Over the last few months, I started spending more time exploring how AI can be used in real Business Central consulting scenarios. The idea did not start with a plan to build an AI product or create another chatbot. It started with a simple observation from my daily work with customers, implementations, and support activities. While working with Business Central environments, I noticed that consultants and developers often spend a significant amount of time collecting information before they can actually start solving a problem.
The challenge is usually not that the information does not exist. In most cases, the required information is already available somewhere inside the system. The difficulty is finding the right information, connecting different pieces together, and understanding the complete picture.
A performance issue is a good example. Before identifying the real cause, we may need to review AL code, understand database behavior, check transactions, analyze customizations, and verify how users are actually working with the system. A problem that looks simple from the outside can require investigation across multiple areas. Similarly, an error message rarely tells the complete story. The error itself is only one piece of the puzzle. To understand the real cause, we often need to look at the business process, configuration, customization impact, recent changes, and the sequence of events that led to the issue.
Code review is another area where I noticed a similar pattern. A developer may review hundreds or thousands of lines of AL code to identify patterns that could create performance issues or maintenance challenges in the future. Some problems are obvious, but others are hidden in small design decisions that only become visible when the solution grows.
This made me think about the role AI could play in our daily consulting activities.
The question was not, “Can AI replace a consultant?”
The more interesting question was:
“Can AI help consultants reduce the time spent on investigation so they can focus more on solving business problems?”
That question became the starting point for my experiments with AI tools for Business Central. I started looking at where AI could genuinely assist consultants, not by replacing experience, but by helping organize information, identify patterns, and speed up analysis.
My First Approach: Making AI Useful
When I started experimenting with AI agents, my first focus was capability. I wanted to understand whether AI could genuinely help in Business Central scenarios by understanding concepts, analyzing information, and providing practical recommendations.
Like many people exploring AI, I initially focused on the obvious areas: selecting the right model, improving prompts, and connecting useful tools. The early results were promising. The AI could explain concepts, summarize information, and assist with different analysis scenarios.
However, as I started testing more real-world Business Central cases, I realized something important.
Building an AI assistant for Business Central is very different from building a general chatbot. A consultant does not need only an answer. A consultant needs context.
They need to understand why something happened, where to investigate, what could be the impact, and what action should be taken next. That realization changed my approach. The goal was no longer just to make AI answer questions. The goal became building an assistant that could support the way consultants actually work.
The Importance of Business Context
One of the biggest lessons I learned was that AI becomes truly valuable only when it understands the right context.
A general AI model may understand programming concepts, ERP processes, or technical terms, but Business Central has many specific areas where deeper understanding is required. Posting routines, inventory movements, financial entries, dimensions, extensions, events, AL development patterns, and performance considerations all have their own meaning in a Business Central environment.
Without the right context, AI can generate answers that sound technically correct but may not be practical for a real implementation or support scenario.This changed the way I approached AI development.
Instead of asking AI to solve everything, I started focusing on providing the right information at the right time and designing workflows around specific Business Central problems.
From Simple Prompts to Real AI Workflows
Another important learning came when I started building more structured AI workflows.
Initially, creating a large prompt with all instructions seemed like the easiest approach. Define the agent role, explain what it should do, describe the expected response, and add examples.
This works at the beginning.
But as the solution grows, the prompt becomes harder to manage. The instructions become longer, the context becomes heavier, and the agent spends more effort processing information before reaching the actual problem. That is when I started moving towards a better approach — using structured instructions, focused tools, and clear workflows to create AI agents that are easier to maintain and more effective in real Business Central scenarios.
Building Tools Around Real Consultant Problems
While exploring AI opportunities, I started focusing less on what AI can do and more on where consultants actually spend their time.
One area that immediately stood out was code analysis. Developers often review AL code manually to identify performance concerns, maintainability issues, or patterns that could create problems as the solution grows. Some issues are easy to spot, but many are hidden inside small design decisions that only become visible later.
Another area was troubleshooting. Many Business Central support issues are not difficult because the solution is unknown. They are difficult because finding the root cause requires connecting information from multiple places — errors, processes, setup, transactions, and customizations.
This is where I started seeing AI differently. Not as a replacement for consultants, but as an assistant that can help organize information, identify patterns, and reduce the time spent on investigation.
The real value of AI is not just answering questions. It is helping consultants reach better decisions faster.
The Biggest Learning: AI Needs Good Design
The biggest lesson I learned from this journey is that creating useful AI tools is not only about connecting an AI model with data and expecting results. The real value comes from designing the complete workflow around the AI.
A successful AI assistant depends on many decisions. What information should the agent receive? What instructions should guide its thinking? Which tools should it use? How should it analyze the situation? And how should it present the final recommendation?
These design decisions determine whether AI becomes a valuable assistant or simply another place where users ask questions.
A well-designed AI solution should go beyond providing information. It should understand the problem, help organize the available knowledge, identify possible actions, and support better decision-making.
For me, the biggest shift was realizing that AI is not only about making models smarter. It is about designing smarter workflows around them.
Looking Ahead
I believe Business Central consulting will continue to evolve as AI becomes more integrated into our daily work. However, the consultants who benefit the most will not simply be the ones who use AI tools. They will be the ones who understand how to apply AI to real business processes and practical challenges.
For me, this journey has been about learning, experimenting, and discovering where AI can genuinely create value. Every experiment, whether successful or not, has helped me understand that the biggest opportunity is not just automation, but augmentation.
The goal is not to replace the experience and knowledge that consultants have built over years. The goal is to help consultants and developers use that experience more effectively by reducing repetitive work, speeding up analysis, and making better decisions faster because the future of Business Central is not only about smarter systems.
It is about smarter ways of working, where technology supports people and helps them focus on solving the problems that matter most.
Leave a Reply