BlockFrauds helps insurers tackle insurance fraud more efficiently and effectively, helping cut $billions of costs that result in higher premiums which prevent the most vulnerable people getting their protection.
As part of their journey on BlockStart’s Acceleration Programme, as a startup developer of blockchain technology, they were looking to further develop their solution and for new opportunities for collaboration with end-user SMEs that could validate and help them improve their “Blockchain of Things Devices” blockchain solution.
We spoke with the CEO Roanna Doe to know more about Blockfrauds and their participation in BlockStart.
In one paragraph, what is Blockfrauds?
BlockFrauds is a B2B SaaS business that helps insurers detect fraud more efficiently and effectively, whether in automated processes or direct customer contact, thus helping to save billions of dollars each year which otherwise get passed on as higher premiums.
Can you tell us how the idea for your startup began?
Soadad had a vision for enabling insurers to streamline their workflows and create personalised value by helping create an ecosystem of their data and tools and applying intelligent automated processing. I helped focus this on valuable use-cases. Initially we got traction with an exposure management solution, partnering with a key industry player to prepare an MVP for multi-party trials planned for winter 2020/21, when so many other things also got delayed.
During those delays we entered R3’s Corda Insurtech Challenge 2020: Fraud Detection, sponsored by B3i – a major insurance industry blockchain consortium – and with a goal of “helping insurers privately and securely share data on fraudulent activity and work together to reduce it.” There are two main challenges there: insurers handle sensitive information and are tightly regulated over its use, such as with GDPR here in Europe, and of course they don’t want to be sharing their customer details too freely with their competitors. We were runners-up and fired with enthusiasm for tackling this crime, so it is now our main focus!
What is Blockfrauds passionate about achieving?
BlockFrauds is passionate about helping insurers detect fraud efficiently and effectively so they can reduce the hundreds of billions of dollars a year in fraudulent payouts and costs of tackling it. This is because those costs get passed on to customers as higher premiums, meaning billions of the most vulnerable people worldwide cannot afford the protections they provide whilst the rest of us get worse service and value with all the checking. We all want to look after our families and our health, and we all need the security that our jobs and businesses and assets provide, and BlockFrauds is fighting hard to help more people get that protection from insurance.
Who are the team members and what backgrounds do you have?
Soadad Farhan is co-founder and CTO. He has over 30 years’ experience in software development and management roles, including for Siebel Systems (Oracle) and various financial institutions. He was most recently CTO and Head of Innovation at Hiscox, a major UK insurer, which stimulated his vision for Fardoe Software / BlockFrauds.
My name is Roanna Doe and I am co-founder and CEO. After Oxford University I qualified as a chartered accountant with Deloittes (London), then had over 20 years’ FTSE 100 experience in a range of senior strategy, M&A, management, finance and business development roles worldwide.
We both live in London, UK, and you can see we have the complementary skills to help deliver our vision!
Can you tell us about the prototype you have developed with the support of BlockStart Acceleration Programme?
It is focused on three main areas: better detection of fraud, enabling collaboration for even better detection, and providing technology that can keep pace with fraudsters as their tactics evolve.
First, we help insurers create an ecosystem of their data and tools, for maximum detection opportunities. These contacts are light-touch in the prototype but the design is highly connective so will easily expand. Then we apply a range of cutting edge technologies, including AI and our proprietary algorithms, to further detect fraud. We have two initial focuses in the prototype. Digital Speech DNA is advanced speech analytics, which not only detects imposters but also suspicious changes in the way the customer speaks, or suspicious characteristics in their speech. Digital Image DNA can flag if an image has been used elsewhere in the ecosystem, even if it’s been inverted or put through a filter, or seems to show the same subject from a different perspective. This flags an initial Claim Credibility Score to the Claims Handler, so they can prioritise which claims to review.
This processing has anonymised the data, so this intelligence can then be shared seamlessly and compliantly on a private, permissioned blockchain. The prototype is built on a Corda sandpit, and only trusted participants are allowed to share intelligence and train the model. It holds a single, immutable source of truth of those Digital DNAs, so it acts like a Fraud Bureau. This gives greater chances of spotting known fraudsters and multiple claims, and the central federated learning can develop faster and more broadly from that wider intelligence sharing, and is better able to keep pace with the fraudster evolution. This all results in a more informed Claim Credibility Score for the Claims Handler, and all network participants get an updated algorithm for their private AWS node so their initial scores are better informed also.
What were you most excited about in taking part of BlockStart Acceleration Programme?
Lots of things! It is always valuable to get feedback and support from experts, and BlockStart has such an extensive and helpful network of mentors. It was also a great opportunity to consider SME engagement. We always think of these huge insurance companies, and for sure they will be a key part of this joint fight against fraud. But there are an increasing number of small insurers – often underwritten by larger companies – offering a range of great new products such as pay-per-mile car insurance. They are less able to access information on frauds and fraudsters so could be a target, and BlockStart really helped us look at this segment too and see how we can bring the whole sector together in the fight.
Which were the major challenges you have faced during the development and market/fit validation of your prototype?
A nice problem to have is that there are now so many insurtechs out there, with so many great ideas, that it can be difficult to make your voice heard. BlockStart has helped in that regard, as indeed have our investors and ongoing support from R3.
What unique aspects does your company and the prototype you have developed have to offer to the market sectors/type of companies you wish to enter/sell and to the blockchain ecosystem in Europe, as a whole?
We offer seamless, compliant sharing of intelligence and central federated learning which has several benefits for the insurance sector. It helps detect multiple claims and the activities of fraudsters across the sector; it helps the AI learn most quickly and widely; the seamless sharing really helps with the efficiency and speed; and it meets regulatory and compliance requirements. Other sharing solutions are typically much more limited and with onerous follow-up rather than being seamless. Very few companies are offering our kind of advanced speech analytics, especially with a focus and integration into insurance workflows, or our kind of advanced image analytics. And it’s a great use-case for blockchain which should really help its take-up across Europe and hopefully the world, by both large and small players.
Is collaboration with SME adopters important for piloting your solution?
Yes, especially with Claim Technology as an adopter partner. They are able to move fast, provide a wide range of insights including across the SME space, and are a route to gaining those all-important first adopters including SMEs. With them we will most easily be able to demonstrate the functionality and thus give confidence to other and larger companies to trial.
Where do you see your company in 5 years?
Present in many regions around the world with hundreds of insurers able to join forces with us to fight fraud. We will have leveraged pilot success into an MVP launch and roll-out, expanding the language functionality from English into the major European languages and several others besides, and with a plan to add more. We will have out-of-the-box connectivity for many more data sources and tools, which alongside our Digital Speech DNA and Digital Image DNA will be detecting fraud at substantially higher rates and far more efficiently than at present. We will have identified solutions for additional unmet needs and be launching those. We will be seen as a force for good, helping fight crime and enabling more people to afford the protection of insurance.
