Analogue vs. AI - Which one is better?

Fraud detection has come a long way from the days of detectives pounding the pavement, zipping all over town from clue to clue to piece together the jigsaw puzzle of the crime in question (thank you Inspector Gadget).

Analogue methods do have a place in the mix but not as the main approach to find the errors and fraud initially. AI on the other hand can crunch the numbers in huge volumes, finding patterns and unusual entries faster than you can say “go-go-gadget”.

The problem with AI is that it's still not perfect. It relies on the quality of the data that we feed it and the algorithms it uses. And then you have a variable factor of the fraudsters. They are a clever bunch of highly organised people who are always testing the systems and looking for new ways to get their foot in the door.

Now this is where VigilantPay weighs in. Rather than you constantly needing to be sifting through the haystack looking for needles. VigilantPay highlights and points you to where the needles are and only when you have to do something with them.
How much time would that save you per year? 

So the best approach, we think, is to use AI to find the patterns and errors, and real people (Analogue) to do the deeper digging and investigation once the patterns have been surfaced. So Analogue and AI don’t work against each other, they actually are holding hands skipping down the street together looking for the next big threat.