London-based machine learning startup Cytora enables insurers to write more premium and improve loss ratios through data-driven decision-making.
An unprecedented amount of high-resolution data about risk now exists online in unstructured text such as news articles and social media. Cytora has created a way for insurers to tap into this data reservoir and is now working with insurers across Europe and North America to improve risk selection, pricing, and understanding about why losses happen in the first place to help businesses reduce their exposure.
The Cytora Risk Engine, powered by machine learning, captures, connects and delivers billions of data points from the unstructured web, giving underwriters a view of every insurable risk.
Why is unstructured web data valuable for insurers?
Today the opportunity is underpinned by the growing conversions of what happens in the physical world and what is reported on the web. Cytora expects by 2020, 90% of what happens in the physical world would be described in unstructured web data.
Unstructured web data is valuable because:
- It allows insurers to capture significantly more observations about a population
- It allows insurers to micro-segment and target high-value niches where loss occupancy is low relative to the premium opportunity.
Cytora’s data provides an information advantage which can drastically impact underwriting decisions. By using Machine Learning, Cytora connects billions of data points together, assembling vast datasets that humans are unable to generate because of the huge volume and velocity of data on the web. This allows Cytora to capture losses as soon as they are reported online.
Every time they capture a loss, Cytora also connects wider attributes such as sector information, occupancy data and financial data. So, for example, when there is a fire in a hotel, they provide information about the history of loss, location, the owner, the revenues, the number of rooms in the hotel, and other potential rating factors.
Cytora streams all the data into a live data set and runs machine learning against it, allowing them to uncover micro-segments and relate that to the expected loss frequency of that segment.
Insurers can use this data, for example, to quantify the number of losses across insurance lines such as commercial fire and product recall to optimise their underwriting strategy and risk pricing. If the insurer has no loss history, Cytora provides the data required so the insurer can enter a new line and price effectively.
Because they collect so many real-time observations they can identify emerging loss trends and alert insurance companies very rapidly.
Why we selected Cytora for DIA Amsterdam
The worldwide web is the largest dataset about risk. Cytora’s revolutionary technology — based on advanced machine learning — turns billions of pieces of web-based information into structured, factual data describing where risk is distributed in the world.
Cytora’s data produces a disruptive impact for all sorts of information, analytics and decision support companies.
We are pleased they will demo at DIA Amsterdam.
Who are Cytora?
Cytora was spun out of the University of Cambridge and founded in 2014 by Joshua Wallace, Andrzej Czapiewski, Aenenas Wiener and Richard Hartley. With a background in geopolitical and supply chain risk, the 20-person startup now works with global insurance companies across Europe and North America.
They recently raised £2.4 million in Series A funding led by Parkwalk with participation from Cambridge Enterprise, iLexIR and a high-profile group of angels including Alan Morgan, Paul Forster and Matthew Grant, a former executive at RMS. The round builds on earlier investment from Paul Forster, the company’s Investor Director, Co-founder and former CEO of Indeed.
Cytora is also the first start-up ever to receive funding from the Judge Business School at the University of Cambridge, after being incubated at Accelerate Cambridge and they’re member of the Plug and Play accelerator in California.
Aeneas Wiener (CTO), Andrzej Czapiewski (CCO), Richard Hartley (CEO)
“We help businesses gain a comprehensive understanding of risk, particularly in new and emerging markets, by leveraging the observation power of the internet. In the right hands, our data provides an information advantage which can drastically impact underwriting and trading decisions.” Richard Harley, CO-Founder