- Enterprises around the world, including in India, are turning to AI-powered chatbots for customer acquisition, knowledge management, and employee engagement.
- The good part of any chatbot conversation is that everything is recorded and digitized somewhere.
Last September, visa outsourcing and technology services company VFS Global, deployed chatbot-Viva-offering round-the-clock assistance to visa applicants headed to Australia. Powered by artificial intelligence (AI), the chatbot can decipher patterns from previous interactions, what customers and VFS Global consider as useful information.
"Viva radically decreases the response time by replacing interaction with call centers through chats. Also, it has helped us understand which queries to focus on to improve user experience," said Benjamin Boesch, digital and e-commerce head at VFS Global.
In India, Raymond has roped in Applicate IT Solutions' Sellina AI assistant to help, train and engage with the textile giant's 5,000-strong dealer network. Startups such as Bengaluru-based Senseforth AI Research are also providing enterprises with multiple options.
"Enterprises have millions of customers, hundreds of employees, partners, and all of them have millions of interactions in a single day. The only way enterprises can scale up and deliver a personalised and customized experience to all of them is by having conversational experiences," said Ritesh Radhakrishnan, chief technology officer and co-founder of Senseforth.
The startup has multilingual bots and a bot builder, which allows enterprises to customize bots according to their requirements. Once a bot has been trained for conversational experiences, enterprises can monitor their interaction using Senseforth's analytics solution.
While there is a lot of hype around chatbots, the technology is still in its nascent stages, particularly because the underlying technology, AI, which makes them so effective, is still taking baby steps. "The technology is in its infancy and needs time to mature. But, of course, as we move along it will pick up and improve," said Arup Roy, vice president-analyst at Gartner.
Despite chatbots, there will still be a natural conversation and dialogue with a customer, but then it won't be open and wide. There will be some boundaries, and users will be presented with some choices. It is done because when you leave it open, a customer can ask any kind of question. Roy pointed out that in a natural conversation, there could be multiple levels of intentions and questions baked into one statement. That is where the technology doesn't do that well, especially in terms of understanding the intent and processing it.
Also, chatbots work fine when an AI model trains five or seven intents or speech interfaces that allow bots to understand verbal requests and convert such patterns into actions. Beyond that, the bots struggle. Managing and fine-tuning them becomes very difficult. There are fundamentally three-four challenges that exist in the adoption of chatbots. Integration with the enterprise's system is one. "It (the chatbot) needs to connect to your back-end systems to be able to fetch that information. That connection and integration is still a long way to go. So, while the front-ends are changing quite fast through a lot of AI, the back-end is still slower," says Vijay Bhaskaran, partner (robotics and intelligent automation) at EY India.
Another big challenge faced by chatbots is the lack of good quality data. "You may not have enough volume of good quality data. Even if you have data, you will not have a good combination of people resources, who understand the business context as well as the NLP (natural language processing) domain for them to be able to carry out this stuff," says Roy of Gartner.
The good part of any chatbot conversation is that everything is recorded and digitized somewhere. VFS Global was using insights from Viva to understand what customers want. Similarly, an insurance or services firm can use chatbots to suggest products based on customer request.
Bhaskaran says there is a lot of opportunities as chatbots provide firms with customer insights that they never had. But all this needs to be synthesized to get the right amount of information. You need to have good data science and analytics to make it more efficient.
Gartner estimates that 85% of all customer interactions will be through AI-based chatbots by 2020. That estimate may not be off the mark, simply because, despite their initial limitations, chatbots are changing the way users interact with companies, and this will only accelerate as we go forward