Like most financial services companies, Mumbai-based stock broking firm HDFC Securities has been an early adopter of technology for day-to-day operations. The firm, a subsidiary of private sector lender HDFC Bank, has partnered multiple artificial intelligence (AI) vendors, Alexa and Google Home to deliver differentiated services to its customers.
In an interview with TechCircle, NK Purohit, head of digital strategy and analytics, spoke about how AI is being utilised within the solutions portfolio, how an increase in compute power has helped gain real-time critical insights into market indices and how the firm plans to take the social media game forward.
Purohit also shed light on how gamification plays a huge role in the acquisition and retention of millennials and how blockchain can increase volumes in the securities market. Edited excerpts:
What is the future of the banking, financial services and insurance (BFSI) industry going to look like, considering it has the highest uptake of the latest and upcoming technologies?
We will be leveraging technology to know more about customers, create personalised customer-centric solutions and offer those solutions without the customers asking for them. AI/ML (machine learning) is hugely used now and it is jargonising the BFSI industry.
Our endeavour has been to make the process as simple as buying a shirt or a book on Amazon. We want to make it purely conversational, want to be the platform where the customers are and where the party is happening. HDFC Securities is already present on Facebook, WhatsApp and Internet of Things (IoT) devices, among others. The customer can transact with us in their channel of preference rather than them needing to download an application or visiting the firm’s website.
How has your conversational AI bot Arya played a role in the growth of the company? Any plans to move to regional languages?
We want a solution that is relevant for all customer types and language is just a medium to converse in. We will use Hindi and other key languages. When we launched Arya a year ago we had a conversational accuracy rate of 75 % and now we have an accuracy level of 98.5%.
In terms of the infrastructure, there is a primary AI/ML-based engine and an MLP (multilayer perceptron) engine.
The AI/ML is primarily used to offer relevant products to the customer while the MLP is used to understand the conversational intent and provide the relevant solution. We have achieved this through three fintech partnerships.
At the backend is my complete transactional platform powered by Tata Consultancy Services (TCS). It is one of most robust platforms that has an uptime of a 100% even on days such as those falling on 23 May, where transactional volumes are at their highest.
What are the different AI partnerships you have entered into?
We are in partnerships with three AI platforms: Niki.ai, Active.Ai and Senseforth.AI.
Niki.ai is primarily for e-commerce-based transactions while Senseforth.AI powers the conversations through resolutions and queries. Senseforth.AI also helps with IoT devices such as Alexa and Google Home. Active.Ai is my core AI engine for queries related to transactions. So we have leveraged three different players for three different implementations.
We have also set up an innovation lab within HDFC Securities, where we work with fintech companies on various use-cases. Depending upon the outcome of these use-cases, we take certain use-cases into production.
Mention a case study that stemmed from the innovation lab and was successful when implemented in the real market.
When a customer opens an investment account, the person wants to invest Rs 5 lakh and opens a demat trading account with me. We needed to engage a welcome call to the customer and understand the risk profile, level of understanding, knowledge in investment and guide the person to make investments through my digital platform -- this is termed as customer onboarding.
Usually when we call, the customer is busy with the onboarding process, or if the process is done, we don’t really know how much knowledge the customer has imbibed.
For the simple customer onboarding process, we worked with Activ.Ai on a conversational onboarding platform that activates the moment the customer gives me an intent of opening an account with me. So the moment the customer comes to the platform and gives an intent of opening a lead, I start engaging through this platform.
I explain the process, various requirements and explain any queries related to investing to the customer. Till the account is opened, I engage with the customer in terms of their knowledge of the industry, preferences and the customer need not download anything for the process.
We have created an app that acts like an application through a simple link without the need for downloading anything. We handhold the customer from the first step of knowing about mutual fund or equity till he becomes an advanced trader.
How much of an impact has gamification had on your solutions portfolio?
At each level of the customer’s lifecycle, we have introduced gamification, so that they can educate and empower themselves. The customer can also share their achievements and critical steps in the journey on social media.
As of today, 90% of my customers are getting onboarded through social media. Hence, the customer can get onboarded right from the day he shows interest. Customers can get onboarded via the channel of his choice, platform and product of his choice.
Gamification is extremely impactful for us. We have realised that for millennials, life is all about gamification.
It has increased our new client activation by two-fold. As far as BFSI is concerned, gamification is in trend. And it applies more to the investment space. In India, we use a simulation model of gamification to enhance customer experience.
How is the trend towards social media channels changing? Is it the future of the BFSI segment?
We are present on a host of social media channels including Twitter, Facebook and IoT devices. Right from acquisition till transaction, we can enable through our channels for the customer.
Social media is gaining ground and we expect the contribution of social media to increase with the announcement of the ‘WhatsApp for Business’ last month. People may not invest directly through social media but they use the platforms to know market prices and to understand the market well. Actively engaging the new social media platforms will help in making new acquisitions easier and help in building awareness.
Will the increase in compute power help the industry predict the markets faster and more accurately?
AI and ML will bring in compute power to the market. Before the advent of AI, it used to take a lot of time to get the relevant information -- AI and ML will improve compute power and we will get inferences out of market movement very fast.
Somebody who is looking for real-time information will be equipped with critical data in no time. People will also be able to pre-empt market movements. However, the individual should be intelligent enough to make sense of and utilise the information in the right manner. If I believe in the bull cycle and you believe in the bear cycle, what critical insights are good for me might not necessarily be helpful for you.
Our platform, built with AI/ML in the background, accumulates data from multiple research houses in the country and provides a transparent view of the data. The platform also has a real-time sentiment analysis feature, which helps make decisions based on fundamentals easily.
There has been a lot of talk about blockchain in BFSI. How can it impact the financial security market? What does the industry need to do to get the most out of blockchain?
Blockchain, as far as banking is concerned, is taking some shape, but the securities market is still in its infancy in blockchain adoption since that would need the entire ecosystem to come together in real time.
Blockchain is an ecosystem and not a technology. What IMPS (Immediate Payment Service) and CTS (Cheque Truncation System) did for banking is what blockchain can do for the securities market.
There are a few use-cases of blockchain in the industry but we have not received any regulatory approvals for implementing the same. It needs broader participation of all stakeholders and intermediaries involved in the transactions, including the exchanges.
Blockchain is fundamentally well-suited for the securities market. My wildest guess, if the regulations are in place and the industry comes together to implement blockchain, is that it can double the volume of transactions.
All of the players need to develop a product consensus regarding the merits of the blockchain and remove the apprehensions surrounding the demerits of the blockchain. The companies in the domain are still not looking at the worth of blockchain and are only focused on the peripheral cost-saving mechanisms.