AINews & Analysis

Why Indian Financial Institutions Need To Harness Conversational AI

Banks can use conversational AI to answer queries in real-time and consumers today do not mind using self-service to get their answers.

virtual assistants

The way consumers engage with brands has undergone massive transformation over the years. Digital natives form a third of the population in India today, and no surprise that they are comfortable shopping, booking shows and ordering meals or cabs on their smart phones. Most consumers today are online 24×7 and expect businesses to cater to their needs anytime anywhere. And banking is no exception.

In such a scenario, it is a given for banks and financial institutions to adopt the latest technologies to engage seamlessly and efficiently with their consumers. In fact the rising popularity of conversational artificial intelligence (AI) or chatbots could be attributed to the need for businesses such as financial institutions to become more efficient, productive and engaged in today’s day and age. For the uninitiated, a chatbot is an AI software that can simulate a chat with a user online.

BFSI’s gain from conversational AI

Banks can use conversational AI to answer queries in real-time and consumers today do not mind using self-service to get their answers. Consumers are always in a hurry and need prompt answers, and they can do so via a chatbot. These common queries could be related to questions about a branch location or ATM locations, credit limit or the details of a recent transaction, etc. Take the case of IRIS, an intelligent virtual assistant rolled out by IndiaFirst Life Insurance. IRIS provides a live chat experience so customers can ask any question and get a quick response in real time. It’s a boon for the sales force as well, as they can shift focus on other critical issues for the organisation.

Chatbots also come in handy when financial institutions need to send in customized push notifications about a new product or offer, say a credit or a new loan scheme.

Customer support, when automated, saves agent time and banks are able to streamline their operations better. Efficiency and productivity increase. A good example is that of City Union Bank’s customer facing chatbot, Lakshmi. Thanks to Lakshmi, the bank has been able to handle more than 10,000 customer queries in a week at 92 per cent accuracy. The bank’s operating expenses have come down by 30 per cent as a result.

With so many uses, it comes as no surprise that many Indian banks and financial institutions are rolling out conversational AI. A recent State of Service report by Salesforce says the use of AI and chatbots by customer service teams in India will grow by 90 and 118 per cent respectively, in the next 18 months.

However, it is important to bear in mind certain considerations before rolling out conversational AI.  Some of the risks that come with introducing chatbots raises critical questions on the process adopted to build the bot in the first place.

Risks and challenges

Any technology meets with some initial challenges, and conversational AI too comes with a few risks. Banks need to ask themselves if the platform they choose can address these challenges.

Banks and bot developers need to identify the exact need that the chatbot will be able to meet. Identifying pain points is half the battle won. When banks identify a problem, they can then lay the foundation for a chatbot that is specific to the customers’ (and employees’) needs.

For instance, BirlaSunlife has rolled out an insurance-specific AI solution, Disha. The bot manages over one million insurance policy queries in a month. The problem the insurance firm faced was that they received a high number of policy requests towards the end of the financial year (January-March). The answer was Disha, which gave them direction. The chatbot is accessible on multiple channels as well.

An important consideration while rolling out conversational AI is to address the question of when the decision making is transferred to a support executive. Also, understanding what users are trying to do/ask is very important. That’s where the question of contextual and semantic understanding comes into play.

Data complexity

Data is a key element of the AI revolution, and a conversational AI solution may only be as good as the data that comes in. According to a Capgemini-IBA report entitled ‘AI Revolution in Indian Banking’, 78 per cent of banking executives surveyed cite the lack of right data to implement AI.

Traditional systems may not exactly be user-friendly and information is siloed. Owing to the lack of integration, bots may not be able to pull up information in real-time. The Indian banking report also underlines the need for AI to be integrated into core business systems. It adds that “plug and play implementation” are just not enough.

Other challenges

The question of regulatory compliance is another challenge, apart from the question of scalability. In many cases, bot platforms may not be able to adopt shared learning models. This is because learning from a certain use case cannot be applied to another one.

Delivering an omni-channel experience is now the holy grail of CX. Going omni-channel means being present where the customer is, and delivering a seamless and unified experience across channels. So, the key consideration for a financial institution would be to ask themselves this: should they build a bot for every channel? Then, what about the costs involved? Also, there’s the aspect of handling security standards across all the channels.

Rewards outrun risks

Experts however point out that these challenges are bound to disappear as financial institutions discover the massive benefits of embracing conversational AI. In fact, choosing the right platform may be half the battle won. Although it is easy to deploy chatbots, getting accurate insights into the customer and leveraging the right engagement model will be crucial for its success.

Financial institutions that play the waiting game run the risk of being left out of the race. It’s better to hit the road running so the challenges can be met head on, while preparing you better for the long haul.

(The author Ram Menon is Founder and CEO, Avaamo)

Leave a Response