Wrap-Up and Q&A

• Summary of where AI adds the most value across both the document-based and decisioning sides of credit work
• Open forum for questions

The AI for Credit Officers course is taught by a financial educator, technologist, and curriculum designer with over two decades of direct international experience training professionals in derivatives, fixed income, financial mathematics, and front-office finance. He has taught in the world's leading financial centres, including London, New York, Frankfurt, Singapore, Hong Kong, and Dubai, and has delivered programmes for global institutions including Goldman Sachs, Morgan Stanley, Credit Suisse, HSBC, Deutsche Bank, Barclays, and sovereign wealth funds such as ADIA, QIA, and KIA.

After gaining a 1st class honours degree in cognitive psychology from the University of Sheffield, which included a dissertation in artificial intelligence, he began his career in banking technology, working as a developer, team lead, and project manager at UBS and JP Morgan, where he helped redesign and relaunch the Pyramid derivatives trading platform for global front-office desks. He subsequently completed the Certificate in Quantitative Finance (CQF) and later lectured on the programme, delivering modules on derivatives and quantitative methods.

Over the past 25 years, he has trained analysts, associates, traders, and senior professionals inside banks, hedge funds, and many other related financial institutions. He has also taught at premier academic institutions such as London Business School and Cambridge University's Judge Business School. 

His financial expertise spans the full spectrum of markets and product delivery:

• Fixed income, bond pricing, FX, equities, and alternative investments
• Derivatives (futures, forwards, vanilla and exotic options, swaps)
• Structured products (capital-protected, credit-linked, yield-enhancing)
• Credit and risk: VaR, CDS, CDOs, ABS, MBS
• Financial mathematics, accounting, and regulation
• CFA Level I and CISI front-office exam preparation
• Traders' macro, behavioural, and Austrian economics
• Python, data science, and Excel VBA for finance
• Machine learning, natural language processing, LLMs, and other associated AI-related technologies

He also designs and develops intelligent training tools. One AI-based platform, created in partnership with the CFA Institute, generated unique CFA-aligned practice questions. He is currently building an AI legal system using LLMs and RAG for financial compliance, together with a machine learning platform for crypto cointegrated pair selection.

He is recognised for making financial education rigorous, modern, and transformative, combining technical precision with global front-office experience.

By the end of this session, delegates will be able to:

• Identify the specific parts of their workflow where generative AI can genuinely save time
• Use AI to extract and structure information from annual reports and loan documentation
• Use AI as a first-pass screen for potential red flags, while understanding its limitations
• Produce and critically review an AI-assisted first draft of a credit committee paper
• Describe, at a working level, how AI-based scoring and PD models function and what governance should surround them
• Ask informed questions about the explainability and reliability of a model-assisted credit decision
• Articulate, from direct experience in the session, where human credit judgement remains essential

This session is aimed at:

• Credit officers and credit analysts
• Relationship managers with credit responsibilities
• Credit risk managers and team leads
• Anyone who prepares or reviews credit committee papers, or works alongside credit scoring or decisioning tools

Prerequisite: None required. A brief refresher on prompt engineering essentials is included in Module 1, and the session is designed to work for delegates with little or no prior AI exposure.

A deliberately practical course, not a generic introduction to AI. Delegates get genuine hands-on time in workshops on document analysis, risk identification, and credit committee paper drafting, alongside a module giving them a working understanding of AI in the credit decisioning layer. Throughout, the emphasis is on where AI's usefulness ends and human credit judgement takes over. 

This 3-hour session shows credit professionals how to use AI to make existing credit work faster and sharper, without replacing the judgement of the credit officer. It combines practical, breakout-based workshops on four core topics (where AI fits in the credit workflow, AI-assisted credit analysis, human vs AI risk identification, and drafting credit committee papers) with a dedicated module on AI in the credit decisioning layer, covering scoring, probability-of-default modelling, and explainability.

The decisioning-layer module is pitched at awareness level for credit officers, covering what these tools do, how they're validated and governed, and what a credit officer needs to know to work alongside them, rather than a technical deep-dive into model construction. A more technical follow-on session can be arranged if required.