“The machines are taking over! Just kidding, we’re at least 30 years from that happening. However, they are becoming a lot more helpful!”


When investing in ERP software, you are not looking for a system to just handle your needs of today, rather your investment should be in technology that will scale and provide cutting-edge solutions as part of the core platform. In this article, we will set the stage by reviewing what Artificial Intelligence and Machine Learning mean, then cover the FREE A.I. extensions that are available for Dynamics 365 Business Central, and most importantly, I will point out some reasons why you should care.

What are A.I. and M.L.?

Before discussing what A.I. extensions are available, it is helpful to understand what Artificial Intelligence means, and in the context of these Dynamics 365 Business Central extensions, how Microsoft can deliver these solutions. Artificial Intelligence is the broader concept of machines being able to carry out tasks in a way that we would consider smart. Machine Learning is an application of A.I. based on the idea that we should be able to give machines access to data and let them carry out tasks and learn for themselves. The rise of the Internet fueled Machine Learning as you can connect neural networks, or computer systems designed to work by classifying information the same way a brain does (EX. Recognize images and classify them according to elements they contain), to huge amounts of digital information from all over the world. These neural networks are essentially a system of probability, that based on the data fed, it can make statements, decisions, or predictions with a degree of certainty. The addition of a feedback loop enables “learning.” By sensing or being told whether its decisions are right or wrong, the model modifies its approach in the future. Microsoft has so much access to data via their cloud applications that they have an established data set to pre-train these neural networks, which allows for quicker adoption by companies wanting to get started with A.I., but do not have a lot of historical data in the cloud.

Built-in A.I.

Because Microsoft has access to a tremendous amount of data, they can deliver A.I. solutions natively within their business applications. Three conceptual model-types provide the built-in A.I. functionality for Dynamics 365 Business Central: 1) Time Series; 2) Image Analysis; and 3) Prediction. These are reflected in the FREE extensions available – Intelligent Cloud Insights, Cash Flow Forecast, Sales and Inventory Forecast, Late Payment Prediction, and Image Analyzer – all of which we will explore further.

Intelligent Cloud Insights

In a nutshell, Intelligent Cloud Insights is a collection of KPI’s and insights based on Machine Learning that can surface potential areas of concern. KPI’s included are Cash Availability, Sales Profitability, Net Income, Inventory Value, all in easy to consume graphs and figures. Insights such as customers with late payments, invoices you haven’t paid on time, and customer and vendor payments due in the next period are included. These insights are all available from your home screen to easily surface information to better run your business. You can also include Power BI reports for additional business intelligence. You can even access these tools with the Team Member license, which is great for senior leadership and executives, plus they are drill-downable so that you can quickly uncover the details and act on pertinent issues.

intelligent cloud insights

Cash Flow Forecast

Cash Flow Forecast is an advanced reporting tool that uses Microsoft Azure Machine Learning (which is free in Dynamics 365 Business Central, just turn it on, however you are limited to the amount of monthly compute time) to model future cash flow predictions based on historical data. Considering various factors such as liquid funds, expected payments via open entries in the customer and vendor ledgers, tax entries, budgets, historical trends, and this model provides enhanced cash flow forecasting. You can set up a variance percentage to specify a range of deviation that you will accept in a forecast, and forecasts outside of that range are considered inaccurate and ignored. Note that you must have at least two years of data to utilize this extension, and you can forecast up to four months ahead with reasonable certainty (the more data, the more accurate your outputs). This revolutionizes how your accountants will ensure the business can meet its financial obligations. You will also be able to better plan for CAPEX and OPEX initiatives by having a clear view of future cash flow.

forecasting model azure machine learning studio

Sales and Inventory Forecast

Sales and Inventory Forecast use Cortana Intelligence to analyze historical data to predict future demand, enabling you to base procurement decisions on accurate forecasts. To make predictions about future sales and inventory demand, the web service consumes data about past sales like posting date, item number, and quantity. You can specify the historical periods to calculate the forecast by and can forecast by month or quarter. For this extension, it requires at least 5 periods of historical sales to predict futures. Otherwise, you will receive the “there is nothing to show in this view” dialogue. This extension even helps you create replenishment requests for your vendors to fully support your customer demand. Use this tool to avoid lost revenue from not having inventory on hand to meet customer demand, and boost brand reputation by always delivering on orders. Optimize your shipping costs by better planning replenishment. On top of these functional benefits, you will discover trends in your sales and inventory management.

sales and inventory forecast

Late Payment Prediction

Late Payment Prediction can help you reduce outstanding receivables and refine your collection strategy by predicting whether sales invoices will be paid on time. Note that you may not be able to use the model initially because of the amount of data, or the data does not contain enough variation or does not meet the quality threshold. You can modify the quality threshold as desired. The data for this extension is derived from each customer ledger entry that has a related posted sales invoice that includes the amount, payment terms, whether applied credit memo exists, and is enriched with aggregated data from other invoices related to that customer, such as total number of paid invoices, the total number of invoices paid late, total number of outstanding invoices, average days late, and several ratios based on the same data set. So, when you open a posted sales document (Quote, Order, and Invoice), a notification will display with a confidence prediction on the payment receipt. Important areas to evaluate are the Payment Prediction, which indicates whether the payment is predicted to be late, and Prediction Confidence, which indicates how reliable you should consider the prediction to be (High or at least 90% certain, Medium or between 80-90% certain, and Low or below 80%). Dynamics 365 Business Central automatically runs training and evaluation to stay current on this prediction, plus you can manually run it. You can even build your own custom models if you are so inclined. Beyond reducing your DSO, this improves the financial health of your organization by improving cash flow. You will also find better agility with customer payment terms and payment methods and surface often delinquent payers.

Image Analyzer

Image Analyzer extension uses the Computer Vision API for Azure Cognitive Services to detect attributes in the images you import for your items and contacts. By default, it only suggests attributes if it is at least 80% certain. There are image requirements for file format (jpeg, png, gif, bmp), size (less than 4MB), and dimensions (greater than 50×50). Analyzing a picture of an item provides attributes that can be added to the item master or inserted into the item description. You decide what actions to perform, whether you ignore, use as an attribute, or use as a category. As mentioned, you can analyze a picture of contact, however that may be less desirable as the attributes are age and gender with no confidence level for either (does it matter what age or gender??).

Why You Should Care!

In three words: Efficiency, Accuracy, Health. Become more efficient by making a business decision earlier and streamlining business processes and data entry. Become more accurate in your forecasts and decisions with the feedback loop. Improve your organization’s fiscal health, inventory valuation, and, most important, your customer relationship health.

In Conclusion

These extensions are exclusively available with Dynamics 365 Business Central hosted in the Microsoft cloud. You can deploy these extensions at any time, just consider the amount of historical data you have in your system. For those of you with on-premises deployments thinking,” I just read all of this and got excited, and you’re telling me I can’t access it?” – don’t fret! You CAN access these tools through the Intelligent Edge, as there are solutions for Dynamics NAV and BC on premises and Dynamics GP. Better yet, just call us, and we can help you take advantage of this futuristic technology today!

Feel free to connect on LinkedIn and Twitter! If you’re interested in more articles on Business Central check out our blog on Top 3 New Features for Business Central