Valida, artificial intelligence for fraud detection

Scroll to see more

seguridad_19_cabecera

 

This technology detects manipulations in any type of digital document to offer an extra level of security in verification and KYC online processes

Premieres at CES 2019, it will be soon available for on-site testing at the MWC-4YFN in Barcelona and at the RSA Conference in San Francisco 

 

The world we know changes each day, we live in a society where digital environments are part of our daily lives. Internet offers us all the products and services that we wish to acquire: the universe at our disposal, either through a computer or with our mobile phone or tablet. Anything, anywhere and at any time.

This reality offers companies great advantages, since it is not necessary for customers to be physically present in shops, offices, banks, department stores, supermarkets… either to buy a phone, open a bank account or take out insurance for their vehicle.  But with the large number of online processes that are carried out daily by users, companies need to have at their disposal mechanisms to authenticate and verify the thousands of identity documents they receive on a daily basis in digital processes.

Evaluating the authenticity of digital documents to prevent fraud in online user verification, digital onboarding and Know Your Customer (KYC) processes is Valida’s main goal. This technology developed by Gradiant, in collaboration with atlanTTic research center of Universidade de Vigo, allows companies to offer an extra level of security in these operations, as it automatically analyzes any type of identity document (and other digital files such as payroll, invoices, receipts, etc.) and detects possible attacks of impersonation and forgeries in the data present in these files.

 

Valida, the perfect partner to ensure document security

Remote and secure user identity verification has become a must-have for digital business. This fact makes it essential to develop advanced, reliable and easy-to-use solutions to prevent fraud, especially in digital onboarding and KYC processes to meet the customers.

Valida is an effective solution for companies with document verification processes. This technology automatically detects forgeries produced in digital documents. By using forensic AI-based techniques, our technology analyses the document and show the manipulated areas, through a heat map that clearly indicates where the forgery has taken place.

Valida supports all types of identity documents and nationalities: passports, identity documents, driving licences, etc. without specific adaptations. In addition, it does not require connection to external databases (e.g. identity document databases) to detect modifications and also warns if a document has been captured from a screen (and is therefore not a photo taken from original document).

 

Specially designed for fintech and insurtech

Although digital onboarding processes -remote opening of financial products and services by identifying customers through the use of biometric technology for user recognition- brings great advantages for customers and businesses such as easy shopping and customer growth, it also carries security problems like fraud or manipulations of documents and user verifications. In addition, the advancement of technology and the democratization of multimedia publishing programs, make forgeries more and more realistic.

For this reason, insurance companies or banks, as well as the new paradigm of fintech and insurtech -which work daily with thousands of documents (ID documents, accident reports, photos, etc.)- already need technology to guarantee remote users verification, as well as the authenticity of the files they receive in digital processes to avoid security problems and economic losses due to possible spoofing attacks, for example. Valida performs this task, automatically detecting in seconds the regions edited and modified in any type of document.

This site is registered on wpml.org as a development site. Switch to a production site key to remove this banner.