Blog

Deep dive into our all-in-one identity decisioning platform

Product.png

Our 1-to-Many Approach to Identity Decisioning

Identity decisioning can seem like a huge, top-of-mind burden for most product owners out there taking onboarding or risk-related products. The stress of having to go out there to find a different vendor for practically each screen of your journey, to then having to hold commercial, legal, technical and cultural conversations for integrations that seem never ending.... What a pain!

But once all those conversations, negotiations, integrations and internal toolings are done (are they ever?), you hold your head high and launch your onboarding. You give it some time to soak in and get some volume flowing to start measuring performance.

And then…one morning you pour your daily cup of coffee, and open your laptop, just to find fire on Slack. The Growth team is complaining because their conversion rates are on the floor, the Risk team is finding multiple fraud incidences, and Customer Success is swamped with customer tickets on this “new onboarding screens that don’t work.”

That doesn’t really add up, this isn’t what was promised by the different data vendors in those sales meetings….

You rapidly shift your priorities and head to your dashboards with your data and engineering team to figure this out, and before you know it, you realize you’ve created a frankenstein! And it looks kind of this way:

Neobank entangled KYC architecture


If this seems familiar to you or you’ve heard a similar horror story, you are not alone. These “common practices” force teams to build complex in-house decision models in order to be KYC-compliant and mitigate some risk. The crucial problem with these individual solutions is that no aggregation is happening to consolidate a 360-degree view of your customer’s identity. Individual boolean validations (Yes or No) end up raising your false negatives with a thin risk threshold, causing ending up with low conversions rates and high fraud incidents. If someone’s digital identity is a jigsaw puzzle with multiple pieces, the 1-to-1 solutions fail at assembling the puzzle together before making a definitive decision on the person behind the pieces.

We believe that to comprehensively solve this problem one should not verify User Attributes (identity elements such as email, phone, address, etc) individually but rather as a whole, gaining a holistic understanding of the User. Instead of evaluating user attribute independently, as product owners focusing on the the big picture of our Users is more productive and here is were these three important questions come to play:

  1. Does the person exist?
  2. Is the person who he says he is?
  3. Can you do business with that person?

Zenpli’s 1-to-Many Approach

To solve this challenge we created the Onboarding API, a single integration identity decisioning solution with a multi-layered approach. This API is capable of looking beyond the conventional single User Attribute <> boolean validation (1-to-1), but rather leveraging Multi-User Attributes based on advanced correlations and analytics to gain conviction on a User’s authenticity (a 1-to-Many approach).

For example, the Onboarding API not only verifies a phone number's existence, but also looks for connections and correlations with Email, Name, Device, IP networks, Social Media, and so on, for all provided User Attributes. By taking this approach, we make highly accurate predictions that all information provided to your business belongs to an organic / real user, and not a fraudster.

We’ve invested heavily in feature engineering, curating the best data sources (online and offline) in diverse markets in LatAm and finding connections across the different User Attributes. Some examples of the features that we have developed to better understand User Attributes’ 1-to-many connections from a customer’s phone number include:

  • Phone number is landline or disposable;
  • Phone account inactive with mobile carrier for more than 60 days;
  • Address user attribute and phone address distance out of range;
  • Email not correlated to phone;
  • No phone record for main messaging platforms;
  • Low phone incidences on consulted databases

Single API, Multiple Solutions

Whether you're in the fintech, insurtech, autotech, proptech or any other n-tech market we’ve got you covered. From account opening to ongoing monitoring, you’ve got full KYC compliance and fraud prevention mechanisms with a single integration.