Insights – 12 Floors Consulting

June 2026

Financial Dimensions vs Chart of Accounts: Using AI to Find the Right Balance

Most ERP implementations run into the same question: how much reporting detail belongs in the chart of accounts, and how much in financial dimensions? Get it wrong in one direction and your chart explodes into hundreds of near-duplicate accounts. Get it wrong in the other and every transaction becomes a data-entry minefield. And it is expensive to reverse once data starts flowing.

I put together a practical guide on how to use AI as a thinking partner to land on a balance you can actually stand behind, not to make the decision for you, but to weigh the trade-offs, run the maths, and stress-test your assumptions before you commit (Click to download the slides).

What the guide covers:

The balance problem: why the chart of accounts and dimensions pull against each other
Why it is hard to get right, and costly to undo after go-live
A 4-step AI workflow: feed it your reporting requirements, run the dimensions vs accounts analysis, stress-test stability and growth, then get a recommendation
Copy-ready prompts for each step, written to give you a usable design, not a generic answer
A sensible starting point: 3 to 5 dimensions and a chart under 300 main accounts
Three worked examples across company sizes: a 40-person consultancy, a 600-person manufacturer, and a 3,000-person multi-entity group
Where AI helps, and where final judgement and governance stay with your team
A clear what to do next so you can start this week

If you are scoping an ERP design or rethinking an existing chart of accounts and want to talk it through, feel free to connect or send me a message. Happy to help.

June 2026

AI in Consolidation: Where It Actually Helps Close the Books Faster

Group consolidation is one of the most repetitive, high-volume tasks in finance, and one of the easiest to underestimate. Intercompany matching, currency translation, journal review, and period-end commentary are exactly the areas where AI saves the time of your controller every close.

I put together a practical guide on where AI genuinely saves time, where it does not, and what to set up before you let it near your close (Click to download the slides).

What the guide covers:

What consolidation is and where AI genuinely adds value across the close cycle
Intercompany elimination: how AI auto-matches transactions and surfaces only the exceptions
Currency translation validation: catching rate misapplication before the books close
Consolidation journal review: flagging anomalies, duplicates, and missing reversals
Period-end reporting: AI-generated variance analysis and first-draft management commentary
Worked examples with time savings calculated for each use case
Where AI does not help and what still requires finance judgement and sign-off
What to set up before you let AI near your close

If you are working on your period-end close and want to talk it through, feel free to connect or send me a message. Happy to help.

May 2026

AI in ERP Migrations: Where It Actually Helps with Inventory Valuation

One of the most underestimated risks in any ERP migration is inventory valuation at cutover. Not just quantities — the cost layers behind them. Get that wrong with FIFO or average cost, and you have a direct P&L problem.

I put together a practical guide on where AI genuinely saves time, where it does not, and what questions to ask before cutover day (Click to download the slides).

What the guide covers:

The problem with inventory valuation at cutover and why it matters for your P&L
Where AI actually saves time: data cleanup, FIFO layer reconstruction, average cost validation
Worked examples for both FIFO and average cost migrations
Where AI does not help and what still requires human judgement
The hidden costs of AI-assisted migration that most projects underestimate
A realistic starting point: how to pilot, measure, and scale
Key takeaways and what to do next before cutover day

If you are working through an ERP migration and want to talk it through, feel free to connect or send me a message. Happy to help.

May 2026

From Idea to Website: A Plain Language Guide Using AI

I built 12floors.com using Claude AI and a basic hosting package. It is a simple site, and for what I needed, an AI assistant was enough to get it done.

In case it helps someone else who just needs a simple website, I made this step guide (Click to download the slides).

What the guide covers:

What makes a good webpage and why it matters
What Claude AI is and how it writes the code for you
Step 1: How to get Claude for Windows
Step 2: The tools you need and how to set them up
Step 3: How to use Claude to build your pages
Step 4: How to choose a web hotel (hosting) and get online
Step 5: How to build, test, and upload your finished website

If you are curious or stuck somewhere, feel free to connect or send me a message. Happy to help.

March 2026

AI for Business: What It Actually Means and What to Ask

Most people I talk to are about to waste their AI budget, the valuable time of their employees, and risk alienating their customers.

I put together a plain language guide covering everything an employee needs to make good AI decisions. Not theory. Practical building blocks, with real business examples throughout (Click to download AI presentation and guide).

What the guide covers:

Why AI matters and a short history of how we got here
Core concepts: tokens, context windows, prompt engineering
AI Workspaces, RAG, embeddings and vector databases
Hallucinations, what causes them and how to prevent them
LangChain, LangGraph, MCP and AI Agents explained simply
How it all connects, from a basic chatbot to full workflow automation
10 questions worth asking your AI vendor before you sign

Feel free to connect or send me a message. I am always happy to talk it through.