AI Forecasting and Scenario Planning

Note: The Live Virtual course is presented in collaboration with CPA Western Provinces. The content is applicable to all participants. If you have questions regarding this course, please contact pdregistration@cpaalberta.ca.

Overview:

Spreadsheet-based forecasting answers one question: what will happen? AI forecasting answers a better one: what could happen and how likely? This course equips CPAs with three core AI forecasting tools time-series models, Monte Carlo simulation, and LLM-assisted projection and applies each to the financial decisions they face every day: rolling cash flow management, budget variance analysis, and capital planning. Every concept is grounded in a running case study (Atlas Manufacturing Inc., a mid-market company with tight debt covenants) so participants leave with frameworks they can apply on Monday. The course closes with a CPA-specific governance module: how to validate, document, and sign off on AI-assisted outputs and what professional liability looks like when something goes wrong.

CPAs who complete this course will shift from producing single-number point estimates to communicating P10/P50/P90 probability ranges, a more honest, more defensible, and more useful way to present uncertainty to boards, treasury teams, and senior management. They will also understand the professional obligations that attach to AI-assisted financial outputs, including the Air Canada precedent on AI liability.

Course Content:

Module 01 — Why AI Changes the Forecast 

• The limits of average-based, single-point forecasting
• Introducing Atlas Manufacturing: the running case study
• Where AI creates a structural advantage: non-linear pattern recognition, probabilistic outputs, automated scenario generation
• AI makes uncertainty visible, it does not replace judgment

Module 02 — AI Forecasting Tools Deep-Dive 

• Model selection rule: explainable models for reporting; advanced models for exploration
• Time-series models: ARIMA, Prophet, LSTM and when to use each
• The Interpretation Gap: why raw transaction data must be classified before modelling
• Beyond the GL: non-finance data sources (sales pipeline, HR, procurement, real estate) and the CPA's stewardship role
• Monte Carlo simulation: replacing point estimates with distributions; reading P10/P50/P90 outputs
• LLM-assisted projection: the Green/Yellow/Red usage framework for CPAs
• LIVE DEMO A: Prophet time-series forecast (Google Colab)
• LIVE DEMO B: Monte Carlo cash flow model (Excel, no add-ins required)
• LIVE DEMO C: LLM variance narrative (claude.ai)

Module 03 — Scenario Planning Frameworks 

• Scenarios as decision frameworks, not forecasts
• Base / downside / upside with linked management actions
• AI-assisted correlated stress testing: why crises don't happen one variable at a time
• Atlas applied: covenant floor analysis under P10 scenario

Module 04 — CFO & CPA Use Cases 

• Cash flow forecasting: failure story, Step 0 classification, P10 as the survival floor
• Atlas Before vs. After: same company, same data, different lens
• Budget variance analysis: from accounting to decision-making
• Capital planning: Is it still positive NPV under pressure?

Module 05 — Governance & Professional Accountability

• The CPA Control Framework for AI: Data, Assumptions, Outputs, Accountability
• Cross-functional stewardship: the CPA's obligation over non-finance data integration
• Air Canada chatbot ruling (Moffatt v. Air Canada, 2024): professional liability precedent
• Building the AI governance audit trail
• If you can't explain it, you can't sign off on it

Learning Objectives:

Upon completing this course, you should be able to:

• Apply three core AI forecasting tools,  time-series models, Monte Carlo simulation, and LLM-assisted projection,  to cash flow forecasting, budget variance analysis, and capital planning scenarios.
• Communicate probabilistic financial forecasts using P10/P50/P90 frameworks and explain the governance obligations that accompany AI-assisted financial outputs.
• Build and validate an AI governance audit trail,  including data classification, assumption documentation, prompt records, and output review, for financial deliverables.

Who Will Benefit:

CPAs working in finance, FP&A, treasury, and advisory roles who want to apply AI forecasting tools to real financial decisions — particularly those managing entities with cash flow volatility, debt covenants, or capital allocation decisions under uncertainty. No technical background required