ChatBots have been evolving for almost 3 years now. However, by 2018, ChatBots stand at a cusp, with the technology and usage, ready to take off in an exponential way; powered by AI, NLP, and Analytics. ChatBots are intended to provide end users with a Cognitive, Conversational and Customized Interface to get their job done.
Initially, building the NLP and AI was expensive and an uphill task because of the corpus of data required to train the AI algorithms on. The amount of data required to train on, to understand the intent was huge, which only some companies had access to. However, with the democratization of these technologies and data becoming more and more accessible, it became an easier task for developers and Data Scientists, to develop the Bots.
Most BOT development platforms today provide the data corpus as well as the complex Neural Network algorithms to train the data on. Platforms like the Microsoft Bot Framework goes one step ahead and provides a full framework not only for Language understanding (LUIS) but also for API integration, Cloud Integration, Chat interface integration and also publishing tools. Similar development frameworks are provided by DialogFlow (Google), LEX (Amazon) and Oracle. Other than this, there are numerous other platform providers as well. Some companies have developed ready-to-use and deploy ChatBots for specific Use Cases and industries.
These are then licensed to end-user organizations. This drives up the cost of implementation for these Bots. Hence, to keep cost low and provide the flexibility of developing Customizable and scalable ChatBots, its best to develop one’s own BOT using some of the grounds up platforms likes the ones that are provided by Google, Amazon, Microsoft or Oracle. This also ensures that the NLP is built on the huge corpus of data that these platforms have access to and the most innovative algorithms in AI & NLP are used. This takes away the worry of an end user about whether to use Bayesian Gradient Algorithm versus pattern matching algorithms or others, as long as the results achieved have satisfactorily accurate and serves the specific purpose for which the ChatBot has been built. It also gives great flexibility in terms of API integrations and developing richer UX/UI.
While the technology and analytics for developing ChatBots are getting more accessible, the application of ChatBots still leaves a lot to be desired. For example, most chat Bots in Banking websites are still FAQ Bots which occupy a small corner or an insignificant link in a webpage or a Mobile App. This ChatBot mostly gives answers & FAQs based on the direct queries of the Banking customer, if the customer wishes to use it. The visibility of these Bots are so low that they are hardly used.
The customer still has to go through most of the website in order to achieve what they want to do. This defeats the purpose of providing a customized and personalized experience to the customer. While most ChatBots have the capability today, they are not implemented in the right way or in a manner that utilizes their full potential. ChatBot for today’s businesses, cannot restrict itself to just providing information inside a webpage.
The whole concept has to be turned on its head, and the Website should be embedded inside the ChatBot instead of the ChatBot being inside the website. The ChatBot here obviously cannot be just an information Bot but has to be fully integrated with the backend with the core banking system so that the Bot not only can serve information but also be able to execute all Core-Banking transactions that a customer would otherwise perform on a banking website or application. This will entirely change the customer experience and bring in the “WOW” element to it. Imagine a customer who in a normal scenario would have to go through and research hundreds of links and sitemaps to be able to achieve one simple task like linking ones phone number to their checking account. Instead of that, now, the customer interacts with the chatbot first, expresses their intent and exactly gets that done, in just a few steps, all within the BOT interface. This provides true personalization for the end consumers and gives an excellent user experience; at the same time reduces the cost for the bank, in terms of time taken by the resources to service their customer. This is when ChatBot evolves, from being just an Intelligent Conversational Interface to being a Cognitive Conversational Personalized Application or Assistant. Similarly, for other domains like Travel & Transportation, ticketing systems, retail transactions, Financial Services there is a need for ChatBots to evolve from being just Intelligent Conversational interfaces to being Cognitive Personal Applications or Assistants. It also opens up the possibility for Bots to use analytics and the latest developments in speech and picture recognition lead Robotic Process Automation to be integrated into the Bots, to provide a truly seamless end consumer experience.
A necessary metamorphosis from ChatBots to Cognitive Conversational Personalized Assistants or CCPA’s.