StreamBase 2.0 Targets Financials

By Lisa Vaas  |  Posted 2005-06-17

StreamBase 2.0 Targets Financials

StreamBase on Monday will unveil a version of its real-time stream processing engine thats focused on financial services and which will be able to snatch live data streams and record them back.

That playback ability in StreamBase 2.0 will let users model market assumptions and trading strategies on archived market data at market or accelerated speeds.

StreamBase 2.0, an update of software that can manage as many as 100,000 messages a second and provides a graphical interface for building applications, also comes with financial application templates and analytics, Java plug-ins for real-time applications, and interfaces to Microsoft Corp.s Excel or auto-generated Java GUI.

StreamBase Systems Inc. CEO Barry Morris said that the rationale behind stream processing is that theres a flood of data coming from sensor networks such as RFID (radio frequency identification) technology, and relational databases just dont have the capacity to deal with it.

"In the not too distant future, everything of material significance will have a real-time sensor connected to it," he said. "Thats a lot of real-time data that has to be handled by infrastructure. Relational technology wont have that kind of capacity. Hence our belief that theres a category called stream processing engines that will emerge."

Financial services are at the forefront of this trend, Morris said, with energy in that sector going to build applications for equity trading, options trading, bonds pricing and other types of fast-moving, data-intensive analytics—pretty much anything that happens on the trading floor.

Click here to read an interview with database legend Mike Stonebraker, creator of Ingres and Postgres and founder of Informix and, most recently, StreamBase.

As a consequence, although the capabilities in StreamBase 2.0 are generally applicable across verticals, including telecommunications, for example, theyre tuned in particular to financial services.

The updates reusable and extendable application components are designed so that customers can easily embed Java or C++ analytics within real-time streaming applications.

This will enable customers to leverage their libraries of existing code, using a "firewalled" or safe development area for application testing and debugging. Also, new "Super Boxes" will enable users to create reusable libraries of StreamBase application components.

"People in financial services may have a secret-sauce model and want to reuse it," said Bill Hobbib, StreamBases vice president of marketing.

"Not to just cut and paste it but to reference it or provide it to a bigger solution a systems integrator may be using."

Next Page: A trend toward component reuse.

A Trend Toward Component

Reuse ">

Mary Knox, an analyst at Gartner Group, said one of the most interesting new features in 2.0 is reusable and extendable application components, since it ties in with a trend toward component reuse.

"It allows for much more rapid time to market with new applications or functionality that incorporate those components," she said.

As far as data stream playback goes, Knox said it will appeal to the niche wherein people are building models. In that space, looking at historical data has been more routine than having the ability to check out real-time data, she said.

Real-time data analytics is drawing increasing attention, demonstrated by a recent announcement from Oracle Corp. that it would acquire TimesTen Inc., a maker of real-time data management software that takes information locked in legacy applications and replicates it to a relational database at what customers say is lightning speed.

StreamBase was rolled out at the Demo Conference in February. Click here to read more.

StreamBases competitive edge lies in the fact that its not a real-time database, Knox said. Rather, the software filters and analyzes data without that data needing to first be entered into a database.

"Thats where you see them being different from a TimesTen, which has a real-time database," she said.

"What [StreamBase does] is apply the filtering and analysis to in-stream data, which allows an even greater reduction in latency. It only matters for extremely time-sensitive applications, but that matters in the trading space, where sub-second delays can make dollar differences. I dont see it as a direct competitor of TimesTen, but as a different magnitude of performance."

StreamBase does have competitors in the realm of low-latency, complex events processing on multiple events streams, such as iSphere Software Ltd.

But, Knox said, StreamBase has the edge when it comes to industry specificity with regards to financial services, and even more so with regard to trading environments and the handling of data streams.

Its users are keen to avoid the spotlight, but StreamBase does claim to count among them a $7 billion hedge fund, a global investment bank, and the worlds largest all-electronic exchange, ArcaEx.

Check out eWEEK.coms for the latest database news, reviews and analysis.

Rocket Fuel