Jana Uhlig, CEO of startup HStreaming, talked about the company’s Hadoop-powered real-time data processing platform inside theCube, live at Strata 2013. The HStreaming solution helps companies convert big data into value by putting it through a strategy engine before being stored. The platform works over Hadoop to get data as it is generated, push it through the engine and then triggering the appropriate business process, everything in real-time.
“It’s critical that companies take actions on the data as it comes in,” Uhlig says. Explaining the HStreaming take on real-time data streaming and how its method differs from competitors, Uhlig explained that unlike other products that rely on real-time queries to help companies look through data for insights, her company takes the process one step further. An “insight does not have value on its own,” it has to further trigger and action, this is where HStreaming comes in.
To showcase how the solution works, Uhlig explained its applications for video analytics in security systems. HStreaming can match real-time data to other sources of data, identify patterns and help triggers the correct course of action. For example, as all video data is stored and allows for real-time searches for patterns, a current event is compared against previous video recordings, helping staff decide if there is a need to take action or flag the real-time event as safe.
“Big Data is bringing new problems, it comes so fast and furious,” Uhlig pointed out, stressing the need to analyze information quickly and turning it into triggering business processes.
HStreaming considers the analyst’s position, making it easy for data to be analyzed, thus being built over Hadoop was the logical choice.
Asked which is more important, throughput or latency, Uhlig states HStreaming benchmarks showed amazing values for both parameters, making it possible to pursue both.
HStreaming is also paramount is mobile advertising, especially through apps, as it allows for better targeting through its action triggering process, increasing revenues for both advertisers and app owners.
HStreaming is right at the beginning, their product is in production and they have received a first round of investment. The company is looking to expand its currently small team by adding top notch sales representatives, marketing experts and engineers that can adapt to the German industry.
Asked to comment on her take on Big Data in the cloud, Uhlig explained that HStreaming works on any Hadoop distro, fron Cloudera and EMC, to MapR or Amazon’s Redshift. She also stated that customers steer clear of cloud solutions, as most of those who “have big data and need fast decisions are big organizations,” thus they are weary about the security risks of cloud solutions.
Latest posts by Alina Popescu (see all)
- Big Data solutions should start at the business level | #IBMInterConnect - February 25, 2015
- Modernized z Mainframe: A bridge to IBM’s future|#IBMInterConnect - February 25, 2015
- The new models and pricing strategy for Internet of Things businesses | #IBMInterConnect - February 23, 2015