FinCrime Dynamics’ synthetic data is at the heart of new financial crime vaccines

The recent news that FinCrime Dynamics has received its first investment for its financial crime vaccines raises an obvious question – what is a financial crime vaccine?

FinCrime Dynamics (formally Ealax) is a UK-registered company based at St John’s Innovation Centre. The company was founded in 2019, based on the work of Dr Edgar Lopez-Rojas in computer simulation and cybersecurity.

FinCrime Dynamics’ product, the Synthetizor, is a financial crime vaccine which enables financial institutions to generate their own financial crime simulations with synthetic data, and empowers users to test and improve their financial crime controls.

Synthetic data is information that’s artificially manufactured using a sample of historical real data. It is created algorithmically, and it is used as a stand-in for test datasets of production or operational data, to validate mathematical models and, increasingly, to train machine learning models.

Described as “the regtech breakthrough of 2022”, a financial crime vaccine entails the use of synthetic data as another layer of cybersecurity protection which can be used to supplement existing safeguards.

Traditional data protection until this new development has largely consisted of analysing real time and historical data. A financial crime vaccine analyses the statistical qualities of this real data to build new mechanisms to defend against potential financial crime attacks. Those mechanisms are based on the principles used to develop healthcare vaccines, meaning a substance which provides immunity against one or several diseases, prepared from the causative agent of a disease, and treated to act as an antigen without inducing the disease.

FinCrime Dynamics has developed the technology to enhance the fight against financial crime by enabling financial institutions to improve their compliance controls. It has been built on the foundation of more than 10 years of leading academic research by Dr Lopez-Rojas in financial crime analytics.

Dr Lopez-Rojas, co-founder and CTO of FinCrime Dynamics, is a researcher and entrepreneur. He has a PhD in computer science from BTH in Sweden.

“Over the last 10 years of my research, synthetic data was a real eye-opener that led me to understand that collaboration is a key ingredient in fighting financial crime,” he says.

“It is very exciting that the UK financial regulator, the Financial Conduct Authority (FCA), has made efforts to improve innovation in financial crime compliance using artificial intelligence.

“Synthetic data has played a major part in these through initiatives such as the FCA data sprint, the FCA digital sandbox and the recent FCA ‘call for input’ on synthetic data. Our FinCrime Dynamics’ team contributed to a number of these initiates and we look forward to continuing our engagement and further the concept of financial crime vaccines.”

The Financial Conduct Authority is the UK’s independent, non-governmental regulator that regulates around 58,000 businesses in the UK’s financial services industry. It aims to be a data-led supervisor, but in today’s rapidly evolving financial landscape it needs all the help it can get.

Dr Lopez-Rojas adds: “My 10 years of academic research in the area of fincrime analytics using synthetic data has led to the creation of the financial crime vaccines.

“We are now at the moment where we can make a real impact in the fight against financial crime. It is a very proud moment for our team and everybody that has been involved in this campaign.”

Stephen Quick, CEO and co-founder of FinCrime Dynamics, adds: “We’re a software product company, we allow financial institutions to improve their defences against financial crime. Banks are one of our main customers, plus payment providers and crypto exchanges.

“Our first product is Synthetizor, which is now at the testing stage. We have been conducting user feedback which has had a fantastic reception. Our attention is now focused on a number of software trials with financial institutions.

“Synthetizor is deployed within a financial institution’s environment and integrated with their transaction monitoring systems. This allows institutions to test and improve their financial crime controls. The real power of the solution can be unlocked if they’re using machine learning.

“We use an agent-based model [ABM] which rebuilds the data – we can look at it, learn everything about it and we rebuild it without the client’s data in it, so there’s no client-sensitive data in it. There’s no privacy concerns and we can strip out any bias, cleanse it, and essentially we label the data to visualise and model new financial crime techniques.

“We are able to enrich synthetic data with labelled simulations. This improves machine learning training and explainability, which is then tested on the transaction monitoring systems with the bank or other institution.”

Stephen joined in September 2021.

“I was working in the City, in investment management, then I did a year-long intensive MBA at Cambridge Judge Business School in 2018. From that I got to know people in the start-up ecosystem, including angel investors – and then I went back to work for my old firm, Rathbone Brothers plc, and started working on their digital transformation. Then I got a call from Edgar.

“The company then joined Accelerate Cambridge, the accelerator at Cambridge Judge Business School, and we did a lot of work on finding our story.”

Sounds like time well spent and, with the FCA now involved – “we spoke with them in 2021 and they’ve now come out with their call for input on synthetic data” – the potential market has become more defined. It includes Tier 1 banks, challenger banks, digital banks, payment providers, consultancies, cryptocurrencies and insurers.

FinCrime Vaccines creates synthetic versions of a pattern of disorder in the same way as a pharma company creates a vaccine which uses a microcosm of the virus to activate the body’s defences.

“Humanity started using technology to fight disease and now we’re using it to fight financial crime.”

There are no competitors, says Stephen: “This is a niche solution we’ve created.”

“The user can use Synthetizor to test their solutions – it’s not an oversight tool, it’s used to test those oversight tools.”

The company is so ahead of the curve it almost built the curve, but that’s all about to change.

“Sixty per cent of financial institutions are expected to use machine learning to combat crime in the next two years. Gartner says 60 per cent of that machine learning will be based on synthetic data, not real data, by 2024.”