Socure's Sigma Solution: A Unified Approach to Identity Verification

When we started the company, we started with using social media data to authenticate users on the web. The reality is that we had to shift, and in that shift, we realized that there was a big need for solving identity fraud at account opening. When we started looking at what we could really do at account opening, we started customizing identity verification solutions.

And the very first model that we’ve ever built was called the Fraud Score. The concept of Sigma came about once we were able to build several clients into production. And we were able to step back and understand what we learned from each of the customers and then aggregate that into the solution today called Sigma.

Sigma stands for summation and aggregation. Consider this concept: if we could aggregate similar clients or industry specific clients, let’s say five different credit card clients into a single model, because of the same identity fraud patterns, we could then deliver and package a solution that every single client or every single credit card could use at the time. And that is the birth of Sigma.

So the traditional way today that people have tried to solve for identity fraud is by stacking a series of individual point solutions in a large kind of orchestration rules engine. Socure has taken a different approach, which we’ve built. Each of the individual elements of digital identity, such as email, phone, device, geolocation, behavioral, all combined into a single unified identity verification solution. We developed tens of thousands of variables across the entire identity, not individual point solutions, which gives us a much more comprehensive view of a singular identity decision.

One of the most important aspects of all of our solutions here at Socure is that every single client that uses Socure provides us performance feedback data. And that performance feedback data, the aggregation, the network, is really what allows us to get more accurate over time.