Mark Gazit | CEO | ThetaRay

ThetaRay: Rendering Exemplary Fraud Detection Services

In an interview with ‘Insights Success’, Mark GazitCEO of ThetaRay shares insights on how the company helps clients at large financial organizations, cyber security divisions and critical infrastructure become more resilient and seize opportunities. Also, he broadly discusses the company’s core competencies and the services it offers.

Below are the highlights of the interview conducted between Mark and Insights Success:

Give a brief overview of the company and its vision. 

ThetaRay was founded in 2013 by acclaimed mathematicians Prof. Ronald Coifman and Prof. Amir Averbuch with the goal of transforming the way the world benefits from data. We are dedicated to helping clients at large financial organizations and Industrial Internet of Things (IIoT), companies that have become more resilient against threats.

Our advanced analytical solutions are based on AI and machine learning technologies, built on proprietary algorithms developed by Coifman and Averbuch throughout 15 years of research. Their breakthrough technology, hyper-dimensional multi-domain big data analytics, has the distinctive ability to fuse and analyze massive amounts of heterogeneous data from diverse sources like network traffic, financial transactions and database records. This holistic, all-seeing technology provides automatic, unsupervised, real time discovery of the “unknown unknowns” — threats and risks that are not detected by existing rule-based solutions.

ThetaRay lets math discover meaning in the data without making any assumptions. It has no need for any semantic or contextual understanding, predetermined patterns, rules or other known elements. The technology operates with unprecedented speed, accuracy and scale, enabling clients to manage risk, detect money laundering schemes, uncover fraud, expose bad loans, identify ATM hacking, and more. With offices in Israel, NY, London and Singapore, ThetaRay is privately backed and has raised over $65M in investment. The company is led by cyber expert CEO Mark Gazit.

How do you diversify your products and solutions in order to benefit your customers? 

Our platform is industry-agnostic, so it can be used in any sector. However, we currently focus on the financial and IIoT sectors. Our three financial services solutions are:

Anti-Money Laundering: Today’s criminal organizations are intelligent enough to bypass existing AML rules and knowledge. When they launder money, they use small transactions that look legitimate. However, ThetaRay identifies anomalous patterns of behavior that suggest money laundering is taking place, allowing banks to intercept the crime in its initial phases.

Fraud detection: With the transition to digital, banks must manage attacks exploiting all their new and existing channels and products. With unsupervised machine learning, they can detect fraud without predefined thresholds or assumptions.

ATM Security: We are seeing organized attacks on ATM control networks. Instead of creating skimming devices and trying to fool the machines themselves, these criminals are engaging in massive attacks that penetrate the management and control networks of ATMs and make them distribute large amounts of money. When this occurs, it is a large-scale event that is almost impossible to identify in real-time. However, ThetaRay can detect it in its earliest stages.

Describe the experiences, achievements or lessons learnt that have shaped the journey of ThetaRay.

We initially launched the company as a cybersecurity provider for critical infrastructure but, instead of hunting viruses, our technology looked for slight anomalies in everyday processes. This allowed it to detect both cyber-attacks and general equipment malfunctions. We later realized that financial organizations face similar risks as critical infrastructure, but with greater financial losses at stake. As a result, we shifted our focus to the financial sector.

What are the evident challenges in the Fraud detection Solutions industry? 

The key issue is that the banking industry’s traditional rule based fraud detection methods do not work anymore, because most fraud now takes place online and most criminals know the rule thresholds. Add to that the fact that financial organizations are generating massive amounts of data, and you can understand why they might feel helpless against new types of threats.

Describe the significance of machine learning in fraud detection space. 

Machine learning is an important trend in fraud detection, since traditional solutions are incapable of detecting today’s complex threats. Even the industry’s regulatory bodies have begun suggesting that banks use new technologies such as AI. However, not all machine learning solutions are created equal, and we are seeing industry confusion over three types:

Supervised machine learning: This is what most of the vendors are using, and it’s essentially the same thing as the old rules-based systems: you tell the machine what to look for and it finds it. Unfortunately, new and unfamiliar schemes are missed entirely.

Unsupervised machine learning: A few companies are offering unsupervised machine learning solutions, which can detect unknown threats based on anomalous behavior. Unfortunately, this all takes place in a ‘black box,’ so banks cannot submit SARs based on these conclusions.

Intuitive machine learning: This is what we call our form of AI, which is unsupervised yet transparent. It is the only machine learning-based solution on the market that completely explains every decision it makes, and thus enables banks to submit suspicious activity reports (SARs), which is a critical point for regulators.

What are the current trends that are driving the industry? 

We see two key trends driving AML and fraud detection today:

Ongoing and escalating penalties: Even though banks are using fraud detection and anti-money laundering solutions, they continue to get fined – and even indicted – for AML violations. A big part of the reason for this is because criminal enterprises are using extremely sophisticated techniques to funnel and cleanse money that is used to finance human trafficking, drug trafficking, terror financing and other illegal operations. These groups know the rules and thresholds that banks’ detection systems look for, and subvert them in some very clever ways.

New regulatory support for AI: The Financial Crimes Enforcement Network (FinCEN) and all four U.S. Federal regulatory bodies recently released a Joint Statement on Innovative Efforts to Combat Money Laundering and Terrorist Financing that not only recommends that banks try AI-based approaches to AML; it essentially guarantees financial institutions that they won’t face regulatory action if AI finds money laundering events that their existing systems were unable to detect!

Where does ThetaRay envision itself in the long run and/or what are its future goals? 

ThetaRay’s ground-breaking unsupervised machine learning technology is the result of over two decades of academic research led by world-renowned mathematicians, Professor Ronald Coifman (Yale University) and Professor Amir Averbuch (Tel Aviv University). Their patented approach has been designed from the ground up to automatically detect meaningful anomalies from massive data volumes with unprecedented accuracy and performance. ThetaRay was launched to transform how the world benefits from big data tackling the most difficult challenges. Originally founded out of Israel, ThetaRay has seen tremendous demand for its solutions on a global scale. Thus, ThetaRay currently has four offices in the US, Israel, UK & Singapore and is providing its solutions and support for each region. Today we are dedicated to helping clients at great financial institutions make giant strides in managing risk. Detecting money laundering schemes, uncovering fraud, exposing loans that are likely to fail, and revealing valuable new customers whose credit scores only tell part of the story. But on the horizon, we see almost limitless potential: to support professionals of every kind, in every industry, make their organizations as resilient as these times require, safeguarding assets, recovering from setbacks, and capturing future growth amid continuous and wrenching change.

Considering the rising number of fraud detection solution providers, how does ThetaRay stand out from its competitors? 

We have four key differentiators:

  • Full transparency: ThetaRay’s machine learning algorithms were intentionally designed from their inception to produce output that is fully supported by easily accessible forensic evidence. To understand why ThetaRay has identified an activity or customer as suspicious, users can click through the full data lineage and see every single transformation that was made to the raw data — including how each anomaly was identified through statistical comparisons to preestablished ‘normal behavior.’
  • High detection rate: Because our technology analyzes massive amounts of data, our detection rate is 5-10x greater than competing soluti
  • Low false positives: The biggest problem with today’s threat detection systems is their staggering false positive rates. In some cases, 99.5% of the alarms generated are false. This creates ‘detection fatigue’ for companies, making it easy for them to overlook real attacks when they take place. Our false positive rate is 10 to 100 times lower than that of competitive soluti
  • Always up to date: Our system doesn’t rely on any rules, so it’s always up to date. It educates itself via deep machine learning, and updates automati

Clientele Assessments 

A multi-national bank that had previously failed to detect its exposure to the Russian Laundromat money-laundering scheme contracted ThetaRay to conduct a review of its correspondent banking activity over the previous six years. The bank was pleasantly surprised by the results of a project they initially believed to be “out-of-reach.” However, ThetaRay algorithms were able to seamlessly parse through 200-million SWIFT messages, which are notorious for their data quality frictions, to illuminate the client’s correspondent-banking black hole. Using SWIFT messages alone, ThetaRay uncovered three new money-laundering patterns contaminating the bank’s business, which they had previously thought they caught all of the money laundering schemes occurring at the bank.

A tier-1 bank engaged ThetaRay to analyze over 12 months of data, including 45-million transactions and over 100,000 business customers, and was jolted by the discovery of five new confirmed money laundering patterns. While we not only detected unknown events, in this Pilot, ThetaRay’s analytics platform detected 100% of the true positives that the bank’s legacy system detected.

About the CEO

Mark Gazit is the founding CEO of ThetaRay and has played a crucial role in growing and guiding the business since its inception. He is one of the top cyber security experts in Israel, with a longstanding reputation dating back to his cyber security service in the Israeli Air Force. Mark is a prominent senior executive with 20 years of experience in Israeli and international high-tech companies. Prior to ThetaRay, he served as Managing Director of Nice Track, which provides software and hardware solutions to government agencies worldwide in the areas of information intelligence and cyber. He was also the Group President & CEO of SkyVision, which he took from the start-up stage to an international company serving over 50 countries worldwide. Mark has held additional pivotal roles in leading companies.

No Comments Yet

Comments are closed