Chapter contents

The fraud threatscape is constantly evolving. Airlines need a new approach to stay ahead. The answer is in using all available data to reach more accurate decisions more quickly. Move beyond manual review. Meet Ravelin.

Ravelin dives deep to uncover more fraud

Ravelin is a single system which pulls in customers and payment data from across your business. From sign-up, source of the sign-up, account creation, route selection, on-site behaviour, through to the booking and payment process - if it’s relevant for fraud, Ravelin analyses it. 

Our enterprise-ready and travel-specific machine learning models aggregate huge volumes of disparate data and use it to provide a straightforward fraud decision right at the time you need it.


With Ravelin everything is automatic, allowing us to grow our user base without fear of fraud." Zalan Lima, Head of Fraud at Easy Taxi

Ravelin: perfect for the fraud challenges of airlines 

Our data-first approach means that we look deeply into your data and your experience of fraud. Fraud in online travel booking shares  many characteristics with other industries - e.g. card velocity, the number of attempts to pay, multiple cards used. 

There are also trends unique to the industry that are strong signals of attempted fraud, such as overlapping journeys, one-way journeys, new accounts attempting same day travel, and flights cancelled in order to gain a credit. 

Our machine learning models are trained on your company’s historical transaction data and adapted to your fraud and domain experience. Powered by your data, this allows us to provide accurate decisions from Day 1. The result? Fewer chargebacks and fewer false positives. 

Screen Shot 2017-01-30 at 17.51.31.png

Fast decisions

Ravelin provides results in milliseconds and updates its scores whenever a customer makes a relevant action. It’s real-time fraud detection.

Round the clock fraud scoring

Our models never sleep. Ravelin is just as effective at night, at weekends and even during peak season. Our clients trust us to make the right calls to optimise conversion without any undue risk of fraud. No more reliance on manual review.

Everything in one place 

One system for all your fraud needs. With ML models, link analysis, graph networks, analyst interface, rules, workflows and reporting, there is no need for multiple systems to provide a fraud determination. One system: enterprise-wide.


"Great support during integration, gthey set up a direct communication channel which helped us with any queries we had. The API documentation was easy to follow, well-documented and full of examples on how to use the API." Ian Vaughan, Development Team Lead

Instant results

Our experience in a variety of industries means our algorithms give accurate scores as soon as the integration is complete. No cold-start problem. At the same time, our graph networks map your entire customer database highlighting and closing down bad actors from day one. 

Built for analysts

We’re confident your analysts will love the Ravelin dashboard. They can access full customer history in an easy to interrogate timeline and so instantly understand what underpins a score and confirm its accuracy. Our workflows will blend seamlessly into theirs making the entire review process more efficient.  

Reduced manual review

The most notable impact of Ravelin is the the massively reduced requirement for manual review. We would expect a maximum of 1% of transactions to be reviewed, usually retrospectively. Time is spent so more more effectively on investigating trends and devising tactics to prevent fraud.

Management and control

You are in control from the initial definition of the thresholds, to configuring rules, defining analyst workflows and deciding on what reports are required and how often.