Chatbots got a boost during the pandemic when remote working and work from home became a norm. Often, chatbots filled in for employees as far as possible. Now, offices have either fully opened or at least follow the hybrid model. Whatever it is, chatbots have stayed on.
Companies have realised that chatbots can do much more than what they did over the last two years. They are being positioned as digital alternatives to human agents in call centres and retail outlets. They also answer FAQs (frequently asked questions), which make them a storehouse for customer data. This data is being used by the companies to understand customer requirements, package their offerings accordingly and thereby improve their revenue.
A further step in this direction is the integration of AI, or artificial intelligence. AI's accuracy and speech recognition software have given a boost to chatbots and voice assistants. The regular chatbots function on sensitive keyword data. But the AI-powered conversational chatbots bring in additional features like NLP or natural language processing and AI algorithms. Together they process the dialogue in a conversation, which can be used for queries raised by customers.
"There's huge potential for growth of chatbots. India is a $3billion market for chatbots. Globally the market is estimated to be $100 billion by 2026," said Ankush Sabharwal, founder and CEO, CoRover at the 3rd NASSCOM XperienceAI Virtual Summit 2022. The company focuses on conversational AI platforms and creates chat and voice bots. It has incorporated a human-centric aspect in the form of virtual assistants that look like humans, somewhat like a Digital Human. "We usually build chatbots for enterprises. We are working to provide virtual assistants in 15 Indian vernaculars. Outside India, we are working on different slangs. In short, the services are more personalised based on demographics," added Sabharwal.
Diverse verticals are customising voice offerings. Large e-commerce brands have voice shopping solutions under their umbrella. This includes brands representing consumer packaged goods, durables and food. Consumers have begun to tap this option as it is hassle-free and contactless. Other than that, healthcare start-ups and job openings are relying on chatbots.
Private companies and the BFSI (banking, financial services and insurance) sector are keen to leverage the conversational AI platform. They are exploring the possibilities of automating messages for a better customer experience and outreach. Fund collection as well as customer-centric services need to be conveyed in the local language for a personalised connect. "We have a voice-based automated conversational AI platform. It is tailored as per specification. Voice has its own complexities, it also has authentication parameters," added Ganesh Gopalan, CEO and co-founder, Gnani.ai.
Automated conversations have to be in the language of the people. Vernacular recordings can be used to train AI solutions. They can also be used to build datasets. In the absence of datasets, voice agencies can use localised audio and build mixed language NLO or Natural Language Query through open libraries. To think of it, the manner in which automated systems talk to people in their language and develop authentication for the same is both a challenge and an opportunity.
We may live in the smartphone era, but we need to remember a large chunk of the population still uses basic or feature phones. Automated conversations of chatbots are being designed for deployment for a wider reach. Voice bots help convey service-transaction messages on feature phones. A case in point is the Unified Payments Interface (UPI) payments for feature phone users launched by Reserve Bank of India (RBI) earlier in the year. RBI has incorporated Interactive Voice Response (IVR) facility. The UPI payment through pre-defined IVR numbers would require users to initiate a secured call from their feature phones to a predetermined number and complete UPI on-boarding formalities to be able to start making financial transactions without internet connection.
RBI, India's central banking institution, which controls the monetary policy of the Indian rupee, also launched DigiSaathi, a 24x7 helpline for information on digital payments. National Payments Corporation of India (NPCI) has initiated it on behalf of a consortium of payment system operators and participants including both banks and non-banks. Information about any of the digital payment products and services is available in English and Hindi. DigiSaathi can be accessed through the website, chatbots and WhatsApp.
When we look at the AI scenario, it has probably been around for around 50 years. "The first 50 years of AI is similar to the first half of the chess board. We seem to have migrated to the second half of the chess board, with the exponential rise of AI that has happened over the last 10 years," highlighted Srikanth Velamakanni, co-founder, group chief executive and executive vice chairman of Fractal. This could probably be deduced from the fact that AI integration is increasing and spreading across verticals. The adoption of voice search optimisation and virtual assistants took off with Apple's Siri and Amazon's Alexa. The use of voice depends on the accuracy of AI and speech recognition. The human element probably happened in 2016. The demand for accuracy of models came and this segment is slated to grow and give a new dimension to the internet. Perhaps here lies an opportunity for voice agencies. Their community could grow as the voice gets louder.
Diversity must be at the heart of equitable AI development
As artificial intelligence becomes a part of everyday life, developers need to make sure that the models it learns from provide an accurate reflection of the real world.