Most frequent use cases our Zoe.AI tool addresses

Usually, in my posts, I focus on a specific use case and focus on its detail. This time I decided to give you an overview of the most frequent use cases and business issues where Zoe.AI can help.

Sure, the most common topic of my posts, I've written about it many times. But can you tell me all the ways of usage of Zoe.ai in the world of online lending? No? Well, let me recap them!

Identifying suspicious behavior

You received an online loan application from a client who appears to be okay, taking into account all available "hard" parameters. But you probably didn't notice - because you have no way how to find out - the fact that when applying for the loan, he changed his monthly income three times, his expenditure four times, but his marital status and education level too?

Well, you should know this! Because our statistics show that such a client is then linked to a proven fraud or default in 50% of all similar cases.

Fraud/default detection

See above for an example, but not only that. We can also detect a specific type of device or network the client is accessing from - e.g. VPN, TOR, etc. Thus, thanks to device fingerprinting, we are able to recognize those who try repeatedly, typically using the personal data of their relatives - which is quite common.

Client segmentation

Who to target your advertising at, or who to leave alone? Who is more interested in arranging some specific product and who is just looking around? Zoe.AI tells you everything you need to answer those questions and therefore invest your time and money efficiently.

Let me divide this vertical into two different processes:

1. Arranging the insurance product

If you want to optimize the online form workflow performance to have the highest possible conversion with minimum risk, you need to:

• segment by interest segment by price sensitivity - use dynamic pricing

Device fingerprint technology together with biometrics tracking enables you to predict a real user’s interest and to track credibility and scale pricing.

Moreover, you can also learn detailed aspects of the user's behavior and make use of them in the future in case any claim occurs.

2. Claims

As you have already learned, Zoe.AI is very strong in identifying a suspicious intention. Just imagine: Your client had a car accident. In the online claim registration process, he changed the description of the insured event twice together with the location where it had happened. Strange, isn’t it?

Do you - in your e-shop - recommend products online using any type of already existing engine? Probably yes, it’s nothing new.

Do you base the recommendation on what people frequently click on and what they usually buy together with items in question? Interesting, but common.

Or do you use also users’ device fingerprints and behavioral biometrics? The behavioral similarity of various users and their preferences entering the model?

You don’t? It is the right time to give it a try. And that's exactly what we've taught Zoe.AI. To easily identify specific preferences of not only entire groups, but every single user.

Thanks to this, we are much more accurate in showing the product they are most likely to be interested in.

Have you recognized your business issues in any of the use cases above? Well, then what are you waiting for? It is really easy to start a “proof of concept”!

Zoe.AI data can also show you that:

• an abnormally large number of users is leaving after some specific form step
• for no obvious reason some user anomaly quickly went through the entire
• workflow - probably a bot? an autocomplete is not working well the layout of
• form elements is not okay

Basically, Zoe.AI is an easy tool to optimize any online form or a whole website in any imaginable way!