Biometric solution reduces customer onboarding time to just seconds whilst dramatically reducing fraud
/ EmbedIT overcame complex, age-old paper-based identity verification problem in the Philippines, cutting onboarding costs and times with a break-through customer identification solution.
/ The innovative biometric workflow and processing solution automatically verifies a selfie taken through the Home Credit mobile app with photo ID of the customer, finding potential matches in the customer database to confirm if it’s a new or existing customer.
/ With virtually all customers now using automated online identification, Home Credit can move forward with business growth plans that rely on fully digitised application processes.
Location
Czech Republic, Philippines
Expertise
Product discovery
UX/UI design
Product management
Development
Testing & Analytics
The challenge
Home Credit Group is an international consumer finance provider offering a range of credit and insurance services around the world. An essential part of processing new applications for its products and services is ensuring that the customer is who they say they are (to prevent fraud) and identifying whether they are an existing customer (to prevent duplicate customers in the core systems).
However, this was difficult to achieve in the Philippines, where there is considerable variance in spelling of names and the personal data that people use to identify themselves, and as many as eight official document types that are valid as ID, some of which can be printed by citizens at home.
As a result, roughly 20% of Home Credit applications could not be processed automatically, with these applicants having to go through a manual identification process instead. This consisted of a team of operators sifting through data until they could confirm or deny the person’s identity. It was not only dramatically slowing down the overall application process, it was also increasing the number of wrongly rejected applications, with human operators turning out to be overcautious and marking many valid applicants for rejection as not matching.
In addition, highly manual workflows were preventing Home Credit from achieving one of its key business goals – to become a 24/7 business with instant, anytime, anywhere product applications.
The solution
With such significant variability in customer data, the only reliable way to confirm people’s identity was through the use of face recognition technologies. So in 2019, EmbedIT and Home Credit started working together to bring this biometric checking to Home Credit. The project consisted of two main workstreams.
/ Enabling biometrics information to be securely generated and submitted by customers and stored within the Home Credit system.
/ Creating decision-making workflows that would use visual information to make reliable decisions about the identity of applicants.
The Home Credit mobile app was integrated with the Innovatrics biometrics engine, allowing customers to take a selfie and securely submit it to the Home Credit database.
The solution uses liveness detection to verify the user is physically present, with OCR technology reading the ID’s data and picture.
A new orchestration layer was built between the mobile app and core systems, which fine-tuned Home Credit’s in-house data matching algorithms and added new cloud-based solutions. It included the Camunda Zeebe workflow engine and smart face-matching algorithms based on artificial intelligence (AI) technologies.
This flexible technology stack, a combination of existing customer databases and new market-leading technologies, enabled the team to implement the bespoke workflows that Home Credit needed. Stored images from photo ID can now be checked to confirm that it’s the same person as in the selfie, and if it is, data matching continues, to ensure that the applicant doesn’t already have a profile within the customer database.
The second part of the solution took just over six months to scope, design, build and implement, with another six months spent running the old and new systems in parallel, to ensure proper training and testing of the process and algorithms.
“The implementation of this biometric solution has provided considerable value in assessing our customers, addressing both underwriting risk and fraud detection with minimal manual intervention. Additionally, it has been a proactive undertaking, laying a solid foundation for our future objectives, including expedited and precise customer identification, eliminating the need for manual deduplication, and ensuring proper management and utilization of KYC data.”
Arvie John Lusanta
Risk Technology and Architecture Manager, Home Credit Philippines
The Benefits
At EmbedIT, we work to deliver clear and profitable business impact, and for Home Credit, this means:
/ Significant reduction in application processing times.
Significantly more applications are processed without manual intervention. And the matching process is incredibly fast too, with images being confirmed and given a probability score within seconds.
/ Reduction in workloads and operating costs.
Image matching has reduced manual matching from 20% of applications to 4%, freeing up a large number of people to work on other tasks, significantly reducing costs.
/ Improved data quality.
Now, only genuinely new applicants are added to the Home Credit customer database, which reduces duplication and the work needed to fix it.
/ Better balance of risk vs profit.
The solution focuses on reducing the number of fraudulent applications and increasing the acceptance rate of valid applications, enabling Home Credit to accept more applications while actually decreasing its risk profile.
/ A flexible, extendable solution.
The system architecture means it’s now easy to add additional workflows that deliver value, for example, updating an existing customer’s record with additional or new data from their latest application.
/ Highly performant solution designed for high volumes.
Based on a SaaS solution that’s flexible, reliable and delivers 24/7 operations, and a structure that will grow alongside the business.
/ Accelerated business progress.
As a SaaS solution, Home Credit is well positioned for future requirements without being punished for growth. This includes rolling out biometric on a larger scale, for example to authenticate daily operations in the app like PIN retrieval or card unblocking, or for multi-factor authentication (MFA).
EmbedIT is now working with Home Credit to roll out similar biometric functionality for business units in the other countries it operates in.
