12 October 2016
Today at Directions EMEA, Klaus Hansen and Andrei Panko gave a seminar on Cortana Intelligence and how it benefits Dynamics NAV 2017. Cortana gathers and analyses data and uses pre-configured analysis tools, machine learning and big data storage.
Cortana can be used as a prediction tool for many areas of NAV, such as predicting future sales based on past sales of an item, cash flow based on payables and receivables, as well as predictions of what a customer might buy based on what similar customers have bought. This could be based on buying patterns, demographic, location etc.
It will also predict if a customer will respond to a marketing campaign based on whether they have followed through from emails and offers previously, combined with some demographic information.
Time Series Forecasting looks at patterns of sales by date and looks for seasonality, patterns etc. and predicts future sales. You can then view the inventory forecast and sales forecast on the item or from a fact box from the Item list. You can also create purchase orders based on forecast predictions which is extremely helpful for longer lead time items or if running “just enough just in time” purchasing.
You can adjust the setup to look at different lengths of historical data, i.e. 12 or 24 months, as well as how far in the future you want to predict; 2 months, 6 months etc. We are also able to amend the max variance so that it will provide you with a range that it is likely to fall within. For instance, +/- 40% might be the maximum variance, if it is outside that range then there is not enough confidence and consistency and the system will say so.
There are 4 new functions that can be called from CAL to an API:
When calculating the predicted sales value it gives a predicted variance as an absolute quantity and percentage. Cash flow forecasting can tell Cortana which document types and data to include in the calculation.
In Dynamics 365, Cortana intelligence is a subscription service. You can view and create models with a monthly subscription (see image). When you've finalised setup of model you can deploy as a web service and you get an API key to connect NAV to your own online model.
With the direct marketing analysis you can pass historical information that you know the answer to and ask it to analyse and compare to real results. For instance, this type of customer looked at this type of product, we then sent a follow up email, did they click through and visit the website? It also allows you to try different modelling methods to find the one that best suits your data to predict if it's worth following up initial interest.
This is obviously just a quick overview of the new features from the new Dynamics NAV 2017. To find out more and to see the new tools in action, why not join us at the Business Software Showcase, the free event in November. We will be going over the latest updates for Microsoft Dynamics, with keynote speaker Robert Pope, the UK Dynamics NAV Channel Manager from Microsoft.