top of page

How Kredivo Prevent Blacklist Customer with Identifai

Diperbarui: 23 Jun 2022

Introduction

Though fraud can be a multi-billion dollar problem for Kredivo, it doesn't have to negatively impact their business. Blacklist Identification technology can help Kredivo detect users with high risk scores and prevent them from gaining access to Kredivo platform.


As the first Computer Vision company in Indonesia, Kredivo sees that the product maturity of Nodeflux through IdentifAI is appropriate and able to detect face attributes when customers verify their faces.


Through IdentifAI Face Recognition, Kredivo can improve the onboarding process without being constrained by error detection due to the use of hijab or other facial accessories.



The Story of Kredivo

Kredivo is one of the online loan service providers that has been established since 2018. The loan services provided are not only limited to cash, but various purchases from marketplaces that have collaborated with Kredivo. As a service differentiation and advantage, Kredivo focuses on accelerating the disbursement of funds or the registration process that prospective customers need to do.


Consequently, with the speed point that is put forward, the services provided by Kredivo do not tolerate errors in disbursing funds at all, or can be said to be zero mistakes for every transaction submitted or executed.


Not only individual loans, Kredivo also offers loans to merchants or MSMEs operating in Indonesia. Broadly speaking, Kredivo targets the underbanked and underserved segments which are not facilitated by conventional banks due to the long SOP process or not covered by the Bank's services.



The Challenge in acquiring Customers Through Digital Platforms Before Using IdentifAI

In the next five years, Kredivo targets the growth of 20 million subscribers in Southeast Asia. Especially in Indonesia, where the Muslim population dominates, Kredivo has to focus on onboarding without any obstacles to face attributes, which in this case is the hijab.


This smooth onboarding not only supports the achievement of the targets above, but also as a preventive measure to select customers who are not eligible to receive funding due to poor credit history or the equivalent amount of credit that can be channeled to each customer.


Blacklist identification is a very crucial process for Kredivo if we look at the target market segment, which are underserved and underbanked. Geographically, this segment comes from rural areas and has a lower middle SES that may not have a credit history or delayed payment settlement processes. So it can be said that there are two important challenges that Kredivo has. First, they must be swift in selecting potential customers to determine whether they are entitled to receive funds or how much funds are received. Second, the speed of the procedure is important for Kredivo to be able to stand out from the competition and create the impression of "convenience" in the minds of the Indonesian people.



Why Kredivo Chose IdentifAI as e-KYC Platform for Digital Onboarding

Prior to collaborating with Nodeflux, Kredivo had used other e-KYC platforms originating from Indonesia. However, errors and engine capabilities in detecting hijab are still far from Kredivo's expectations. As a result, the onboarding and blacklist identification processes are hampered. Of course, this is a big homework for the Kredivo business team and developers. Thus, Kredivo conducted a research regarding the e-KYC platform that has good product maturity so that it is able to answer the problems being faced.


In mid-2021, Kredivo contacted Nodeflux to try out IdentifAI. After doing POC, Kredivo is interested in using Identifai Face Reconition and Face Enrollment Service. This selection is based entirely on the dataset owned by IdentifAI in detecting various face attributes, especially hijab and glasses, so that the accuracy possessed can precisely provide output that is in accordance with the real conditions of potential customers whether or not it is feasible to get a loan from Kredivo.



How Kredivo Used identifAI for Blacklist Identification Needs

Kredivo implements Face Recognition to check the person is on their blacklist database. The system takes an unknown face and compares it to a large database of known blacklist faces to determine the unknown person’s identity is on the blacklist or not. By using Face Recognition technology, the verification system can identify most similar face candidates to be checked as the blacklist suspects.


The process start from user blacklisted face enrolled on the database as face embedding. When the user doing verification, the selfie will be compared on the face embedding database to validate the user is not blacklisted person.



The Results

Kredivo's total hits using IdentifAI have reached 1.3 million hits. This number will continue to grow in line with the targets that have been set every year on Kredivo's side. In addition, the onboarding process for customers who use face attributes is no longer a problem, because IdentifAI can detect it properly.


In the midst of the current competition for new customer acquisition, determining the right potential customers is very important. We don't want our customers to have a bad credit history that interferes with ongoing business processes.


Therefore, in order to minimize this risk, you need to pay attention to several aspects in choosing an e-KYC platform, including:

(1) how reliable the SLA is given, (2) has there been any application of XAI in analytic liveness? (3) How robust is the analytics in detecting various face attributes used by customers? These aspects are already present in IdentifAI. It's not only product maturity that we have, but as a native Indonesian company, it allows us to provide the best service very quickly and precisely whenever you need it.


Come visit the following page to find out more about how identifAI can help you improve Blacklist Identification Process and mitigate fraud here


If you have any questions, you can also contact us through whatsapp


0 tampilan0 komentar
bottom of page