The Infosys Labs research blog tracks trends in technology with a focus on applied research in Information and Communication Technology (ICT)

« December 2018 | Main | May 2019 »

March 25, 2019

Continuous Authentication

The world is rapidly changing; smart phones, laptops, and wearables have altered the way we live. Today people frequently switch between and share devices with each other. Over the years, organizations and regulators have mandated the authentication of users in an attempt to prevent fraud and protect personally identifiable information.

Increased focus on user engagement has forced companies to avoid excessive authentication related tasks in order to enhance customer experience. However, by adopting this strategy they risk compromising customer data and hacking. Conversely, fraud avoidance is also a primary concern for organizations and they enforce extra barriers such as two-factor authentication and short login timeouts which increase account security, but it comes at the cost of user experience.

Continuous authentication has emerged as a promising option for companies trying to provide secure services to clients in a digitally vulnerable world. Continuous Authentication is a dynamic, risk based authentication that changes the procedure of authentication such as event based authentication used today to a process. Instead of users logging in or out, the application continuously authenticates the user through behavioral biometrics.

Continuous authentication uses behavioral biometrics to enable organizations to move away from one-time authentication, thereby increasing security and improving user experience. Behavioral biometrics uses machine learning to continuously monitor the user's behavior based on interactions with apps and websites. Malware and frauds with login credentials are easily identifiable with the application of continuous authentication. Yet, behavioral biometrics' s biggest advantage is that the technology is capable of working with existing infrastructure without requiring any modification.

Another key advantage is the ability to assign user action constraints based on tolerable risk or context. These constraints can be based on factors such as location, presence of other people or the time of day. Infact, behavioral biometrics has been approved by the PSD2 as a valid authentication option.

Startups like BioCatch, authenticate users continuously when their online and protect them from cyber threats. Some examples are Trojans, account takeover and other malware. The company is already providing real-time fraud prevention for 2 billion sessions a month.

Continuous authentication is the first true alternative to pesky passwords and an effective method to prevent fraudulent transaction which amount to around $130 for mobile transactions and about $115 for tablets on a normal day. The technology stands to secure the way interact with the digital world while streamlining the entire experience.


Wearables in Banking

Wearables are electronic devices incorporated into items that can be comfortably worn on a body or embedded into accessories. According to IDC, the wearable tech market will see shipments almost double to 240.1 million by 2021. Wearable adoption has been driven by higher acceptance and adoption by newer generation due to its positioning as a standalone device and the evolution of operating systems which are more user friendly.

Wearables enable bankers to develop customer centric solutions and offer hyper-personalized offerings to clients. The technology enables organizations to collect real-time behavioral data which can be leveraged to provide personalized and real time offers similar to google and amazon. To Illustrate, consider wearable fitness devices which monitor the wearers movements, location and activity level to advise users. Similarly, banks can leverage the technology to become their 24/7 personal banker.

Another advantage that the technology provides is the autonomous verification of customer identity. Wearables enable users to authenticate transactions without having to remember password by using biometric technology. The feature would enable banks to streamline services and customers to initiate payments or withdraw money from ATM with the wave of a hand. For instance, The Australian and New Zealand Banking Group (ANZ) is now accepting cash withdrawals via smartwatches at 2,400 ATMs across Australia. Similarly, Gemalto's MiniTags and MicroTags have already been certified by Visa, MasterCard and Amex. And can be used at all locations where these companies accept contactless payments.

Advancements in biometric technology have helped financial institutions combat rising concerns of fraud and identity theft. An example is the Apple Watch which utilizes plethysmography to identify if a body part (in this case, the wrist veins) has increased or decreased in size to identify the user. Banks will also need to leverage architectural design and prevention analytics to protect customers against threats such as Bluetooth theft, signal interceptor issues, and virus attacks.

Multiple financial organizations have already taken the lead in the adoption of wearables. Barclays has partnered with brands like Topshop, Garmin and Mondaine to design devices compatible with its bPay chip in order to enable customers to make payments with a range of wearable devices. Similarly, ABN AMRO bank is testing smart rings, watches, and bracelets as a new NFC payment method. Companies like US Bank, Wells Fargo and Citibank have also embraced wearables by provide balance tracking and notification to customers through smartwatches.

In conclusion, wearables are going to revolutionize the way organizations interact with their customers and prospective clients. Organizations can only meet rising consumer expectations by using the technology to interact with their customers outside the branch. Wearables are poised to become an integral part of our everyday life and provide a doorway for companies to provide cater to our needs.


Voice Analytics in Financial Services

Voice analytics is the use of a voice recognition tools to analyze a spoken conversation. The technology has the potential to increase the performance of client facing operations and the value of every customer interaction. According to Grand View Research, the voice analytics market to reach $1.64 billion by 2025 and Opus Research expects companies 57% of companies to increase sales and collections.

Customer satisfaction is one of the key drivers of voice analytics, the technology offers the capability to understand the emotions hidden behind the customer's words and analyses whether an expression is neutral, negative or positive, and to what extent. The solution offers a trustworthy alternative for post-call surveys and helps identify where agents fall short and excel to help analyze and tweak agent performance in real time. An example is ING Bank, the bank has leveraged voice analytics to uncover issues that require improvements in customer experience, agent performance, sales potential, and service quality. In addition to enhanced customer experience, the bank the solution reduced telecom costs thanks to shorter calls.

Increasing regulations is also a critical driver of voice analytics. Voice analytics based compliance is a trending topic which helps banks by offering risk-based quality monitoring. The solution leverages analytics to identify phrases that indicate mis-selling or other compliance failings. The banks can then isolate, follow up and remedy the issue, within hours. Another aspect of compliance where the technology has great potential is authentication, voice analytics helps banks to provide faster access to services by verifying customer identity in an ambient and seamless process. For example, Barclays reduced the authentication process by an average 20 seconds using voice recognition. It minimized identification and authentication issues which resulted in increased customer satisfaction.

In conclusion, voice analytics is set to radically change the way we interact with customers. While offering enhanced insights into their customer's satisfaction and interests, the technology also helps banks adhere to regulatory protocol. Hence, the adoption by banking giants such as Ally bank, ABN AMRO, US Bank, TD Bank, Santander, Citibank and Bank of America.



Computer Vision in Banking

Computer vision is an AI application that provides computers the ability to visually understand the world. The technology is capable of automatic extraction and analysis of relevant information from images. According the Grand View Research, the global machine vision market size is expected to reach USD 18.24 billion by 2025.

KYC is an important usecase for computer vision. The technology enables banks to shift their focus from duplicate databases and use biometrics for identification of prospective clients. The technology streamlines the KYC process by allowing prospective customers to open accounts over their phones. The European bank BBVA is attracting customers by allowing prospective customers to open an account with a selfie or video call. The solution helps banks increase customer convenience and move further in the customer centricity journey.

The technology also helps organizations analyse customer behavior in real time through emotion recognition by analyzing micro-expressions, pupil dilation and eye moments. Banks can capitalize of the information by offering personalized products and modifying process to make it more convenient for clients. Similarly, computer vision can track intent of individuals near cash points and detect threats in real time, thereby providing a safe environment for their customers.

Back office operations would also benefit from applications of computer vision. The technology would streamline back office operation by reducing the paperwork through automated data extraction. The solution would help banks save time and cost involved in processing of documents by humans while increasing accuracy and efficiency.

Computer vision also helps traders by providing granular data for making trade decisions. Investors and economists can track data such as automotive activity in a retailers parking lots and shipping container movement in ports to identify trends in the economy.

The rise of AI has opened an exciting opportunity for the financial sector which has been burdened by the mounting paperwork and regulations. Computer vision is poised to revolutionize everything from back office operation to how investors evaluate a company, ubiquitously.


March 20, 2019

Future of Personalized Assistant

The adoption and usage of smart assistants is gaining popularity with more and more people using google, apple, or amazon assistants for navigation, searching, scheduling, doing mails etc. With advent of new age of communication and application based digital economy the usage smart assistants will be ubiquitous. Apart from helping people in their day to day repetitive tasks smart assistants will eventually integrate into our lives as our alter egos. But that might take couple of decades to become reality. 
Other take is how smart assistants can help us in learning new skills or moderate our actions to improve existing capability?  In today's world smart assistants can certainly act as information or knowledge provider; can they go beyond and play the role of personalized coach? Well, the answer is yes; with growing power of data processing and advancements in ML and AI techniques smart assistants are poised to take up such roles. By leveraging right sets of data we might be able to infuse capabilities human personas of champions of fields in the smart assistants. Imagine someone is working on a presentation note that needs to be rearticulated, and he takes the help of an assistant which is powered by Shahi Tharoor like grip in English language. Or a guy preparing a desert recipe gets culinary suggestions from Sanjeev Kapoor; or a young tennis player is getting backhand shot tips from Federer.
However, multiple challenges from technical feasibility of such assistants remains. Nevertheless, such avatar of personas enabling exclusive assess of their (champion of the field) capabilities to everybody can be next evolution step for smart assistants. Thereby fundamentally changing the way we interact and leverage them. These Avatars powered with multi facets interaction capabilities like text, speech, natural language, gesture, extended reality, as well as haptic interfaces will extend the boundaries of association, experience, and application.
Apart from the consumer centric applications mentioned above personalized assistants have multiple opportunities in enterprise environment as well. Areas such as asset monitoring and installation, safety training, maintenance work, employee upskill-reskill etc. have multiple touchpoints where smart assistants can complement the enterprise users for efficiency and effectiveness. Moreover, in any field or industry there is always need for coaches, and what can be better than the very champion of the subject being the coach, very own personalized assistant paving the learning path for employees.  Industries like engineering, utilities, construction, manufacturing can leverage personalized assistants to optimize workforce productivity and resource utilization.
Whether it is individual user or enterprise user the relevance and important of smart assistant will increase significantly in upcoming years. And the essence of the new age assistants will be personalized Avatars. So, boarding the bus of smart assistant evolution should be key focus for system integrator to be in race for enterprise adoption as well as relevant among Googles of the world.