Teradata Talks Apps, But Don't Think AppExchange

10 In-Memory Database Options Power Speedy Performance

What if big data could be app easy? Teradata thinks it can be, as it announced with the launch of the Teradata Aster Apps Center this week, but be sure you understand the kind of apps it's describing.

As you might suspect, Teradata's not talking about apps you can download onto a smartphone. Big data can't be that easy. The Aster Apps Center is also not a Force.com-like app-development platform-as-a-service (PaaS) or a Salesforce AppExchange-like community where companies can offer their own apps built on a PaaS platform.

The Teradata Aster App Center is a place where you can get templates for analytic applications designed to run specifically on Teradata's Aster database management system. The center is launching with 29 app frameworks in six industries: retail, telco and cable, healthcare, travel and hospitality, entertainment and gaming, and consumer financials.

Teradata has custom-built these sorts of apps in the past, so it decided to turn them into repeatable templates, including pre-built application logic, data visualizations, and configurable user interfaces, according to Chris Twogood, Teradata's vice president of product and services marketing.

"Think of it as pre-built services," Twogood told InformationWeek in a phone interview. "Instead of writing multiple authentication services for each individual app, for example, we wrote one service that all of the apps use."

[ Want more on big data DBMS options? Read 10 In-Memory Database Options Power Speedy Performance. ]

Each app template gets you about 80% of the way to a finished application ready for production use. The remaining 20% of the work entails configuration and custom integration work that's handled through Teradata consulting services.

"We want to get more adoption of Aster, and we think scaling out apps faster is the way to do it," said Twogood. "In the future we may get to a point where customers could configure these app templates themselves, but we're not there yet."

You say you're not an Aster DBMS customer? Teradata offers the option of deploying Aster and related apps on its private cloud. That cloud runs in a third-party data center and offers elastic (spin-up, spin-down) deployment in monthly increments (though not by the hour or day).

Keep in mind these are analytic applications. If you want to actually deliver product recommendations in retail, prevent churn in a telco scenario, prevent fraud in a consumer financial services scenario, or execute on whatever the analytic insight might be, you'll have to connect these analytic apps to customer-facing transactional applications or to frontline human decision-makers.

In other news from Teradata this week, the company announced support for JSON objects, metadata partitions, and international currencies in the 2.4 release of Teradata Loom. Loom is the metadata-management, data-lineage, and big data-wrangling tool for Hadoop that Teradata acquired in last year's acquisition of Revelytix. Teradata says it's making Loom available as a free download on its Hortonworks Data Platform-based cloud sandboxes, but the software also works with Cloudera and MapR Hadoop distributions.

In yet another announcement, Taradata's Think Big big data-consulting unit, which was acquired last year, has launched an eight-week, fixed-cost Data Lake Optimization Service. Aimed at companies that have deployed Hadoop-based data lakes or that are contemplating doing so, the service is designed to improve data governance and establish best-practices for structuring metadata and other data-management processes for Hadoop.

Emerging with the last two years, data lakes (or data hubs, as Cloudera describes them) serve as catch-all repositories for long-term storage of all enterprise data. In practice, however, data lakes have yet to be designed or managed consistently, according to Twogood.

Teradata Aster App templates are available immediately. Teradata Loom 2.4 will be available in March. Teradata said pricing for Think Big's Data Lake Optimization service is still being finalized. No doubt the more data and data sources filling your lake, the more costly the fixed-cost, eight-week service is likely to be.

Just 30% of respondents to our new Big Data and Analytics Survey say their companies are very or extremely effective at identifying critical data and analyzing it to make decisions, down from 42% in 2013. What gives? Get the The Trouble With Big Data issue of InformationWeek Tech Digest today. (Free registration required.)