Article

Finding elegant data solutions – without the complications

Jason Crabtree put his military ambitions on hold to help solve the world’s most complex data-centred problems

Growing up on a farm west of Seattle, Jason Crabtree’s life revolved around the outdoors and cattle. It gave him an early education in how to do business.

“I grew up raising and selling registered Angus cattle,” Mr Crabtree says. “I loved it: it was a wonderful chapter in my life.”

Mr Crabtree’s early, “hands-​on” approach on the cattle farm and in construction drew him to the practical field of engineering; election as a Rhodes Scholar took him to Oxford University for postgraduate study.

“My background was in civil engineering, buildings and structures. I always liked this idea that there is opportunity for people to construct and create a better future, a better structure, a better framework in which to interact,” he explains.

However, deeply moved by the 9/​11 attacks, he volunteered for the elite US Military Academy at West Point, where he trained as an engineer and was eventually named the First Captain of the academy’s Corps of Cadets. The next stop was a tour of Afghanistan, leading infantry troops in combat, and then a return to US soil to eventually join the Army Cyber Command in Virginia. There he was tasked with figuring out solutions to some of the most difficult problems, working with the Defense Advanced Research Projects Agency (DARPA), US Cyber Command, and intelligence agencies on problems like preventing or mitigating cybersecurity attacks on key infrastructure such as power grids, banks, and military facilities. For someone who loves simulation and catastrophic risk modelling, it should have been a dream job.

New direction

It was at this juncture that Mr Crabtree’s life took an unexpected turn. As he began to realise the challenges of fixing complex problems amid the rigidity of the military personnel system, he saw new opportunities to have a greater impact beyond his long-​established career plan.

His US Air Force colleague, and former contemporary at Oxford University, Andrew Sellers, felt the same limitations.

“We couldn’t move as fast as we wanted to. We got to the point where we didn’t think that we could build what was missing inside the institutions that we were a part of”

“We couldn’t move as fast as we wanted to. We got to the point where we didn’t think that we could build what was missing inside the institutions that we were a part of,” Mr Crabtree explains.

Even outside the military, the two men couldn’t see any institution they could join that would sustain the necessary focus and specialise in building the tools they needed: a common source for data and a common source for truth. These tools could help them reason about individual risks, as well as “the systemic risk associated with these terrifically important problems where we needed data-​driven solutions to help us avoid unnecessary harm or losses”.

Mr Crabtree and Dr Sellers realised that lasting solutions would require simulation modelling above and beyond data science and machine learning: this was the genesis of QOMPLX. The name reflects the company’s underlying ethos – “intelligence, applied” to solve the world’s most complex problems at the nexus of operations, finance and risk management.

Tackling complexity

Mr Crabtree’s background in large-​scale simulations and AI-​driven optimisation and his collaboration with Dr Sellers – a recognised expert in unstructured data collection and analysis – led to the exciting possibility of “hoovering up and smashing together” the reams of unstructured data generated, he says, leveraging a mix of retrospective statistical and machine learning models alongside generative simulations.  This approach is necessary to evaluate fat-​tailed risks like economic or supply chain disruption during pandemics, or financial contagion where “historical” data is non-​existent or very hard to come by.

“QOMPLX’s goal is to give people a clear understanding of the practical problem scenarios they face in the world and offer products that help address those risks, in a way that doesn’t overcomplicate or obfuscate them,” Mr Crabtree says.

Stephen Hawking said that the 21st century would be the century of complexity. This is precisely what drives Mr Crabtree and Dr Sellers. Having grown to a complement of 200 full-​time staff since its launch in 2015, QOMPLX now offers data handling, analytics and simulation capabilities, as well as advanced applications of its streaming analytics platform, knowledge graphs, and several modelling paradigms for cybersecurity, insurance, finance and general analytics use cases.

A quick scan of recent headlines reveals the magnitude of such challenges. Travelex was held hostage by cyber attackers for almost two weeks recently, forcing its staff to serve customers with pen and paper. A.P. Moller-​Maersk nearly lost its entire global corporate network due to NotPetya Ransomware. Both Wall Street and Main Street banks have been fined billions since the global financial crisis for all manner of unethical behaviour, ranging from collusion to insider trading.

QOMPLX targets such immense problems by taking companies’ structured and unstructured data and making sense of it – at scale. From cybersecurity software for some of the world’s largest technology and finance firms; to underwriting cyber insurance; to time-​series data analysis and telematics; QOMPLX tackles some of the world’s toughest data problems.  Some clients leverage QOMPLX to consolidate terabytes of data from dozens of countries around the world to make near real-​time determinations. Others are rushing to leverage simulation technology after grappling with the limitations of much-​hyped machine-​learning techniques during COVID-​19.

A bright future

Forbes recently named QOMPLX one of the 25 start-​ups that raised the most money in 2019. The tech company announced $78.6 million in equity financing in July last year, led by experienced financial services-​focused investors Cannae Holdings and Motive Partners.

Its commitment to the “intelligence, applied” ethos is evident in its growing pool of patents, with more than 20 issued or allowed between the beginning of 2019 and mid-​2020.

With more than 60 additional US filings pending and a growing global workforce, the company is rapidly filling the gap that Mr Crabtree and Dr Sellers first identified. Its rapid growth to tens of millions in revenue and the quality of the QOMPLX management team are just two of the reasons for Mr Crabtree being named to the Washington Business Journal’s 40 under 40 list for 2020.

QOMPLX’s ultimate goal is to make its technology so cost-​effective that it is available not just for corporations, but the everyday business or common man

QOMPLX’s ultimate goal is to make its technology so cost-​effective that it is available not just for corporations, but the everyday business or common man. It may take time to move from the world’s biggest companies to medium-​sized enterprises and eventually to individual households. For this to happen, we must have clear control of data supply chains, empowering consumers and businesses to regain and maintain trust in the information that impacts their lives.

“We want our services to be the proverbial decision tool for businesses, people and ultimately their families to make good decisions and take actions when it comes to risk,” according to Mr Crabtree.