The Era of Conversational Banking

Historically, banking has been among the most resistant industries to embrace digital transformation. Calling in or waiting in the teller queue is no longer a reasonable customer expectation for any kind of banking service. Customer service is often limited to business hours.

Today’s tech-savvy customers want to interact with the bank on their own terms. Technology has also made it possible for banks to address customer demands with online and mobile banking services to their customers.

Online and mobile services have helped banks make strides in the right direction, and internet banking has helped simplify bank processes. Although not completely seamless and user-friendly, as customers are inundated with information and made to scroll and filter through hordes of it, it’s just a matter of time before these issues are ironed out.

With digitization, any problem/limitation can be mitigated or eliminated with an appropriate solution.

The Solution

Digital conversational banking means interacting with a non-human interface to discuss your finances or even seek advice. It is about managing your financial life through voice or text. Conversational banking is about using chatbots to text or ask questions and issue commands.

Today, conversational banking has made visiting a branch a matter of customer choice.

AI-powered chatbots fill the gap between humans and websites. Banks can rely on AI to better understand their customers and reduce the number of basic inquiries a bank’s support staff spends time answering, thus freeing up their staff for other customer-facing or value adding tasks.

A customer asks a question, and the bot processes their request to deliver the most relevant information. Machine learning improves the accuracy of chatbot responses. Major banks have adopted conversational banking products to simplify banking processes, which will soon be the norm in the banking industry.

The lines between the capabilities of man and the capabilities of machine are slowly but steadily blurring as machines continue to take up more and more core tasks. Customers can engage with these machines in a natural way.

Rise of Chatbots

Chatbots represent the next step in customer-facing evolution in the banking industry. Chatbots are computer programs that use artificial intelligence and imitate conversations with people. They can transform the way a customer interacts with the bank from a series of self-initiated tasks to virtual conversation. For example, Google’s Allo, powered by AI learns how the user replies and then offers relevant suggestions. It has its own search assistant bot that extracts knowledge from existing Google’s services.

Chatbots at banks will allow customers to ask questions about their financial accounts, initiate transactions and get financial advice through either text messages or any other messaging services.
The promising chatbot tech in banking is expected to quickly reach a level of sophistication that allows these bots to perform every task performed by a bank service officer.

Conversational AI

Virtual assistants are nothing but bots combined with conversational user interfaces and machine learning. Natural language, the fundamental element used in conversational user interfaces is thus also the key element for conversational commerce.

Virtual assistants have the potential to impact customer experience and loyalty.

These technologies facilitate two-way interactions that allow banks to establish genuine relationships with their customers while providing a more seamless experience across multiple digital arenas.

However, it is not yet possible to provide flowing natural-language conversations, real-time adjusted flow conversations with a software, though efforts are concentrated towards that.

Once this is made possible, chatbots can slowly inch towards automated responses and push on-demand information / choices to users in real time as opposed to just answering queries or requests. Conversational chatbots will can also contextualize and personalize their conversations with customers. So the sooner they start learning customer preferences, the better.


In the evolution of banking technology, internet banking reduced the operational time by enabling customers to transact real-time. In the future, conversational banking will reduce operational costs further by providing instant communication.

Building Smarter Businesses with Edgeverve’s Business Applications Powered by AI

Profound, Disruptive Effect of Artificial Intelligence on Global Businesses

With the change in pace of technological innovation, businesses grapple with driving digital transformation. The need of the hour is adoption of the right technology without disrupting the existing IT infrastructure. Businesses need to become future ready by leveraging enterprise data for useful business insights, which can be powered by AI. The business world is abuzz with talk of how artificial intelligence (AI) is impacting value creation and realization at enterprises. Technologies such as machine learning (ML), deep learning (DL), natural language processing (NLP), speech recognition, image recognition, among others, are often mentioned as examples of what AI brings to the table. These are completely transforming the way products are manufactured, delivered and consumed. The latest McKinsey1 report states that a few early AI adopting firms have already begun witnessing some real-life benefits, making it more urgent than ever for others to accelerate their digital transformation.

Enterprises are Grappling to Implement these Newer Technologies

Despite understanding the potential AI can unleash, enterprises are trying to figure out what works best for them and how to implement it without disrupting their existing Infrastructure. In addition, enterprises face several other barriers such as uncertainties over how best can they utilize AI and make it effective for their organization.

Despite these challenges, 75 percent of senior decision-makers believe AI is fundamental to the success of their organization’s strategy, according to the Infosys1 CIO report. And, 65 percent of decision makers believe that the impact of AI will bring out the best in their organization’s people. The report further adds that by 2020, those currently using or planning to use AI technology estimate an average 39 percent boost to their organization’s revenue. A report by research firm IDC3 points out that 67% organizations worldwide have already adopted AI or plan to do so over the next 5 years, but 73% lack the necessary data science or AI expertise.

Several such reports and studies conducted reinforce the need for simpler, viable solutions built on AI.

Built On AI – Edgeverve’s Business Applications Addressing Specific Business Problems

It is important to understand that the gaps for implementing these technologies often differ from one industry to another – whether it be pharmaceutical, manufacturing, retail, financial services, or something beyond. The key is to employ a solution that can resolve a real business challenge and mitigate risks in real-time. For instance, if a hurricane in Florida disrupts its supply chain, a manufacturer would be faced with fulfilling demand in the face of compromised supply lines. In this scenario, instead of an expensive and inefficient rudimentary approach, the manufacturer would be better served if they were to use a solution employing AI and ML to predict which supply line would be disrupted and then automatically (using automation) switch to the next available supplier.

Through extensive research and discussions with clients, EdgeVerve has converged on the viewpoint that AI should be consumed as micro applications or micro business solutions to specific business problems. Leveraging the XtractEdge™ platform to create a foundation of automation and AI for these Business Applications, these solutions can solve real-life business problems.

This means, a specific business solution for a procurement officer grappling with simplifying procurement, can simply sit atop the existing data, without the need to disrupt the infrastructure. It, being built on AI (using open source technologies) can utilize the data and derive insights which can help steer a business towards cognitive capabilities. Similarly, a banker grappling with a collections and customer service problem can utilize a pre-configured business app that will expedite time to value.

Edgeverve’s Suite of Plug & Play Business Applications for Every Enterprise

It is no secret that enterprises tend to have a plethora of existing apps. These legacy apps are often incapable of intelligent performance. EdgeVerve’s Business Applications can seamlessly integrate AI capabilities without ripping apart the existing landscape.

For example, addressing the problems faced by lenders and debt collectors, the data-driven, intelligent XtractEdge Loans Loss Mitigation business app predicts which customer will be delinquent next month, or maybe the next quarter. It doesn’t stop there; it goes ahead to offer actionable insights such as personalized negotiable strategies created after behavioral and financial analysis for each consumer, guidance to recovery agents to call the right customer, setting the right tone and helping gain an upper hand in negotiations.

As we go forward, realized digital transformation will require more extensive use of automation and AI. The EdgeVerve suite of plug-and-play Business Applications built on AI provide that boost to enterprises that will make a difference in how they realize value and compete. Their ease of use and deployment are designed to help enterprises get over their hesitation to go further with automation and AI. Designed to integrate seamlessly with existing infrastructures, these business solutions built using the XtractEdge™ foundation can be deployed by enterprises of all sizes. We, at EdgeVerve are excited about an emerging new world of built-on-AI – in our case, built-on-XtractEdge – Business Applications. Join us in this revolution.

Read more about the #BuiltOnAI campaign here.

Don’t do it in a Hurry!

Across industries, businesses, financial institutions, and service providers, there has been a constant focus on establishing omni-channel presence for access to the services offered to the clients. This has resulted in a challenging competition for designers to present innovative and simplistic designs to catch the attention of the customers and the industry. The end-customer is always the beneficiary receiving new updates with a wow factor on the devices. It’s a constant endeavor by designers to understand the psychology of the consumer’s mindset, behavior patterns, predictive decision making and to design the systems accordingly.

On the other hand, the demand of the consumers and the reviewers on this front has increased manifold. There is a constant ‘invisible’ audit. It’s such a thing where each consumer of the service can have their own point of view. We have teasers, pre-launch reviews, launches, forums, day-1 review, day-7 review, day-30 review and many more. This has been fantastic for customers, who are well informed before actually making an investment.

However, in my opinion, the role, industry and requirement for each business user is of paramount importance at the time of freezing or finalizing the design. The requirement of each business, each user, each role is very unique. This should be the primary and the most important factor for decision-making in the constant effort of the service providers to beautify and revamp the screens. There are lots of efforts in terms of ideas which technology companies and service providers invest in to incorporate them into the platforms. The inputs of subject matter experts are critical in designing. The nuances of every business are best understood by the experts themselves. The client accessing the information on a device app for inputting information is totally different from the profile of a back office user of an organization. The ease of usability and the depth of information required is directly proportional to the role of the user. For instance, considering an Insurance business application, the front-end design for a prospect customer navigating an app for policies vs a customer actually buying a policy vs the company representative selling a policy to a customer vs a back office user creating an actual policy for a customer have absolutely varied needs. While a prospect customer requires an intuitive screen which can show him the comparison of the features and premium amounts with minimal information, the customer who buys a policy would require to know all the fine print in the policy before making the premium payment. The sales representative would require analytical data and key selling points of a policy, whereas the back office user in the insurance company would require an application which has the facility to capture detailed information about the customer in the most efficient way.

Although we understand less is good, designers have to do their research, take help of subject matter experts, understand the nature of business, geography of business and many more factors to be able to meet the expectations of the users and receive a pat. For those businesses, financial institutions, service providers who are yet to take the plunge, are being forced to think and re-think to invest in redesigning the UI to increase business and meet the demands of their customers, get appreciation from internal users, peers, competitors, market analysts. This is a critical decision with long-term impact which has to be taken by the product managers along with the overall strategy of the organization. So don’t do it in a hurry.

Corporate Credit Appraisals and Decision making

Corporate Loans occupy a significant size in the asset portfolio of large banks. Most banks have a well-defined corporate credit department comprising relationship managers, credit appraisal officers, risk officers, approvers and department heads. Corporate loans not only generate interest and fee based income but also build the asset side of a bank’s balance sheet. Related businesses of corporates such as current accounts, forex businesses, salary accounts of staff, financing options for employees of corporates, travel cards for top management/executives, remittance business, cash management are an added attraction for banks to pursue corporate banking clients.

Banks follow multiple methods for corporate credit appraisals and assessment. Credit appraisal and assessment has evolved from manual processing for balance sheet analysis, credit monitoring and assessment (CMA), credit worthiness and credit rating checks, to partially excel based tools for CMA analysis over a period of time. The complexity of the processing for corporate credit necessitated the need for development of a software with all the rules/checks/workflows built in, but it took a large time for a vendor to build the same. However currently most banks use software for corporate credit appraisals, decision making, etc.

The software is required to have organization level policy checks, customer specific checks, industry checks, credit worthiness checks and financial / cash flow comparison to ensure automatic credit decision. There could also be support in the software to overturn the workflow based system decision for credit approvals by manual intervention based on manual approvals/committee approvals with deviations and exceptions. Depending upon the criticality of exceptions to be overridden, multiple levels of approvals can be enabled. Any automated decision / manual process decision needs to be captured in the audit trail / report in the software, and stored so that the same can be referred to at a later point in time. Many a times, decision making is dependent on corporate relationships, reputation of share-holders, past history and precedence which may not reflect totally in the financials or cannot be captured in the credit report / software. It is important to have a decision from the software and record / capture the reason for overriding system approvals by a credit officer. The software can play the role of a watchdog oblivious of customer stakeholders and subjectivity. The software may also open up the hidden nexus between staff and corporate.

Corporate loans can go bad because of multiple reasons – non-performance or project failures, market risks, diversion of funds, operational failures, bad structuring of finances, people/process issues in the company, management issues etc. Needless to say, the quality of credit assessment using software or manual processes cannot be compromised, the credit projects should be in line with prevailing market conditions and the approving authorities should be aware of market trends for the future. The checks and expertise do not prevent a corporate loan from going bad.

However, the bank and the credit committee are always looked at with suspicion by the larger public when a loan goes bad for reasons beyond inaccurate credit assessment by the officer. The transparency which a software can provide with data capture, validations and workflow cannot be equated to a manual process, and the blame game can be avoided if the data and measures taken to approve with exceptions and deviations are recorded in the software with details of approval authorities.

To conclude, the software should have the capability and workflows for manual/automatic decision making, so that the policy makers at a bank can decide on the course to be followed to ensure transparency in the system. With mounting NPAs and bad loans, bankers are looked at with a great degree of skepticism by customers, fellow bankers, and the general public. It is high time banks come up with a mix of manual and automated credit decision skills where each and every action is captured, monitored, tracked and recorded. The staff cannot prevent some of the loans from going bad and cannot control the external factors contributing to bad loans. Inherent risk control measures need to be inbuilt in the software for monitoring and control. Furthermore, automated credit monitoring system during the life cycle of the loan must be put in place to provide an early alarm to banks to taking adequate measures before loans go bad beyond repair. This can go a long way in reducing the gross NPA of a bank.

AI and Humans – The way forward

We are in an era of fourth industrial revolution with rapid advancement of Artificial Intelligence from personal assistants to autonomous vehicles. Today’s AI is more of narrow AI meant to perform specific tasks, and there is a transition towards strong AI for cognitive tasks that require an ability at par with humans. Final phase would be achieving super intelligence which is expected to surpass human intelligence in all the areas. Does this mean AI will replace humans or can they collaborate for betterment of the society?

In 1950, British mathematician Alan Turing, spoke about machines that think for the first time. The word “Artificial Intelligence” was introduced to the world by John McCarthy, an American scientist in 1956. World chess champion Gary Kasparov was defeated by IBM DeepBlue in 1997, this was one of the greatest achievements of AI. Though researchers and scientists have worked on AI earlier, AI has gained more popularity only in the recent years due to the increase in data volume, high computing power and cheaper storage.

How are we using AI?

We are using AI in one or the other form, be it smart phones or vaccum cleaners. AI is revolutionizing many industries like healthcare, finance, retail, supply chain, transportation, etc.

What’s the future of AI?

AI will bring more smarter machines to make our lives easier and better. Once luxury, Mobile Phones have become important part of our life. And, AI becoming part of our daily lives is not too far.

What about jobs?

AI will definitely make some jobs obsolete. Repetitive, mundane tasks will be taken over and at the same time will bring newer opportunities. Companies should ensure new jobs are created for the employees, and reskill them for innovative jobs. Did the ATM take away banking jobs? Banking has become so easy compared to standing in long queues for getting money.

Is AI a Threat to Humans?

Eminent scientist, Stephen Hawking warned that development of full artificial intelligence could spell the end of human race. Elon Musk, chief executive of SpaceX fears AI could be our biggest threat.
AI in the hands of wrong people and wrong applications of AI can be a threat. It’s similar to questioning whether nuclear energy is threat to humanity? AI driven weapons could be as dangerous as nuclear weapons. There is definitely a need for proper government regulations around AI.


We can embrace the AI revolution with open mind. Technology replaces less creative jobs and opens up doors for innovation. We can use AI for betterment of our personal, professional lives and society.

Making Virtual Workforce Real

8 AM. A typical Monday. Sarah is excited for the week ahead. As she reaches work and opens her mailbox, she is relieved to see fewer emails than usual that need her immediate attention. She first looks for an email from someone called Mars. She is thankful to have received this very systematic email with important reports. She didn’t have to probe into multiple apps! Having gone through the emails, she wonders how much her life has become simpler since Mars has joined her team. He is very good with numbers, hardly makes mistakes, and is quite fast too. In the short amount of time, he has learnt quickly, in fact become quite adept at many tasks! Also, he does jobs on her behalf that she used to hate, namely, having to send reminder emails to people! He has become quite an asset to the team. Smiling to herself, she sends an email to Mars- ‘Thank you for the reports. She gets a reply right away from Mars saying- ‘You’re welcome. Happy to Help. 24*7’

This out of the world member of Sarah’s team, Mars, represents what some of us call, a robot; a software bot, who is a Multitasker, always Accurate, works Round the clock and is Superfast.

Robotic Process Automation (RPA) and AI are beyond buzzwords today. With the adoption of automation across business functions on the rise, what is important and goes beyond operational and financial benefits is that the virtual work-force made up of software bots seamlessly works hand in hand with humans.

A bot wears different hats for different stakeholders and can deliver a great experience across all with well-designed handling of the interaction between bots and relevant stakeholders.

During one of my interactions with an F&A associate, she said “I don’t like sending reminder emails, I think people must hate me that I am always on their head! Glad a robot can do that and take all that hate for me”. A bot can execute transactional processes in a touchless manner and take the mundane out of a potentially mundane task, so that the associates in the team can focus better on more important, judgment based activities. An RPA solution can work best when the bots are configured to direct exceptions to the right team/person instead of just highlighting the fallouts.

In a different scenario, where processes are not entirely rule based, software bots can work on parts of processes with smooth handovers between bots and humans by exchanging emails, and with the use of notification alerts, shared folders etc. just the way different people across teams work jointly on a process today.

For a team leader, a robot adds value with high quality execution of processes within defined SLAs. “Our teams are especially stressed during the book closing period, during month and quarter ends. I would want the robots to work full-time on some of the closing activities and help us avoid over time”, highlighted a team leader in a record-to-report function of a retail company. By automating various tasks as period end reconciliations, postings, and reporting, a multi-skilled RPA bot can reduce a lot of manual efforts during period ends and execute other transactional processes in the rest of the month. With the capability of assigning and re-assigning robots across multiple processes, a team leader can manage spikes and put the total robot capacity to its best use.

A business process owner can have clear visibility into the performance of the robots in terms of time taken, success/failure rates, reasons for failure etc., all in real time. Advanced reporting features that enable tracking metrics at a process, bot and transaction level go a long way in identifying areas for improvement and maintaining an audit trail.

In the case of employees, partners and customers, not only do they have their tickets, requests or orders handled faster, but they also have real time response and resolution to their queries through well-trained and integrated chat bots and email bots. Bots having the capability of handing over the chat or directing an email in case of unsuccessful resolution to a domain expert based on the context of chat are integral in providing a seamless experience to the customer.

As with any new employee on boarded, some bots ‘wow’ from day one and some bots experience a learning curve depending on the process rules, data quality and applications involved.

If a bot is configured to process a certain fraction of transactions (say 70%-80%) to begin with, it can be trained better with time to manage complex cases and unseen exceptions to enhance the automation rate.

Reporting features can help pin-point the reasons of failure if any, such as missing data or missing rule or application issues and enhancements can be done in configuration to manage them if necessary.

Process improvements based on identified areas can also lead to increase in addressable scope for automation in processes.

Comprehensive key user trainings help effectively manage the interaction and handovers between humans and robots.

With the right tool, approach towards automation and attitude to adapt, virtual workforce consisting of software robots can become a sweet reality!

Making banking a delight for the elderly


While old age brings with it wisdom from the experiences of life, it can pose difficulties in certain areas such as Financial Management. Managing life time earnings is fraught with challenges for the elderly. A decline in cognitive ability, general health, hearing, vision, or mobility impedes the ability of the elderly from performing even the basic banking transactions. Older people are often not digitally savvy and therefore may not be adept at using digital banking channels which need them to remember passwords, navigate through websites or understand call center menus.

There is a high dependence on children, relatives or caretakers to withdraw cash, check account balance or apply for things like cheque books on their behalf. This makes the elderly vulnerable to financial fraud. Studies show that the ability to manage one’s personal finances deteriorates with age and hence old people are at risk of messing up their finances or making bad investment decisions.

It is therefore imperative to offer the elderly appropriate banking experiences that help them manage their money easily and safely.

Growth potential

While there is a latent need for old age-friendly banking products and services, banks need to first assess the potential of the target market.

According to Census 2011 India, there are nearly 104 million elderly persons (aged 60 years or above) in the country, including 53 million females and 51 million males. (Source: PHD Research Bureau). This offers a huge potential for banks to expand their customer base and build a reputation for old age-friendly banking service. While banks and financial institutions do offer products for senior citizens, no one seems to have developed a comprehensive program to fully address the middle to low income older adults’ unique financial concerns.

At present, Indian banks offer some specialized products for senior citizens, such as Life Plus Senior Citizens Account ( ICICI Bank), Senior Privilege Savings Account (Axis Bank), Senior Citizens Account ( HDFC Bank) etc.

Drawing on the banking experiences of my father and other elderly family members and relatives, here are my thoughts on how to make banking not just simple, but even delightful for the elderly.

My thoughts

Dedicated customer service counters

Studies show the elderly prefer visiting the branch and interacting with bank personnel to using their mobiles or laptops for banking transactions.

Banks can

Provide exclusive counters for senior citizens in their branches.

Train the staff / tellers to listen to and understand the needs of old people, identify instances of potential fraud/financial abuse, and respond appropriately.

Provide facilities such as picking up cash and instruments against receipt, delivering cash withdrawn from accounts, delivering demand drafts, or submitting Know Your Customer (KYC) documents and Life Certificates at the customers’ doorstep.

Simplified product design

Design products with a transparent low fee structure, and benefits such as zero minimum balance, higher interest rates on deposits, credit and debit cards with zero maintenance fee, free bill pay services etc. These will be preferred to premium accounts with complex features and fee structures.

Provide easy access to simple financial products and services that can help older adults with low or moderate income to effectively manage a limited budget.

Customized channel offerings

Provide customized digital channel offerings such as easier navigation, large and bright fonts, simple text, and easy access and authentication.

Enable older people to structure their accounts online such that caretakers or relatives are allowed only limited privileges (for example, may only view).

Old age-friendly marketing materials

Development of old age-friendly materials such as marketing collateral printed in a large font.


Collect and analyze data such as spending patterns, products used, retirement savings, loans outstanding etc. and use it for rolling out customized products and services for their elderly customers.

Chart out financial protection guidelines that deal with the various aspects of financial health of such customers.

Provide retirement and financial management plans helping the elderly customers to plan their future cash flows as well as meet any unforeseen medical expenses.

Conduct financial seminars to create awareness of products and services relevant to that group and help them make informed financial decisions.

Design senior-friendly bank branches with adequate seating, non-slippery floors, rest room access, railings for support etc.

Final thoughts

With increasing life expectancy, banks and financial institutions have the opportunity as well as the responsibility to serve a growing elderly population (beyond 65 – 70 years) with the right products and services. Banks need to examine the holistic financial needs of their senior customers in the light of a changing retirement structure and provide a framework that ensures financial inclusion and access.

Ease of access to banking services, simply designed products at transparent costs, and the fulfillment of typical and atypical needs will make banks old age-friendly, and establish trust and the much needed financial security among senior customers.

Leverage Spend Analytics to Drive Digital Procurement Transformation

Digital technologies have rapidly evolved over the past few years but their adoption in procurement has been very slow. One of the key reasons is data quality and its (lack of) availability in real-time. As per a 2018 Deloitte CPO Survey, more than 45% of procurement leaders believe lack of integration and poor quality data are key barriers to the effective application of digital technology in procurement. This is because most organizations today are still using legacy spend analytics tools that have the following shortcomings:

Procurement leaders have very high expectations from spend analytics as they understand that it is key to transforming their function digitally. These expectations can be classified into four types:

Spend Reports – Most procurement teams want the spend analytics tool to help them respond to queries from business in a timely manner. They want to address problems as they happen and take corrective immediate actions. Besides control, they demand the best possible user experience and visualization tools that they can use on their own without any training.

Spend Insights – Procurement professionals want to manage and track their KPIs and also benchmark them against best in class organizations. They want to correlate spend data with market intelligence to find improvement opportunities. They also want to review both strategic and operational KPIs at one place to be able to see a complete picture and take comprehensive corrective actions resulting in savings and better compliance.

Spend Predictions – They would also like to leverage Artificial Intelligence to proactively monitor problematic trends and take preventive action to improve compliance and meet their savings target. A typical example of this is Price Forecasting. By leveraging procurement teams understanding of price influence drivers, we can build a category forecasting model that can predict price changes based on various drivers like currency, inflation, weather, raw material, demand index and more. This model can then be extended to suppliers to predict their item pricing specifically for tier 2 and tier 3 suppliers (assuming they are technologically challenged).

Spend Actions – Procurement teams also want to use spend analytics to generate actions that can drive automation and improve decision-making. This is a core requirement for achieving digital procurement transformation because first it reduces the need for cumbersome processes by bringing it required automation and second it truly transforms the user experience and business operations. For instance, if we can automatically create purchase requests for even Indirect goods/services based on historical trends and future business requirements, it will automatically ensure that the purchase is from right suppliers at the right contracted price and the transaction is pre-classified to right spend category.


Fig. Spend Analytics Evolution Roadmap

For successful Digital Procurement Transformation and to meet the above expectations, the organization should first ensure that its current spend analytics tools are aligned with advancements in data processing, visualization, AI and big data analytics. If not, it’s time for a change.