Tecman Talks AI Field Notes: How we cut time spent creating Business Central Word layouts from hours to minutes 

In this edition of Field Notes, our no-nonsense AI blog series, we’re looking at how one of our Business Central consultants took a task that could swallow up several hours and turned it into a process that can take just minutes. 

The challenge: recreating customer documents in Business Central  

Microsoft Dynamics 365 Business Central contains hundreds of reports covering different areas of a business, from sales and purchasing to warehousing and invoicing. Each report has its own dataset: the fields that Business Central can pull into the finished document.  

The layout itself is based on a Word document. When it’s printed or sent through document delivery, it might be converted to PDF, but the underlying design and data still come from that Word layout.  

For customers moving from another system, this can create a deceptively time-consuming job – rebuilding documents created outside of Business Central as valid Business Central Word layouts.  

A single report can contain 30 or 40 fields, so even a relatively straightforward layout can involve a lot of searching, matching and checking. It wasn’t unusual for one report to take around two to two-and-a-half hours. 

Where did AI come in?

One of our consultants began by creating a prompt in Cowork and testing how far it could take the process.  

After giving Cowork the right supporting information, the consultant was able to use it to rebuild a Word layout around the customer’s design and map the content to fields that genuinely exist in the Business Central report dataset. 

In recent tests, the whole process – exporting the layout, running the prompt and importing the finished version back into Business Central – took just five minutes (although it’s worth noting that’s a baseline rather than a guarantee). That’s an impressive 95+% potential time save.  

What was the AI actually doing?  

The clever part is not simply making a document that looks right. It’s making sure that the document is built on data Business Central can actually supply.  

The prompt maps fields but it also knows not to invent values that aren’t already in the customer’s dataset. Instead, it flags where values are missing, making the gap instantly visible and prompting the consultant and customer to consider removing the item, using an existing alternative or including extra development in the project scope.  

This is a good example of AI working within firm guardrails. It’s not creating plausible-looking information; it’s being asked to work with a defined dataset, show its workings in the finished layout and flag anything it cannot support.  

Importantly, this process speeds things up – but the consultant is still needed. Any AI output must be reviewed by an expert eye, but instead of using several hours to search for and bind fields manually, the consultant can review the output, resolve exceptions and focus on the more complex parts of the project.  

Why this matters for customers  

Traditionally, Word layouts are tackled later in an implementation, often when the project reaches testing. This can mean that developers are brought in late in the project if anything’s been missed.  

When a layout can be produced in minutes, the conversation can move much earlier. A consultant can ask for sample documents during an initial workshop, run them through the prompt and identify missing fields while the requirements are still being shaped.  

For customers, the immediate benefit is less time and budget spent recreating documents. But the value doesn’t stop there: 

• Multiple layouts can be created more consistently, reducing small differences that can creep in during manual work  
• The finished file remains a standard Word layout, so it can be maintained and edited in the familiar Word interface  
• Business Central captions are retained, supporting translations and future caption changes  

The result is not just a quicker document build. It is a more transparent process, with gaps surfaced early and fewer surprises later in the project.  

Could the same approach work in your business? 

This example is specific to Business Central Word layouts, but the principle is much broader. Many businesses have tasks that involve recreating documents, comparing a desired format with available data or checking whether a system can supply every item a process requires.  

AI can help with that groundwork when it’s given clear source material, firm rules and an expert who understands what a correct result looks like. The goal is not to let AI fill in the blanks. It’s to make the blanks easier to spot.  

What we’ve learned  

The biggest lesson is that speed creates value only when the output can be trusted. In this case, the prompt is useful because it stays anchored to the real Business Central dataset and clearly flags anything it can’t find.  

The second lesson is that a useful AI process improves through real examples. Because our consultant was already receiving plenty of layout work, he could test the prompt repeatedly, correct problems as they appeared and add those fixes to the rules. When a defect is corrected in the prompt, every layout created afterwards can benefit.  

It’s still a work in progress, but the direction is already clear: repetitive layout work can move from hours to minutes, while the judgement, quality checks and customer conversations stay firmly with the consultant.  

What’s next? 

Ready to see where AI can fit in your business solutions? Get in touch today.