Good morning Elishba, thanks for taking the time out for a quick corridor chat! So, what do you do in your day to day role?
No problem! I’m the product owner for Pay360’s Optimize suite.
As a product owner, I act as a conduit between our customer facing and development teams so that I can steer my product vision for Optimize. I currently have two scrum teams, one based in Bath and the other in Pune, India. Before that, I had two development teams in Bucharest, Romania.
My days can really vary (mainly depending on the random emails that land into my inbox!) but overall, I work closely with the software developers, testers, UX designer, internal fraud analyst, solution architect and scrum masters to build and improve the Optimize product.
I also spend time answering product-related questions as Optimize is still a fairly new product so sharing knowledge is pretty critical.
What’s your background? Have you always worked in fraud and risk?
So, I studied BA (Hons) in Psychology and Sociology at the University of Westminster which, interestingly, was in the building right behind our Berners Street office in London! I always had huge career aspirations of working in threat intelligence or in cybercrime.
In the summer after graduation, I was recruited by Amazon and interviewed to work in their fraud department just as they were breaking into the UK and European markets. I was trained up from day one on building their fraud risk products and strategy planning. I grew my career there while working on all their newest launches including requirements for Amazon Web Services so I learnt a lot and saw incredible initiatives developed live into the hands of our global users.
When I got the job though - and I always tell this story - my mum was so confused about what exactly I was doing and just kept asking “What’s online cybercrime? What’s fraud? Why don’t you just go and work in a bank in Canary Wharf?” Like most people she didn’t fully comprehend the magnitude of cybercrime and what it meant! Even now, people still struggle to understand what a fraud team does. I find it fascinating, though.
What’s the fraud landscape like at the moment? How have things changed?
We all know there’s still a huge drive for artificial intelligence and machine learning and I was very fortunate that Amazon made that shift early, in 2005. Since then, I’ve had hands-on experience of working in a matrix team where we ran experiments on huge amounts of interesting and quirky data from all over the place. I can still remember the very first time I was being taught about the algorithm ‘Random Forest’ and its importance in anomaly detection for us.
The landscape back then was slightly smoother to navigate. Logging into the dark web to really get close to the more sophisticated crime rings and gathering intelligence to fight cybercrime before it hit Amazon’s borders was so exciting and fun. It was a massive adrenaline rush and one which I’ll never forget!
Fraud is like finding a needle in a haystack. It’s about looking out for the noisy signals in the data and finding associations with hidden pieces, almost like a jigsaw puzzle but with a lot of the pieces missing. Everything is always shifting, so analysis is fast and you’re always trying to discover missing trends in digital behaviour. The fraud landscape is a tricky area and extremely hard to define. Things evolve and develop so fast, so algorithms need to be kept fresh. Today, fraudsters are still organised in crime rings and target companies programmatically. It’s really not about transactional data - that’s very basic for a fraudster - and for a fraud analyst you can’t determine the level of threat from that, you need much more.
So, is Optimize a fraud protection tool or is it more than that? I’ve heard some people say it’s also an onboarding or payments tool…
As the product owner for Optimize and the driver of the product vision, let me say that it’s not just an onboarding tool or a transactional payments tool. It’s a data tool for anomaly detection - that’s the correct term. This means that you can use Optimize to help assess for checks like Know Your Customer and help with Anti-Money Laundering by analysing the behavioural analytics of the customer at the centre of your investigation and by that customers weighted associations. These things all fall under a wider umbrella of what a fraud team would do in their working day. I’m cautious when I use words like anomaly detection and risk mining as I’m mindful of the fraud maturity of our target market. We’ll grow the product as they become closer to using Optimize step by step.
Thanks so much, Elishba. We’d love to find out more about Optimize and risk mining - can we chat to you again soon?
Of course! I really enjoy chatting about fraud management so I’d be happy to give more info and maybe even a retrospective of Optimize next time.