Blog / machine learning
Why we have our Investigations team
We often get asked questions about our Investigations team - who’s in the team and what do they do? We spoke to Ruth, our Head of Investigations to get all the answers...
What’s the role of the Investigations team at Ravelin?
I’d describe it as a blend of talking with clients, fraud investigation, and data science. We work closely with our clients to help them solve issues and we’re also nerds at heart - we like crunching numbers and extracting actionable insights. I sometimes say our role is part data analyst, part fraud therapist.
As well as helping clients with what’s happening in their business now, we also act as a funnel for all client requests to the product and machine learning detection teams. We’re the voice of our clients inside Ravelin, we distill their concerns, ideas and hopes for what Ravelin can do for them and act as their advocate to get things done. This is a key part of our role and it means that we often see clients having a tangible impact on a Ravelin product within a few weeks or months - that’s very rare in this industry.
What’s the typical background of the team?
There’s no typical background within the team. We have people with backgrounds in military intelligence, police intelligence analysis, private investigations, academia (economics and cognitive neuroscience), data science, business development, marketing and management consultancy.
Having a diverse range of skills and experiences is a huge benefit for us. Team members with a police/intelligence background have unique insights into fraudster tradecrafts and specialist experience investigating fraud rings and other organized crime. The balance between commercial roles and intelligence/police experience gives us all access to different perspectives as we work together. People with academic backgrounds have teaching experience which means they are great at helping clients, answering queries and providing analytical support.
How many clients do you work with at once?
Each analyst looks after a handful of clients at a time. This means they build up a great relationship, and the analyst really knows their client’s business data very well. This also means each client can have daily conversations with the analyst when they need to. Every business faces slightly different fraud, and so each client has different levels of interaction with their analyst. Depending on the business, some clients speak to us every day, some have fewer requests and interactions
What does your team do in an average day?
The average day in the life of an Investigations Analyst at Ravelin would be:
- Answering client questions
- Querying and analysing data
- Monitoring chargeback rates and overall performance
- Improving performance - score threshold and rule analysis, model improvements
- Uncovering integration / data quality issues or bugs
- Outwitting fraudsters - working out which loopholes they’re trying to exploit, and counteracting them with new features or rules
Regular activities we do weekly or monthly:
- Catching up with clients in weekly / monthly calls
- Visiting clients in person in the UK or abroad
- Dashboard training sessions with clients, either in person or via video call
- Meetings with product, detection or integrations to discuss new clients, product improvements, performance enhancements
What have you found most surprising in your role?
I’ve been surprised by how easy it is to commit fraud, how many tutorials there are available out there. This is because fraudsters communicate and share knowledge on an incredible scale on online forums.
This teaches us that it’s so important for us to share knowledge. We can’t keep any information to ourselves, as the other side certainly isn’t. This doesn’t mean sharing any client data - but as a team we share what we’ve been working on and what fraudulent activity we’ve seen. In fact, when we see an incident or notice something unusual on one client’s data we check for the same signs on other clients and we’ve caught a lot of issues by doing this.
I’ve also been surprised just how ingenious fraudsters can be at the highest level. I’ve watched them change their behavior to adapt to new rules and features. We’ve seen new types of fraud emerge, such as supplier-side fraud, to get around advances in fraud detection on the consumer side - for example instances of suppliers in taxi companies or food delivery services repeatedly placing orders and cancelling them to get a cancellation fee, or creating fake customer accounts and placing orders with themselves to get commission.
What do you like best about working in investigations?
The fact we get to help our clients is the best part of working in the Investigations team. We’re often talking to someone about a problem which is really stressful for them and it’s their job on the line if they can’t find a solution. It’s hugely rewarding to be able to lend our experience with data and fraud in different markets to help them find that solution.
I really enjoy having a role which helps to improve the product. Lots of new product ideas come from client requests that we pass to the Product, Engineering and Detection teams, and it’s great to watch the idea transform and develop and get launched. I also love getting to grips with the data and outwitting the fraudsters. When we find that they have found a new way of getting through, we work quickly to make a solution. It’s like a constant game of cat and mouse.