Introducing Aster Guided Analytics

Teradata Guided Analytics
Teradata Employee

Let's start by dispelling the myth that a data scientist can come in a box or that any analyst can become a data scientist with the help of some magical analytics software package. For a while, a number of vendors were telling customers and prospects that they could suspend their never-ending quest to hire a data scientist because their product could fill this role just fine. But I think everyone has now come to the same realization regarding the value of the data scientist and marketers have started talking more about bridging the gap between analysts and data scientists rather than transforming analysts into data scientists.

Aster Guided Analytics is about trying to better understand the analytics division of labor and giving the less technical business analyst exactly what she or he needs to add the most possible value. There are currently three Aster Guided Analytics Interfaces (GAI):

1. The Customer Journey GAI

The CJ GAI features eye-pleasing visualizations driven by from Aster's market-leading Path & Pattern capabilities. Not only can an analyst specify patterns of interest, she or he can then view the results through a variety of different visualizations, interactively drill-down into those results, and also create targeted marketing lists via path-based behavioral segmentation.

gai_cj_final.gif

2. The Customer Satisfaction Index (CSI) GAI

The CSI GAI puts the power of Aster CSI analytics in the hands of the business analyst. The user is able to create and apply rules to generate CSI scores for customers (or any other unique identifier). The GAI is the perfect environment for exploration and iterative analysis. Once the right set of rules has been created and tested, the CSI process can be formalized and automated using with Aster AppCenter or some other kind of workflow scheduler.

gai_csi_final.gif

3. The Text Analytics GAI

The Text Analytics GAI can be pointed at any text data set and a number of prescribed analytical techniques can be executed against the free-form text. Currently, the techniques include popularity/trend analysis over time as well as sentiment analysis and sentiment trending. It should be very possible to install this GAI on a customer server, point it at an existing data set, and have valid insights in minutes.

gai_text_final.gif

The GAI's may be included as part of an Aster Solution, but they can also be sold on their own. These are field-developed, field-supported Minimum Viable Product (MVP) tools. These are not heavy-duty enterprise software packages with intensive installs and months-long implementation timelines. The GAI's are installed alongside AppCenter in the Aster cluster. Start-to-Finish installation time: less than 30 minutes.

Each of these interfaces presents the business analyst with a front-end that can guide she or he through tried and true analytics workflows with high potential for yielding insights. And while these insights are often valuable in and of themselves, the intention of the Guided Analytics Interface is to present the business analyst with the right results and the right amount of information to hand-off to the data scientist for deeper and more complex analytics inquiries. In other words, the business analyst is able to conduct their analytics investigation in the friendly environment of the web form, and once results are available, the data scientist can then take over and proceed into a more advanced investigation with tools ranging from the SQL Client for Aster SQL-MapReduce or SAS or R or Spark or whatever other tools she or he might prefer.

In a world of analytics that is becoming more and more complex and customers are wary of paying consulting group after consulting group for help, the Guided Analytics Interface is there to tell the customer: "you can do analytics too."

For more information about the Aster Guided Analytics Interfaces or the Aster Analytic Solutions, please contact your friendly neighborhood account executive or send an email to ryan.garrett@teradata.com.