Senseon: Going Beyond the Anomaly

David Atkinson, CEO

Over the past few years, machine learning technologies have been beneficial in the cybersecurity space by helping enterprises detect new and abnormal behaviors. Unfortunately, these technologies are often incapable of differentiating between unusual and malicious behavior, resulting in flooding security teams with false-positive alerts. With an absolute passion towards addressing the core issues in the cyber landscape, Senseon is the next phase of AI for cyber defense, moving beyond rules-based systems that are too rigid to keep up with emerging threats or ineffective AI systems. “Our Sensory AI approach moves beyond the current limitations of machine learning and compensates by gathering additional context and automating investigation to produce only genuinely useful alerts” says David Atkinson, CEO of Senseon.

Unprecedented technical advancements and rapid mass adoption of new technologies has created an additional layer of noise for attackers to hide within, making the detection of both interesting and malicious behavior increasingly difficult. A multiple sense approach gives Senseon’s customers complete visibility across their entire digital estates—detecting threats emerging from their endpoints and networks, with an additional outside-in layer of intelligence from Senseon’s Investigator Bots. This approach not only improves accuracy but also reduces cost, frees up resources, and strengthens the security posture.

Unlike single-point solutions, Senseon uses its unique ‘AI Triangulation’ to correlate and understand threats across its multiple senses, following the attack path and allowing Senseon to produce context-rich, meaningful, narrative-based alerts.

Our Sensory AI approach moves beyond the current limitations of machine learning and compensates by gathering additional context and automating investigation to produce only genuinely useful alerts

Despite the advanced capabilities of Sensory AI, the Senseon platform is simple to deploy. All the raw data collected from network, endpoints, cloud environments and investigator bots is standardised and stored. Machine learning algorithms are adopted within the Expert Reasoning and Machine Reasoning frameworks to analyse this enriched data. When an anomalous behavior is detected, the platform takes relevant actions through a cycle of observation and hypothesis. The Conclusion Engine makes the decision of alerting in case of any malicious activity. These alerts are then flagged for further investigation in the Case Visualiser, which shows a step-by-step account of activity.

Today, enterprise networks are in a constant state of flux, meanwhile rigid solutions that are based on rules and signatures are limited to only identifying previously known threats. Their blanket method of ‘one size fits all’ cannot adapt to the often-polar needs of the myriad of businesses in the market. “At its core, the Senseon platform is fundamentally adaptive. Its artificial intelligence learns and evolves in unison with changing business needs and structure,” claims Atkinson.

Atkinson believes that most threats cannot be accurately determined by single-point solutions alone. When installed at a London based financial organisation, a combination of Senseon’s network appliance, endpoint agent, and investigator bots accurately detected malicious remote administration tools giving backdoor access. Senseon’s AI Triangulation was able to correlate the various activities across the organisation, providing rapid detection and alerting which led to response and containment before serious harm was inflicted to the firm.

Senseon is already witnessing great commercial traction in the UK and has customers in nearly every sector including financial services, hedge funds, and global market research firms.The company believes that the market opportunity in cybersecurity has never been stronger and there is still room and appetite for innovation. “As the product evolves, and indeed as the needs of the market change, we will respond by adding additional senses such as IoT support and browser extensions into the Senseon platform, we’re also excited to improve our capabilities to include automated response,” concludes Atkinson as part their future endeavors.