Jumio is a leader in the trusted identity-as-a-service (TIaaS) sector. The company works with firms in the banking, cryptocurrency, retail, and travel sectors, helping these companies to digitally transform their customer onboarding process through the use of Jumio’s digital identity verification platform. Identity verification as a service is a market McKinsey predicts will surpass $16 billion by 2022.
Jumio uses advanced machine-learning based algorithms to quickly analyze identification documents – coupled with a selfie – to verify a person’s digital identity matches the real-world identity in seconds. To find out more, Digital Journal caught up with Jumio Head of Global Marketing Dean Nicolls.
Digital Journal: What is Jumino’s Identity-as-a-Service platform?
Dean Nicolls: Jumio’s Trusted Identity-as-a-Service combines machine learning, AI, computer vision and identity experts to provide the highest level of accuracy and fraud detection without falsely rejecting potential customers that are so critical for expanding businesses.
DJ: What services does Jumio provide?
Nicolls: When banks and other financial services companies onboard new customers online, they embed Jumio’s ID Verification, Identity Verification and Document Verification solutions into the customer journey. With each solution:
ID Verification: Determine if an ID document (e.g., passport, driver’s license, ID card, etc.) is authentic and belongs to the user.
Identity Verification: Ensure that the person behind a transaction is whom they say they are and that they’re physically present (via Liveness Detection) by capturing a selfie and comparing that to the image on the ID.
Documents Verification: Quickly scan documents, such as utility bills, credit card statements and extract key information to provide proof of address.
DJ: How do financial services use this?
Nicolls: Financial services use Jumio in a number of important use cases. The most popular is for customer onboarding and eKYC (Know Your Customer) compliance checks. Jumio is embedded within the account set-up process to ensure that someone is whom they claim to be with a high degree of reliability and security.
Our solutions help financial services companies. There is:
Detect & Deter Fraud: Jumio’s identity verification solutions help prevent account takeover and other types of online fraud. Jumio’s solutions can detect if a government-issued ID is authentic or if it has been doctored or manipulated. Plus, by requiring a selfie and comparing that selfie to the picture depicted on the ID. Many fraudsters are reluctant to include an image of themselves (selfie) as they don’t want it known that they’re defrauding the company. So, identity verification provides a powerful chilling effect on would-be scammers.
Optimize Online Conversions: Most of our customers care as much about ensuring legitimate customers are quickly verified without undue friction. Jumio helps its financial services customers create a simple and seamless experience that streamlines the onboarding process with a verification methodology that is increasingly familiar.
Meet Compliance Mandates: Jumio’s identity verification solutions help financial service companies comply with a variety of global regulatory mandates and directives, including KYC, AML, GDPR, and PSD2.
Increasingly, financial service companies are leveraging Jumio for emerging use cases including high risk transactions including password resets and high value money transfers (such as a wire transfer to a Swiss bank account). For these types of transactions, more and more companies are leveraging Jumio to confirm that the person requesting the password reset or money transfer is in fact the account owner.
DJ: How did you develop your machine-learning based algorithms?
Nicolls: As it turned out, one of the very best application areas for machine learning was classical computer vision, though it still required a great deal of hand-coding of specific features to get the job done. At Jumio, we use classical computer vision to create ID templates for specific ID types (for example a California driver’s license or a French passport). We created hand-coded classifiers like edge detection filters, so the program could identify where an image of an ID started and stopped and advanced Optical Character Recognition (OCR) technologies to recognize characters and letters in different languages. From all those hand-coded classifiers we developed algorithms to make sense of the ID document and “learn” to determine the type (driver’s license, ID card or passport) and country of origin of the ID.
Jumio has processed more than 130 million IDs over the last few years and this data contains important patterns — patterns often indistinguishable to the human eye. Jumio leverages machine learning and deep learning to build production models that perform three vital functions related to online identity verification: data extraction, fraud detection, and risk scoring. These models have helped Jumio to dramatically improve identity verification accuracy and verification speed.
There are still instances where the machine and deep learning models fail to identify what is wrong with an ID. This is where verification experts can apply their expertise to determine what error occurred and teach our algorithms how to spot the issue in the future. This feedback loop creates a different method of learning where new information is constantly being input to the learning model so that the model can improve.
Jumio has pioneered the concept of compliant machine learning that builds in data privacy and security at every stage of the machine learning workflow including initial data capture, ID preprocessing, data tagging, algorithm training and model deployment. In the wake of GDPR and because of our global scale, Jumio can now build machine learning models that don’t co-mingle data from different customers.
As a result of these efforts, we have built highly predictive models even when customer data is disaggregated. Machine learning has already been refined to reduce the time it takes to verify an ID document or an online identity which, in turn, is helping Jumio’s customers reduce their abandonment rates and increase new account conversions.
DJ: How are these services different from your competitors?
Nicolls: There are a number of ways in which we differ from our competitors. Our biggest differentiators are with ‘Verification Accuracy’: We measure this in two important ways: our ability to quickly verify your good users and our ability to detect online fraud through a combination of machine learning, AI, biometrics and verification experts. This is crucial in terms of accurate fraud detection.
Second, ‘Machine Learning & AI’: When it comes to machine learning, he larger the dataset, the smarter the algorithms. Machine learning models can be created for many different ID verification use cases, such as blur detection and bad image quality detection which requires a large volume of good transactions to better train the algorithms. Our scale gives us a big advantage over smaller competitors in our space who have a fraction of our total transactions.
Third, ‘Scale’: With over 130 million online verifications performed to date, Jumio has a leg up on its competition because of the benefits that derive from this scale. This includes being able to feed our machine learning with more data that, in turn, translate into smarter, more accurate algorithms. This scale helps us build the largest commercially-available database of fraudulent IDs in existence which helps Jumio quickly flag and reject repeat offenders. This in turn translates into higher levels of fraud detection and overall verification accuracy. Jumio supports ID documents in more than 200 countries and territories, including older versions of those document types, paper and plastic. This includes passports, driver’s licenses and ID cards, including older versions of these documents.
Fourth, ‘User Experience’: Jumio is committed to helping our customers reduce abandonment rates which would in turn lead to higher conversion rates. We do this by providing real-time feedback within the user experience that lets the user retakes a picture of their ID (for example) if the image take is blurry or has excessive glare. We also help our customers reduce the number of screens (steps) that a customer has to complete in order to complete the verification process. We also share best practices in terms of customer-facing message.
DJ: Which types of companies do you work with?
Nicolls: Jumio works with a wide range of global customers across a range of industries including banking, financial services, sharing economy, telecommunications, healthcare and cryptocurrency.
In a follow-up interview, Dean Nicolls assesses the general cybersecurity risks to businesses. See: “Q&A: Why cybersecurity needs to be part of digital transformation.”
