Andrius Sutas, CEO and co-founder of Aimbrain reveals his highlights from 2018 and the challenges ahead in 2019.
Tell us about how your business has grown and evolved over the past year.
2018 has been a terrific year for AimBrain. We’ve built and added more modules to our solution suite; being server-side and API-based, it’s easy for organisations to select and configure any of the modules to suit their range of business cases. We also launched AimFace//LipSync this year. The module was first-to-market, a combination of facial authentication with a randomised voice challenge and lip synchronisation analysis. As spoofing fraud gets more advanced, protection needs to evolve but still be accessible to all. Our solutions are built to work on any device, not just the latest technologies. We also opened up our platform – AimBrain On Demand – to developers this year, and have seen nearly 100 new applications now connected and trialling biometric authentication ahead of the commercial conversations. Finally, it’s been the year for external recognition too – being recognised as a Gartner Cool Vendor in IAM and being recognised personally – with co-founder Alesis – on the Forbes 30 Under 30 list.
What’s the most exciting innovation you have seen to date in biometric technology?
We’re seeing incredible results from the large-scale testing of our AimAnomaly Detection module, being able to identify and stop the fraud that is currently eluding even the most advanced teams. This means that behavioural biometrics are being used to identify fraud before a user’s behavioural profile is even known. I believe that behavioural authentication is going to underpin so many applications in 2019 onwards, and it’s exciting to not just have a proven solution in that space, but to also have a range of active authentication steps to draw upon too.
Where are you seeing the greatest demand for biometrics?
Like most, our first clients were in financial services and this is still a natural alignment. However with AimBrain On Demand, we’re seeing all types of organisations registering, from individual application designers to global technology and infrastructure organisations. With so many breaches in 2018, it’s no surprise that ditching passwords for stronger authentication is extending way beyond financial services.
What are the challenges that you see in 2019?
Fraud is getting smarter and smarter, as we can see by the evolution of spoofing fraud and deepfake for example. Also, annotated data (labelled data, for instance examples of onboarding records known to be fraudulent) can’t be scaled, so we expect to see a lot more effort put into unsupervised learning (where algorithms draw inferences from datasets comprising input data without labels), self-supervised learning (a form of unsupervised learning in which some part of the data is withheld and models trained to predict) and transfer learning (one approach in machine learning where models are trained on one task and then reused as a starting point for a different task usually harder to learn, or has less data available). So there are challenges, but opportunities within them for vendors to shine!