By using data science technology and intelligence tools from Mango Solutions, UK-based not-for-profit fraud prevention membership organisation Cifas has been able to discover hidden patterns that are almost impossible for human beings to detect. In keeping up with evolution in fraud prevention, Cifas can minimise risk for members and individuals.
Cifas is a not-for-profit fraud prevention membership organisation. Based in the UK, it manages the largest database of instances of fraudulent conduct in the country. Members are organisations from all sectors, securely sharing their data to reduce instances of fraud and financial crime. Since 1988, Cifas has helped its members and customers to protect themselves from over £14bn worth of fraud losses.
The challenge
Cases of fraud and financial crime have reached epidemic proportions. According to research and advisory group UK Finance, the sector prevented £1.6bn of unauthorised fraud losses in 2020, the equivalent of £6.73 in every £10 attempted.
Cifas helps individuals, public sector and private sector organisations by collaborating to create a non-competitive fraud prevention environment. Cifas members reduce their fraud risk by checking details of applicants and customers against known fraudulent conduct. Cifas Intelligence Analysts also search for patterns in the data to alert members to emerging trends and changing behaviour.
To accelerate that capability, Cifas needed a way for members to be able to focus on the most meaningful matches and weed out false positives: that is, those matches that don’t result in uncovering fraudulent conduct. Cifas also wanted to uncover those emerging patterns more efficiently and effectively in order to help members reduce the harm caused by earlier fraud.
“We were providing a very relevant and useful service, but we wanted to go further to unlock the power of the data and speed up members’ ability to identify threats and risk factors,” said Sandra Peaston, Director of Research and Development at Cifas.
“Through improving our data quality and adding more intelligence tools, we were aiming to eliminate more ‘casual’ fraudulent conduct and be able to address more organised, more sophisticated fraud groups which are a significant threat to members and individuals.”
Vendor selection
To offer members and individuals the best protection against fraud, it was necessary for Cifas to select a vendor which was well-respected for its data science tools and deep expertise in this sector. The deployment was required to deepen Cifas’ intelligence capabilities as well as further broaden its understanding of crime networks involved in fraud.
Mango Solutions was chosen for the consultancy’s established and proven data science methodology with a long-term focus on prioritising tangible value aligned to business challenges. The partnership with Cifas would deliver strategic advice as well as innovative tactical solutions to give data-driven insight within the organisation.
With enhanced data quality, Cifas would reduce false positives to make members’ lives easier and give better data for intelligence. Using technology to provide an enhanced matching service to members would enable scoring on the certainty and severity of fraudulent activity.
The solution
In 2020, Cifas began exploratory talks with Mango Solutions to explore potential synergies and ways to make Cifas even more effective. In August of that year, the two companies announced a partnership under the terms of which Mango’s data science tools and expertise would be deployed to deepen Cifas’ intelligence capabilities and deepen its understanding of crime networks involved in fraud.
Peaston said: “The investigation stage enabled Cifas to focus on core areas of opportunity, to build effective stakeholder relationships, while enabling internal empowerment. We realised that the agile way in which Mango works would be essential, to enable us to be more proactive in dealing with environmental conditions.”
Specifically, Mango is applying data science to refine searches and create optimised results that let members further protect themselves and see rapidly when there are risks and trends emerging of certain fraud types or activities. Cifas is adding more intelligence tools rapidly uncovering the valuable intelligence hidden in the data that will allow members to nip emerging fraud trends in the bud.
“Fraudulent activities constantly morph and evolve,” said Peaston. “We needed to develop our offer at least as fast as attackers are moving to increase their threat capacity. Working with Mango to apply predictive data science to that challenge means we can provide an advanced, self-learning augmented platform that cuts through faster than we would otherwise have been able to offer. Cifas can also be more confident in its ability to advise on high-level anti-fraud strategy through our enhanced data and intelligence capability.”
Rich Pugh, Co-founder and Chief Data Scientist at Mango, said: “It was a perfect meeting of minds where two complementary capabilities could come together. We are augmenting expert knowledge by providing a combination of human insight and knowledge and what machines excel at – processing large data sets without errors.”
The approach
Using Mango’s established data science methodology, Cifas has prioritised extracting tangible value aligned to business challenges.
Mango advised Cifas which parts of intelligence should be automated, added in more intelligence tools for members in two critical areas. First, to avoid patterns that could be picked up by algorithms – this was done horizontally – using different data, and different types of patterns. Then vertically by moving from descriptive to diagnostic, predictive and potentially prescriptive analytics. To stay ahead of perpetrators who are continually gaining in sophisticated techniques and adapting their behaviour to break into the information.
Cifas is increasingly able to offer members an easy-to-use service for detecting patterns of activity with high levels of confidence in results. The smooth transition from the investigative phase has enabled Cifas to focus on core areas of opportunity, whilst enabling internal empowerment.
The key benefits realised since implementation
Together, Cifas and Mango have developed a best-practice framework and ensured a smooth transition from building valuable stakeholder engagement to hard results. The benefits of applying the combination of data science and intelligence tools have been to reveal hidden patterns that are almost impossible for human beings to detect.
Specifically, Mango successfully created a probabilistic match engine with member data.
Cifas is applying data science to refine searches and create optimised results that let members further protect themselves and see rapidly when there are risks and trends emerging of certain fraud types or activities and support law enforcement agencies.
Peaston said: “Through Mango’s data science consultancy, we’ve been relieved of the burden of manually trawling through reams of charts and data manipulation. In replacing this with giving our technical experts relevant and timely information at their fingertips to help identify current and emerging fraud threats. This has been an enormous influence on Cifas approach and decision-making in fraud prevention and strengthened the organisation’s position as a compelling thought leader in the field of fraud.”
A key benefit has been the perfect blend of expert human skills and technology within Cifas to offer optimal fraud prevention to its stakeholders. Rather than replacing employees with machines, Mango’s approach augments their knowledge with data insight, in tandem with automation for manual tasks. This ideal combination of the strength of people and technology has the potential to process huge amounts of data and find quantitative trends and patterns while human experts can make sense of multiple unstructured bits of information and evaluate the bigger picture.
Cifas can now utilise intelligence more effectively and efficiently to inform its members and the wider fraud prevention community. In keeping up with the evolution in fraud prevention, Cifas can minimise risk for members and individuals.
What’s next?
Cifas and Mango will continue to innovate and use data science to unlock fraud and e-crime, refining algorithms over time to become even more effective in countering criminal activity and finding ways to stay ahead of malicious actors. Together, they are seeking to help automate processes to better protect members and secure their data assets.