Securonix: A Cognitive Data Defender

Tanuj Gulati, CTO, Securonix Tanuj Gulati, CTO
Technology has welcomed a multitude of new opportunities into the banking ecosystem, but these opportunities come with additional risk. American Banker predicts that banks and financial institutions will realize the need to increase their monetary investments in security and data analytics—by 75 percent and 60 percent, respectively by next year. Conceived in 2009, Securonix operates at this intersection where security meets data analytics. Spearheaded by Tanuj Gulati and Sachin Nayyar—two experts in machine learning algorithms—Securonix is carving out a niche for itself in behavior-based anomaly detection techniques.

“Securonix eliminates the process of security analysts having to manually work their way through hundreds of possible threats,” begins Tanuj Gulati, Co-founder and CTO of Securonix.

Securonix automates enterprise security with intelligent software that prioritizes threats and only recapitulates the ones that need intervention. It mimics the work of a security analyst by detecting abnormalities with real-time analytic capabilities, as opposed to manual processes that may span months, or even worse, never detect the threats at all. Securonix displays a prolific array of pattern-matching algorithms—individual behavioral patterns from entities like users, systems and network, and peer group analysis that compares and contrasts one individual or system against others. By comparing unusual activities against an organization’s baseline of normalcy, Securonix identifies the critical threats that require human intervention, and advises analysts on a course of action.

“There is a money saving factor, but more importantly, there is a focus on the right set of threats that protects financial corporations from a lot of breaches,” asserts Gulati.

Addressing the diversity of systems in an organization—from firewalls to operating systems and external applications—Securonix provides over 300 different types of out-of-the-box connectors. Easily deployable and integrated with the environment, the software aggregates event-driven data and inadvertent contextual data as well.

We eliminate the manual process of security analysts having to work their way through hundreds of possible threats

“Information including travel itinerary and employment timeframes of employees are gathered along with their user roles—from tellers at the counters to database administrators—and their appropriate security clearance to classified information,” explains Gulati.

The analytic engine traverses through the data, gathering insight into the organizational workflow between people, systems and networks, to identify patterns of concern that may range from high-volume data download from critical repositories, to data upload into personal mails or dropbox. In one such instance, a New York based financial institution was facing laborious impediments with an insider-threat scenario, despite their 18-member analyst team. When Securonix intervened, they were able to zero-in on the security threats with a contextual knowledge.

“They went down from monitoring about 7,000 incidents to only seven per day, and within a month they had detected four perpetrators who were involved in data-exfiltration activities,” says Gulati.

The breadth and diversification of Securonix are among its most distinguishing features. Securonix is a one-stop-solution for CIOs to derive benefits from various datasets including identity, event-driven, and banking transactional data.

“To stand out from our competitors, we deliver continuous updates to our customers, constantly testing them on real data to ensure minimal false-positive rates – an average of four percent to five percent being the best,” delineates Gulati.

With the announcement of SNYPER—a Big Data security analytic engine – Securonix is on a quest to forage further into Big Data security analytics hosted on Hadoop platforms.

“Securonix has recently opened upto the European market and we are expanding into the APJand LATAM markets, as we speak,” concludes Gulati.