Cortana Analytics Suite: first look

About Cortana Analytics

Cortana Analytics is a fully managed big data and advanced analytics suite that enables the transformation of data into intelligent action.

It brings together and packages a range of related Microsoft products and services, including the full Azure suite of offerings – from machine learning and analytics to data warehouse and processing, as well as Power BI and the Cortana personal digital assistant.

End-to-end analytics

Bringing these products together in the cloud allows Cortana Analytics to deliver an end-to-end solution. It’s now possible to capture, prepare, analyse, present, publish, consume, learn from and take action on data, on a single platform.

‍

That confers some significant benefits and exciting possibilities, says Jim Taylor.

‍

“Cortana Analytics marks a shift from reporting about data to being able to use data to predict and change things. It’s engineered to help us build coherent pictures using data, where everything flows.”

Making analytics easier

Solving big data problems and generating meaningful insights from data that can be operationalised – turned into action and ideally a working application - is tricky and resource intensive. But with the Cortana Analytics platform it’s easier. Developers can turn machine learning models – from the Azure ML gallery or their own trained models - into REST APIs. Models can be written within Azure ML studio or in open source languages like R and Python. The Cortana Analytics platform makes it simpler to build the connections needed, and brings some of the agility that is commonplace in software development now to the world of analytics.

‍

“It’s like having a very powerful toolkit at your fingertips” says Jim Taylor. “You can learn from and implement the models others have built, and use APIs to connect software to the platform and build working intelligent, data driven apps. You no longer need a data scientist’s understanding of how every aspect of the process works, you just need to know how to use the tools.”

‍

To that end, the Cortana Analytics Gallery features a collection of solutions – including experiments and machine learning APIs – built with Cortana Analytics.

Cortana Analytics in action

One of the highlights of the workshop was seeing some of the ways Cortana Analytics is already being used.

• ThyssenKrupp Elevator is combining cloud-connected elevators with Azure machine learning to deliver predictive and even preemptive maintenance and significant business benefit.
• Self-service predictive analytics with Azure machine learning is delivering smart, energy efficient buildings.
• Schools are developing churn prediction models to anticipate at risk students and intervene before they drop out.
• Retailers are using perceptual data to deliver personalised, context-specific offers.

‍

‍

Perceptual APIs “sort of like magic”

“The session that really captured my imagination was Perceptual APIs, which are sort of like magic,” recalls Jim. With Perceptual APIs inside Cortana Analytics, you can quickly develop applications that require insights from multimedia streams such as text, audio, and image. These general-purpose machine learnt models have already been tried and tested by Microsoft and are now available as part of Azure to power applications. Some examples:

• Computer Vision & Face APIs helped power http://www.how-old.net, the age prediction website, and let your code understand and manipulate image content.
• Speech APIs enable you to communicate with users using audio, thanks to both speech recognition as well as speech synthesis.
• Text Analytics determines a range of attitude from text using sentiment analysis; in addition, apps will be able to identify key phrases behind sentiment.
• LUIS brings natural-language understanding to any application through a simple model creation UX that relies on active learning to improve the model with use over time.

‍

‍

Browse Microsoft’s Project Oxford to see the magic of perceptual APIs for yourself.

Cortana Analytics at Theta

Our research and innovation team is already looking at the possibilities of using machine learning for predicting retail up-sell opportunities, and using that to drive targeted special offers. This service could be integrated into existing applications, with the aim of delivering increased revenue from the shopping process.