AI and Power Platform Part-1 - AI with Power BI

 AI has become ubiquitously exploding and seems like every business application is now infused with it. This is also the case with many of the Microsoft business apps - Dynamics 365 and Power Platform. In this blog series I will explore the infusion of AI in the Microsoft Power Platform and business outcome benefits from it. To start this blog is about AI with(in) Power BI. 

There are many wonderful reasons to use Power BI -

1.    It is now the 'defacto' visualization tool for data with Microsoft

2.     Analytics using Power BI is easy to learn 

3.     It has built in AI capabilities that can be used without writing any code.

If you were not aware of the #3 above than now is the time to explore some of the AI capabilities. 

Power BI has many AI and AI related capabilities OOTB as detailed below -

1. Key Influencers - Provides impact and influence of key contributors on the result.

Key Influencers

The above report helps to answer Sales questions such as - What is the biggest influencer for an opportunity to be won? (or Lost) (discount as in this case and you may not need AI for that). The point is that we now have the power to quickly identify the key influencers on a particular business results. There is no more detailed modeling or manipulation of data needed and helps in understanding patterns and taking actions based on the key influencers.

2. Decomposition Tree - ability to quickly do a root cause and ad-hoc analysis using AI Splits.

Decomposition Tree

In this case we can drill down from Sales to a hierarchy of Sales by Category - by Product - by Region. The AI here is where to drill down using AI Splits to get to the next highest or lowest value. For example below shows that David So has the highest Opportunity values path.

AI Splits - Highest Opportunity value

This can be useful in identifying the biggest opportunity owner from an ocean of Sales data and making sure that the business results are met.
 

3. Anomaly detection - Auto detect anomalies in data and generate dynamic text summaries. Identifying anomalies in data may not be new but auto generating a paragraph that explains the anomaly can be very useful. It can also provide answers to specific questions. Lets look at the revenue line chart where there is a sharp decline in revenue. 

Anomaly detection

I can select the lowest point and right click > analyze > explain the decrease and Power BI shows me various possible reasons to identify the anomaly.

Analysis results

So now I know all the possible reasons there was a decrease in revenue in plain English. (If only it could tell me who is blank 😀 )

Additionally there are many other AI options like Q&A and Insights which provide great benefits to users.

You can add the AI insights to a report page using the Insert > AI visuals option or pick the AI visuals from the Visualizations pane.

I hope this helps in getting you started using AI in Power BI and let me know about your experience.

I will write more about AI in Power Platform in my future blogs. Thanks for reading.

@mihircrm

365WithoutCode

Increase Sales with Dynamics 365 AI for Sales - Part-1

This is part-1 of my blog series on AI with Dynamics 365. In my earlier blog I discussed about the need for systems of intelligence and how Microsoft has introduced a suite of AI apps with Dynamics 365. This blog is on Dynamics 365 AI for Sales.

Enterprise clients want a healthy sales pipeline with motivated sales force. It is vital to provide all necessary tools for the sales person, to be able to focus on the right leads and opportunities and close sales faster to deliver on sales targets. Dynamics 365 AI for Sales aims to do just that –

·       Help increase Sales – Business management

·       Analyze sales team performance – Team management

·       Get insights on Customer interaction – Call Intelligence

Dynamics 365 AI for Sales

Why use Dynamics 365 AI for Sales? – Large enterprises gets 1000’s of leads and opportunities from different channels and sources. They are all collected, cleansed and fed into a sales management system like Dynamics 365. The plethora of leads and opportunities can become overwhelming for a sales person, and it becomes difficult for the Sales team to focus their efforts on the important lead or opportunity and build the right customer relationship. 

Predictive Modelling - using Azure ML

Dynamics AI for Sales removes the manual guess work out of the Sales persons hand. It analyses the data within Dynamics 365 and Office 365 (exchange). It than uses Azure machine learning to creates a predictive model that scores and categorizes the leads and opportunities and identifies trends on a real time basis.
This provides indicators for the Sales team to identify the right lead and opportunity to meet their Sales goals.

Identifying trends

With the right information on the Opportunity or Lead such as the score, grade & trend is improving or declining – the Sales person and team can than work on it and get it to close faster. This will help increase the Lead qualification and Opportunity close rate and improving the overall business.

Initial observations –

1.       Dynamics 365 AI for Sales only works with data from Dynamics 365. So if you need to integrate data from multiple systems outside Dynamics 365 it may not be feasible.
2.        Requires a separate license and it is a different app. Users may not want to log into to multiple App. Though the data is also shown in Dynamics 365 – like the Lead / Opportunity scoring, grading and trends. But the AI app has some cool visualizations for management.
3.       It currently only collects data from the OOTB Sales entities – Leads, Opportunities, Goals etc. If you are using custom entities than they will not be used for the predictive modelling. Also if you are not using OOTB entities like Goals the app may not be that useful.
4.       The models cannot be customized as provided by AML. 

      Considering this is still in preview & version 1.0, I would expect it to only improve over time. My thoughts are, this is a good way of proving how you can integrate Dynamics 365, Office 365 & Azure ML and make an AI app to help the organization do more.

To achieve the above is not an easy task and needs understanding of the organization sales business process and Dynamics 365 Sales integration. Systems of intelligence cannot be built overnight and needs the underlying data from systems of engagement and systems of records. 

I hope this gives you a brief idea of the upcoming capabilities of Dynamics 365 AI for Sales. In my next blog I will write more about the implementation of Dynamics 365 AI for Sales. Keep reading.

Mihir Shah

365WithoutCode

Analytics using Power BI with Dynamics 365 - Part-3 - Data Insights

This is part-3 of a multi-series blog on Power BI and Dynamics 365. The earlier blogs are on integrating Power BI with Dynamics 365 (Part-1) and the Power BI Desktop (Part-2). This blogs focuses on data insights from Power BI with Dynamics 365.

Once you have integrated Power BI with Dynamics 365 and built your data-sets it is ready for you to get insights on your Dynamics 365 data. Microsoft research has developed advanced analytical algorithms to enable for getting quick insights from your data without having to write any code. 

It also helps to be able to quickly analyze your data without missing any significant insight. 

Below are some of the insights that I got using the sample data sets for Sales Analytics and Customer Service Analytics with Dynamics 365 and Power BI.

Email and Web are the dominant origins for Cases

Most of the Cases are questions and not problems

Cases are resolved mostly in the middle of the month

You can pin the insight to an existing or a new dashboard.

You can also expand the insight visual which opens the visual into a Focus mode. You can filter the visual in focus mode to get to the data insight you want as shown below.

Pin to a dashboard or Get further insights on this visual

Also you can get further insights on the data within the insight visual only. This is scoped insights as the data is now restricted to the visual only and not the entire data set.

Note – The insights is only possible on data uploaded to Power BI. It will not work on Direct Query.

How to get quick insights? -

In Power BI online go to My workspace > Dataset. Select your specific data set and click on the … (ellipses) in the Actions column. Than click on “Get Quick Insights” option as shown below.

Insights from a Dataset in Power BI

Now Power BI will start analyzing your data and building the insights for you. Once done it will let you know that your insights are ready for viewing.

Hope this helps you get started with getting the insights from your data. In my next blog I will write about the types of insights that you can get using Power BI and Dynamics 365.

Mihir Shah

365WithoutCode