Course Format and Delivery

Delivery Method: Instructor-led LiveOnline

About this course

Every example in this session runs on your own numbers: the forecast, the risk and opportunity register and the segment bridges you work with each month. Generic prompt courses teach a good prompt against an invented example. They do not show what happens when that prompt meets a real forecast tab. That is where finance AI use stalls. The risk is not that your team ignores AI. It is that your team tries AI on real data without knowing where the tool breaks.

A field study of 79 accounting firms in the Journal of Accounting Research found AI adoption shifted roughly 8.5 percent of accountant time from data entry to quality assurance and shortened the monthly close by about seven and a half days. Experienced accountants gained the most, and 62 percent worried about errors in AI output. This session is built around both findings. It runs three use cases from your own work, each demonstrated live on a sanitized copy of your data with a verification step, and closes with the governance rules a finance team needs before AI touches a real reporting cycle.

What you will learn

• Explain how a language model produces an answer and predict where it will fail on financial data

• Ask five questions before any AI use, starting with whether the data may leave the company

• Run spend, scenario and register analysis in Gemini and NotebookLM, and verify each answer to source

• Explain how a Gemini-written Apps Script builds a management deck, and what to review before it runs on a real cycle

• Apply three governance rules: a named owner, verification proportionate to consequence, and a clear line on what never enters a public tool

Participants receive a Prompt Pack with a verification step for every prompt, lab workbook access for practice, and a five-slide reference on how finance teams use AI and get started.

This course is great for

FP&A, controllership, business finance and management reporting teams.

Topics Covered

1. Open and outcomes. What the session is and is not. Poll 1: where the room stands on AI today.

2. AI literacy for finance. How a model produces an answer and why it can be fluent and wrong. The five questions. The demo data rule.

3. Spend and working capital. Gemini in the Sheets side panel on a flat file. Quarter-on-quarter movements by account, cost centre and site. Where it breaks and why.

4. Forecasting, scenarios, R&O. NotebookLM on the forecast summary, bridges and register. A live scenario. The register reconciled against its own legend.

5. Hands-on. Pairs run one prompt from the Prompt Pack on the lab workbook, verify one figure to source, and report back.

6. Management reporting. Variance commentary from the customer and site bridges. A pre-built, tested Apps Script turns a bridge into a six-slide deck.

7. Governance. Three rules, and what still needs a decision in your environment.

8. Poll 2, Q&A, getting started. The shift from Poll 1 is the point. From first session to implementation: one process, one owner, one verified result, then scale.