Wednesday, September 27, 2017

Rise of the machines : machine learning and artificial intelligence in Dynamics 365 Part 1

A couple of months ago, I wrote a blog post Microsoft Azure Machine Learning – the power to predict which provided  a high level view  of Microsoft’s service offering around machine learning and artificial intelligence in the cloud. In the next series of posts I will delve a little deeper into a number of scenarios where we can leverage Microsoft AI offering (machine learning and Cognitive services)  together with Dynamics 365/CRM.

If you look at most legacy CRM solutions, they are nothing more than digital contact lists and notebooks where you can look up contact information, check personal details or see when someone on your team has last contacted a customer. They contain a huge amount of data, but without built-in machine learning or predictive analytics, defining priorities on which prospects to contact first for a potential deal is like searching for a needle in a haystack. The added value of a modern CRM system is, that it should guide and support your sales team so that they can focus on what they are meant to do – sell. By automating tedious tasks such as sending follow up emails and reminders, you will free up more time to focus on selling. If you are not sure of the current state of your CRM solution, just take a look at what your mobile phone experience.

Microsoft is betting big on AI (See One year later, Microsoft AI and Research grows to 8K people in massive bet on artificial intelligence) and Dynamics 365 is a platform which can benefit greatly from these investments. Although, the first built-in functionality is still quite rudimentary and some of it is still in preview – the potential seems enormous (also see How machine learning will change Dynamics 365 from Ben Hosking )

Figure 2 Dynamics 365 Customer Insights, relationship analytics and relationship assistant.  

The table below lists functionality/features within the Dynamics 365/CRM platform which uses machine learning and artificial intelligence.

Functionality Availability Remarks
Sentiment analysis within Microsoft Social Engagement Understand the public perception using sentiment analysis GA Native sentiment detection is be available in the 17 Core languages (English, Danish, Dutch, Finnish, French, German, Greek, Italian, Norwegian, Polish, Portuguese (Portugal), Russian, Spanish, Swedish, Turkish, Japanese, Chinese traditional) and Arabic
Social selling assistantThe Social Selling Assistant recommends personal actions for you to  leverage social networks (e.g. which posts to share etc …)GA
Use Azure text analytics with Dynamics 365 (Preview) Suggest similar cases US region only – Preview The rules you create can use either Cognitive Services Text Analytics or some features let you use
the built-in similarities matching engine in Microsoft Dynamics 365. Notice that you can only create one similar records suggestions rule for each entity type.
Use Azure text analytics with Dynamics 365 (Preview) Automatically suggest knowledge base articles US region only – Preview Uses Azure Cognitive Services Text analytics services
Use Azure text analytics with Dynamics 365 (Preview) Topic analysis – identifies topics/categories for
cases to allow to better manage and resolve
Currently not available
Azure Cognitive Services Topic Detection API was removed,
so this functionality is not available anymore. Expected to be updated to use the new Topic API.
Use Azure text analytics with Dynamics 365 (Preview)Using Document suggestions , Dynamics 365 can presents a list of suggested documents to the user while
the user works on an opportunity/quote etc …
GAThe Document Suggestions feature doesn't require a connection to the Azure Text Analytics service. If you choose not to use Azure Text Analytics, Document Suggestions will use the built-in keyword matching logic available in Microsoft Dynamics 365
Product recommendations (Preview) Dynamics 365 uses the Cognitive Services recommendation service to build an advanced recommendation model for automatic cross-sell product recommendations that are based on historical transaction data. US Region only – Preview
will end February 2018
and feature will be removed
Lead scoring and contact and account data augmentation

through Versium Predict
Versium predict can use CRM data to generate predictive models to score leads, to augment data on contacts or accounts, and to acquire lists of new prospective consumers or businesses from Versium's database. Microsoft has an OEM arrangement with Versium Predict which allows Dynamics 365 customers to install and use Versium Predict.
Relationship assistantProvide guidance and assistance on activities that you should perform based on interactions with customersUS Region only – previewRelationship assistant is part of the bigger relationship insights functionality which includes e-mail engagement and auto capture.
Dynamics 365 Customer InsightsAzure-based SAAS service which enables organizations to bring together transactional, web and social customer data to visualize it in a Customer 360°. DCI also allows you apply advanced data transformation algorithms (KPI calculation and machine learning) on this data to generate role-specific insights.
(*) There currently are a number of changes ongoing where some of the functionality which was powered by Azure Text Analytics and Product Recommendation APIs, will need to use the newer Azure Cognitive Services APIs.

In the next blog posts I will delve a little deeper into some of the available built-in functionality as well as look at how you can extend Dynamics 365 to integrate with Azure Cognitive Services, Azure Machine Learning, etc ...

Interesting links:

No comments: