Free AI Finance Masterclass: Your Complete AI Stack for Finance [Download]

Free Training · 5 Modules · 3–4 Hours Each · LLM-Moderated Personal Coaching · No Registration

Finance teams spend up to 40% of their time on tasks that AI can handle today. The gap isn’t missing tools — it’s a missing system. This Masterclass closes exactly that gap: in five sequential modules, with a continuous case study and a concrete deliverable after every session.

What if AI were your personal Finance Trainer?

Other courses explain AI tools. This one is moderated by an LLM (ChatGPT or Claude) as your personal trainer. The presentation slides are the visual anchor; the language model guides you through explanations, exercises and feedback. From minute one, you work live on the included Nordlicht AG case study. No extra effort for your own data, no registration required.

Who is the Finance Masterclass for?

The training assumes Excel knowledge; no prior AI technical skills are required:

• Finance Controllers and FP&A Managers who want to automate reporting routines and forecasting
• CFOs and Finance Executives who want to build a governance framework for AI in finance
• Accounting and Treasury Teams who want to introduce AP automation and RAG document access
• Internal Multipliers who are looking for a structured, immediately applicable AI finance training for their team

What will you learn in this Finance Masterclass?

The Masterclass consists of five sequential modules. Each module is structured as LLM-moderated personal coaching: no frontal lectures, but immediate application to the Nordlicht AG case study. Each module ends with a deliverable you can take and use right away.

🔷 Module 1 – Getting Started & Quick Wins

Fundamentals · CRAFT Prompting · Data Privacy · Building a Custom GPT

You understand how generative AI actually works — tokens, hallucination, context window — and why vague prompts still produce usable results. You apply the CRAFT Framework (Context, Role, Action, Format, Target) to your own finance tasks from the very first session. In the Building Session, you create your personal Finance GPT in 8 steps: custom system prompt, knowledge base, quality check.

• CRAFT Prompt Framework, immediately applicable to variance comments, reports and analyses
• Recognise hallucinations and validate them with the P-Q-R-Check (Plausibility · Quality · Review)
• GDPR traffic light: data classification for public chatbots
• EU AI Act Art. 4, AI Literacy certificate fulfilled by this Masterclass
• Build and test a Custom GPT / Claude Project in 8 steps

🔷 Module 2 – Data Analytics

PowerQuery · KPI Dictionary · Forecasting · IBCS Dashboard · AI Narrative

ERP exports become clean, reproducible data pipelines in PowerQuery — no VBA, no manual corrections. You learn FORECAST.ETS for 12-month forecasts with confidence bands and create an IBCS-compliant management dashboard. The AI narrative via CRAFT prompt replaces two hours of manual writing with ten minutes of structured prompting.

• PowerQuery: load, clean and transform ERP exports — automatically and reproducibly
• KPI Dictionary: consistent definitions, built once, used always
• FORECAST.ETS + confidence interval, 12-month forecast in Excel
• IBCS Dashboard standards: colours, axes, scale, no 3D
• AI Narrative: monthly commentary in 10 minutes via CRAFT Prompt

🔷 Module 3 – Virtual Finance Team

Agentic AI · 3-Agent Cascade · Make.com · HITL Gates · GDPR

AI agents receive a goal and plan the path themselves. You learn how a virtual finance team of Analyst, Reviewer and Finance Proxy is structured, and in the Building Session you build your own 3-agent cascade. Guardrails, Human-in-the-Loop Gates and GDPR-compliant workflows ensure AI suggests and humans decide.

• Agentic AI: what an agent can do and when it’s better than a prompt
• 3-agent cascade live: Analyst → Reviewer → Finance Proxy with P-Q-R-Check
• Automation stack: ChatGPT cascade · Make.com · n8n decision matrix
• Human-in-the-Loop (HITL): which gate when, GDPR-compliant
• GDPR + EU AI Act for agent workflows, audit trail and compliance

🔷 Module 4 – Accounting & Operations

RAG · AP Automation · Fraud Detection · SoD · GoBD

RAG (Retrieval-Augmented Generation) turns your financial documents into a knowledge base: credit agreement, budget policy, annual report — answers with source references in seconds instead of hours. AP Automation with OCR, LLM plausibility check and HITL gate reduces manual invoice processing by up to 80%. Fraud detection using Benford’s Law and IQR runs directly in Excel.

• RAG: credit agreements, policies and annual reports as a searchable knowledge base
• AP Automation: OCR → Plausibility Prompt → HITL Gate (3 levels)
• Fraud detection: Benford’s Law, IQR, duplicate check — no special tool required
• SoD (Segregation of Duties): applies even when AI is the originator
• GoBD Audit Trail: 10-year retention, including AI-generated documents

🔷 Module 5 – Strategy & Governance

EU AI Act · AI Roadmap · ROI Business Case · AI Governance Pack

Module 5 closes the circle: you build the governance framework for everything created in the previous modules and calculate the business case that makes your investment justifiable to the board. The AI Governance Pack (6 building blocks) and the ROI calculator are the deliverables for the final project, with which you receive the EU Art. 4 certificate.

• EU AI Act: 4 risk classes with finance relevance — what is Minimal, what is High-Risk?
• AI Governance Pack: data classification · AI Policy · SoD Matrix · Literacy certificate
• ROI Business Case: break-even, conservative scenario, pilot KPIs for the board
• AI Roadmap: PoC → Pilot → Rollout with abort criteria
• Change Management: why AI projects fail at the human level and how to avoid it

What you can do after the complete Finance Masterclass

• Build a complete AI workflow for finance — from data pipeline through agents to RAG document access
• Systematically validate AI outputs with the P-Q-R-Check before they enter reports or bookings
• Assess and prioritise finance processes by automation readiness
• Correctly classify the EU AI Act for your finance function and complete the AI Governance Pack
• Calculate an ROI Business Case and present it to your CFO or board
• Transfer the entire AI stack to your own organisation using the Nordlicht case study as a template

The red thread: Bernd and Tanja

Throughout all 5 modules, you are accompanied by Bernd (Head of Controlling) and Tanja (CFO) of the fictional Nordlicht AG — with all corresponding documents (credit agreement, annual report, dataset, budget policy). Each module builds on their progress. You don’t need your own company data.

How to start the AI-guided training

No installation, no registration. You only need a free account with ChatGPT or Claude.

One-time preparation

1. Download all files (download section below)
2. Unzip the training materials ZIP — the Nordlicht case study and all templates are included
3. Open your LLM account: ChatGPT (chatgpt.com, free) or Claude (claude.ai, free)

Starting a module (2 minutes setup each)

1. Open a new chat — always a fresh chat per module
2. Upload the system prompt: KI-Finance-Masterclass_EN_Systemprompt_v1.md
3. Upload the module content: e.g. KI-Finance-Masterclass_EN_M1_Content_v1.md
4. Have your training materials ready — keep the files from the ZIP open
5. Type “start” — the AI trainer takes over from here

🎓 Free Download — all files, no registration

Share your feedback

This training is Version 1.0 and it’s meant to get better. We welcome any feedback:

• 💬 How was your experience? Did the LLM moderation work?
• 🔧 What can be improved? Content, structure, slides, pace?
• 💡 What training would you like next? AI in Finance? AI for HR? Automation with AI agents?

Simply leave a comment below – every piece of feedback goes directly into the next version.

And if you’d like to support our work: Treat us to a coffee

Frequently Asked Questions (FAQ)

What do I need to get started with the AI Coach?

Just access to a large language model like ChatGPT or Claude (a free account is sufficient). The training materials include all necessary prompts and system instructions — you don’t need to develop or program anything yourself.

Can I use the AI Coach for my company?

Yes. For internal, non-commercial training, use with attribution to foundic.org is permitted free of charge. For commercial use, please get in touch beforehand.

How long does the training take?

Getting started is possible in 1–3 hours (Module 1). The training is modular — you can use individual modules from the 5 selectively, without completing the entire training in one go.

Is my data input with ChatGPT & Co. safe?

The training works entirely with fictional or anonymised scenarios. Do not enter real personal or company data — you will still receive practical, transferable results.