Here’s a common scenario in the financial sector – high volume customer support tasks are being standardized and executed by chatbots or software applications that can be scaled up accordingly and are streamlining operations. From insurance discovery, buying and renewal to handling day-to-day customer transactional inquiries, AI-powered chatbots have moved out labs to redefine banking processes.
Earlier last year, ICICI Bank became the first bank to roll out Software Robotics with 200 plus robots carrying out 10 lakh banking transactions every working day. The end result – customer response time cut down by 60% and a surge in productivity. ICICI’s MD & CEO Chanda Kochhar remarked in statement on the paradigm shift, “We have created new paradigms in the financial services industry by taking the lead in introducing path-breaking innovations such as internet banking, mobile banking, Tab banking, Touch Banking branches and banking on social media.”
And then there is Digibank, India’s first chatbot-staffed mobile bank and fields all banking-related queries in real-time. An offering of Asia’s largest bank, DBS Bank, the chatbot is powered by New York based start-up Kasisto that trained its AI platform KAI on the millions of banking queries fielded by customers.
Bots have emerged as the preferred interface as more and more searches shift from search to voice. While globally, banking bots have moved beyond answering transactional queries to full-service mode, Indian banks and insurance sector bound by regulatory norms is still in a nascent stage. Though, reportedly talks are on about Axis Bank launching its own chatbot in partnership with Singapore based fintech startup Active.ai. The chatbot can function on the bank’s native app and on social media apps like FB messenger as well. Rajiv Anand, Executive Director, Retail Banking at Axis Bank shared the idea behind launching an “intelligent banking chatbot” is – “engaging customers in a personal way, boost repeat value of transactions and to become an early adopter of conversational banking in BFSI space”.
Analytics India Magazine gets Rajesh Kamath, Head of Financial Services Solutions and Incubation, Incedo Inc. to weigh-in on why finance is betting big on chatbots and the technology driving it:
Earlier, the broad set of activities that got bunched under the term ‘AI’ were reserved for the labs, and if they ever saw light of day, were severely constrained by what the technology of the day could deliver at a viable cost. AI, always had a lot of value and contribution from computer scientists in terms of algorithms and maturity. Primarily, well understood by scientific community, it was used extensively and successfully in complex machines, robotics and scientific applications such as vision, medical devices, space programs et al.
Previous surges in AI had unrealistic expectations of immediate consumer applications that could never be accomplished given the limitations of data and techniques available at the time. Within the last five years, the combination of enormous amounts of data and database technology to effectively utilize it along with increase in computer horspower to process it, have enabled AI to move from scientific and academic usage to enterprise software usage and become viable as mainstream business solutions.
This time around, the ‘AI’ movement seems to have tailwinds in the form of a few critical enabling and supporting factors:
- Technology driving AI capability at the right (low) price points (including the availability of platforms from the biggies in the field like Google, Microsoft, Amazon)
- Mainstreaming of practice — slowly building critical mass of practitioners who leverage these platforms
- Mainstream customer interest and demand reflected in real world ‘everyday’ use cases
- Increasing mass of data that is waiting to be exploited which cannot be done by solely human means
- Changed consumer expectations from what is doable using technology, which is further driving innovation in a secular manner (e.g. Alexa)
Advancement in Natural Language Processing (NLP), Deep Learning fuelling the rise of banking chatbots
Chatbots are undeniably the next-big-thing in the evolution of financial services. The efficiency of bots has increased considerably and factors like automation, voice-to-text and text-to-voice conversion, Natural Language Processing (NLP), machine learning and analytics, have played a huge role in fuelling the rise of chatbots, especially in the financial services sector. Incedo, co-invests and partners with financial services clients to help them smarter solutions through the Incubation Labs, building accelerators and Proof of Concepts (POCs). The clients and prospects often discuss automation solutions to their problems, and develop automated financial reporting process using low footprint off–the–shelf technology.
How Incedo is leveraging the chatbot advantage for financial service clients
For clients in the asset and wealth management space, Incedo is helping them improve their cost structures or refining the customer experience by letting them get to information faster, or do transactions through the chatbots. Incedo’s most interesting use case that is currently in the pipeline also integrates Alexa to allow users to use voice commands.
Some of the specific use cases in the pipeline are:
- A chatbot for internal users to get important financial and management information about their business
- Chatbots to help financial advisors access information about their client portfolios
- Chatbots for self-service use cases in a variety of industries
The evolution of bots
Bots did a little more than answer basic queries. Nearly 80 per cent of organizations want to be empowered with chatbots by 2020, according to to Gartner report that makes the time just right for bot stores. Apparently, over 50 per cent of a salesperson’s time is taken up in fil
ling forms. With automation, they can focus on the core areas and achieve higher levels of work. Besides, if a company owns or generate tonnes of data, features like NLP and AI-based knowledge engines that understand voice, convert voice to text and do a sentiment analysis can prove handy.
Chatbots use a combination of machine learning and NLP to predict with great accuracy what a customer needs. For instance, when it receives a call, it can check what the customer was doing five minutes before he called and predict the possible issues. Enterprises will be able to learn faster and provide customized solutions based on behavioral and predictive analytics.
Try deep learning using MATLAB