Aster AppCenter – Sharable, Repeatable Big Data Analytics

I recently spent a week at the headquarters of Teradata Aster in Silicon Valley working on the recently announced Aster AppCenter.

The basic idea of the AppCenter is to democratise access to so called big data analyses. The AppCenter allows a user to pre-package SQL, SQL-Map Reduce and SQL-Graph code into a one-click App which can then be shared with business users. When Apps are run they produce a report which can comprise tables, visualizations or a combination of both. The most basic type of App has its data source and input columns hard-coded in and will produce the report off whatever data is in the source table. Portable Apps allow the user to specify table and input columns – thereby running the App on different data sources.


AppCenter Dashboard

Any analysis which can be coded and run on Teradata Aster can be packed into an App.

This means that previously labour intensive tasks such as importing CSV data, moving data from other databases or from Hadoop, parsing semi-structured data such as weblogs, XML or json, and executing path, text and graph analyses can all be pre-packaged into a user-friendly one click app.

The way I see it, the AppCenter is part BI tool (think Tableau or Cognos), part code repository (think Git or Subversion), and part collaboration and sharing space (think Confluence) – but it is not going to replace any of these. It is a way of making analytics repeatable and sharable.


Source: D3 visualizations, sharable through AppCenter

The AppCenter caters for a range of users. Business users can point and click on a pre-built app, SQL folks can package and share their code using built in visualizations, and java programmers can build customised functions and customized visualizations. For example they can write java converters to make D3 visualizations available via the AppCenter.

To top it all off, there is a RESTful APi to facilitate the integration of AppCenter output into web applications and make them available on mobile devices.

The AppCenter is one example of a wider trend (e.g. BigML) towards making analytics accessible, repeatable and sharable. In a way, this is all to the good as it leaves more room for data scientists and analysts to pursue those aspects of analytics which still require human thought and ingenuity such as designing experiments, studying and improving algorithms, explaining insights and understanding business problems and matching them to appropriate analyses.

Ross Farrelly is the Chief Data Scientist for Teradata ANZ, Ross is responsible for data mining, analytics and advanced modeling projects using the Teradata Aster platform. Previously Ross ran Datamilk, an independent bespoke data mining consultancy specialising in data mining and advanced predictive analytics. Ross is a six sigma black belt and has had many years of experience in a variety of statistical roles including Business Development Management at Minitab and as a SAS Analyst at New Frontier Publishing. Connect with Ross Farrelly on Linkedin.

The post Aster AppCenter – Sharable, Repeatable Big Data Analytics appeared first on International Blog.

Teradata Blogs Feed