SparkCognition: Cognitive Security Analytics Provider

Amir Husain, Founder and CEO

Amir Husain

Founder and CEO

The increase in data analytics and extraction of meaningful data has become a core component of running business, supporting innovation, driving both productivity and growth. However, making sense of data is not the only factor needed for organizations to stay ahead of the curve. The increasing amount of sophisticated data breaches has made security a key focus for organizations. Enterprises are using various kinds of Security Information and Event Management (SIEM) systems to get a holistic view of organization's IT security. "A traditional SIEM can ingest terabytes of OS and application logs and run simple analyses on them, but is unable to extract insights that require a deeper level of understanding," according to Amir Husain, Founder and CEO, SparkCognition. With an aim to provide organizations with security tools, SparkCognition leverages massive quantities of data that is interpreted by Artificial Intelligence (AI), and offers solutions that find patterns of attack, identify vectors, and model attacker behavior. The company's platform is built on patent-pending pattern recognition and machine learning techniques that enable unparalleled cognitive capability.
SparkCognition's security offering,SparkSecure, employs cognitive algorithms to identify patterns of normal behavior. The software detects potentially malevolent deviations,and uses machine learning algorithms and heuristics to identify possible threats. Additionally, SparkSecure is tied into search engines so it can research anomalous log entries and determine threat. SparkSecure analyzes structured and unstructured data in the form of server logs and natural language sources such as social media to identify possible attacks or vulnerabilities in a cloud environment. It can automatically generate signatures for new threats and provide patches to protect against them. "Benefiting from our cognitive security repository, SparkSecure is able to learn about the latest threats found and use that information to detect a new threat which may share similar traits," claims Husain.

SparkPredict, another solution of SparkCognition enables truly predictive capabilities that can deliver billions of dollars in cost savings and operational efficiency improvements to machine operators. The software is used mainly in the Industrial Internet and IoT spaces to use sensor data and automatically predict system health. SparkPredict can provide estimates of the remaining useful life of large expensive assets, detect and identify specific problems, calculate measures of efficiency and assist in cross-asset, system wide performance optimization. "Our cognitive algorithms automatically develop, evolve and curate individual models for each asset," states Husain. The company differentiates itself with its AI technology,
best-of-breed algorithms and enables automated model building. Hussain takes pride in SparkCognition's team consisting of PhDs in high-energy physics, mathematics and researchers in AI,computer science and operations research."Our deep knowledge of specific industrial internet problems, and security challenges make us unique," asserts Husain.

SparkCognition caters to a vast array of industry verticals and has been empowering organizations with its security solutions. For instance, a leading utility client wanted to test out SparkCognition's claim that its cognitive algorithms were agnostic to the data and would detect anomalies and shifts from normal working conditions. The client provided SparkCognition with a data set that had two failures and challenged SparkCognition to find those two failures. "Our algorithms came back with four identified failures in the data set," says Hussain. "When the maintenance engineers at our client facility referenced maintenance logs, they found that there had in fact been four failures, but their existing analytics products, simply hadn't found all the issues. It's safe to say that they were impressed with our algorithms and capabilities."

Moving forward, Spark-Cognition plans to continue its innovation in terms of products and enhance its cutting edge cognitive analytics platform for IoT applications. "We are challenging ourselves to engineer an organization that executes flawlessly to drive customer
success and build the best Machine Learning and AI team in this space," c o n c l u d e s Husain.