Thriving as a specialty data analytics platform, Kentik Technologies, Inc. is doing pretty well in its lane. The network performance analyzer has raised $23 million in Series B funding in a round led by Third Point Ventures, with participation by existing investors August Capital, Data Collective (DCVC), First Round Capital, and Engineering Capital, and new investors Glynn Capital and David Ulevitch.

Kentik was founded in 2014 and is known for Kentik Detect, a cloud-based network visibility and analytics solution able to process huge amounts of data per day, delivering actionable insights to large scale service providers as well as enterprise network operations teams. Emerging from stealth just a year ago, Kintek differentiates itself with highly scalable, real-time solutions built for a world in need of faster analytics than Hadoop-based offerings can provide.

The new round of funding will allow Kentik to scale and meet growing demands for its big data-based network traffic and performance solutions. Kentik also plans to use the capital for new hires and customer acquisition. 

Since the launch of Kentik Detect, the company has collected over 60 customers, including some of the world’s largest web enterprises such as Box, Yelp, Pandora, Shopify, Dailymotion, Neustar, OpenDNS and Instart Logic.

Kentik gained fame as one of five global startups chosen by the innovation arms of service provider giants Deutsche Telekom, Orange, Singtel and Telefonica through their Go Ignite alliance.

“We founded this company because we believe in the value of data to make businesses run smarter, faster, and with more innovation. We’ve been extremely gratified to see customers respond so positively to our vision and solution for network analytics,” said Avi Freedman, Kentik CEO. “This latest round of funding validates the traction we’ve achieved with our approach to the market, and will help us continue to revolutionize what network traffic intelligence means for numerous use cases across network operations, engineering, and security.”

A purpose-built platform

According to a recent report, the global analytics market is expected to exhibit a 25.68 percent compound annual growth rate (CAGR) between 2016 to 2020 because of the increase in use of analytics in various industries such as telecommunications, healthcare, managed service providers, media and entertainment, government sectors, and banking and financial service providers.

The enterprise can choose from a number of big data and analytics services available today, one of which is streaming analytics, projected to grow from $3.08 billion in 2016 to $13.70 billion by 2021, at a CAGR of 34.8 percent. Though streaming analytics is gaining traction, it does not meet all the requirements.

In a statement sent to SiliconANGLE, Alex Henthorn-Iwane, VP of Marketing at Kentik, explains that streaming analytics has its use when it comes to ingesting data at high speeds, but not useful enough for network traffic and performance analytics as it doesn’t allow much depth for data storage, as well as lacking the flexibility and power in terms of how data can be queried.

This void in streaming analytics is what prompted Kentik to innovate and come up with a solution that takes some of the characteristics of streaming analytics and combines them with raw data storage, and the ability to process queries fast.

“[W]e saw the need for further innovation to meet the full functional requirements to deliver truly actionable intelligence for network traffic analysis.  This is why we innovated our own big data engine that not only ingests at very high speeds, but has scale-out storage and processes extremely large and complex, ad-hoc queries in a few seconds,” Henthorn-Iwane stated.

Just how fast is Kentik’s platform? According to Henthorn-Iwane, it is able to process and make data available to query in around three second, which is considered streaming analytics speed. It offers a standard SaaS for 90 days of raw data storage per customer, which could translate to hundreds of billions of online data records per customer. Kentik boasts of a system-wide 95 percentile query response time of around two seconds, which encompasses queries on tens to hundreds of billions of data points.

While Kentik has already evolved beyond the big data processing methods established by Hadoop, the startup still chose to build its own platform instead of tapping Apache Spark, an open source offering keen on streaming analytics. Henthorn-Iwane sees Spark as a powerful, scalable performant, but found that building its own was more cost effective, and customized to support multi-latency. He explains that the “Kentik Data Engine, alongside its general scale and performance, make it possible to offer big data network analysis as a service that is competitive with legacy appliance products while delivering a quantum leap in performance and functionality.”

photo credit: blese via photopin cc