Chatbots have arrived in the enterprise almost unannounced. In India, both businesses and start-up suppliers are figuring out how to do more with them
Earlier this year, ICICI Bank announced a contest among developers to create virtual mobile apps in almost all major areas of banking. Labeled ICICI Appathon, the contest offered a diverse set of over 50 APIs from ICICI Bank and its group of companies and partners, Visa and NPCI, to the contestants to create apps. While over 2,000 participants took part, two of the top three honors, including the first prize, went to applications for enabling conversations based on natural language processing and artificial intelligence.
The ICICI Appathon was just putting an official seal of approval on an undercurrent that businesses, especially those dealing directly with consumers, are witnessing globally. India itself has a host of start-ups who are creating such solutions for business applications—and not just banking.
Interestingly, this is one area where a big revolution is happening simultaneously around consumer space—with big techies like Apple, Microsoft, Google, Facebook and Amazon all creating newer capabilities—businesses are now rushing in to use them to improve customer service, enhance efficiency, and better understand customer needs.
These solutions, called chatterbots or chatbots, are computer programs that try to impersonate human beings while ‘conversing’ with people. Bots have been there doing repetitive jobs, such as web-crawling or posting/following on Twitter, based on automated tasks (scripts) assigned to them. In the last one year, the AI element has crept in with focus on natural language understanding. The ICICI Appathon winner, simply called Conversational Banking, is an example of that. Today, these chatbots are so popular, thanks to their use by consumer platforms, such as Facebook, Google and Microsoft that in common parlance, ‘bots’ now mostly refer to the ‘chatbots’.
To put it mildly, the quest for making machines understand and respond to natural language is not an entirely new phenomenon. In fact, the efforts have been on since 1950s when the first chatterbot actually came up in 1966. Called Eliza, it was created by Joseph Weizenbaum, with just 200 lines of code, to imitate the language of a therapist. While humans interacting with it knew that they were interacting with a computer program, many grew emotionally attached to the program during trials.
Unfortunately, with low hanging fruits of dramatic productivity improvements using IT taking precedence in the 90s, the thrust on artificial intelligence and natural language processing took a backseat. With analytics, especially text analytics, maturing as a field now, the focus is back on natural language understanding.
Today, AI is a trending topic. In 4 out of 5 cases, when media, analysts, businesses and developers refer to AI, they refer to natural language understanding and chatbot type of applications.
And yes, businesses are not just watching from the sidelines. They are actually leveraging it. In fact, it promises to be the new frontier in user experience (UX).
They are here, already!
Safexpress, an Indian logistics player, has at any point over 4500 GPS enabled vehicles traversing a distance of 600,000 kilometers. Using bots, it can now easily track every single shipment of a confinement of goods.
Safexpress is not alone. Oyo Rooms, a startup in the hospitality sector, is leveraging a bot built by an Indian AI startup, Niki.ai– to help travelers ease the process of searching and booking hotels. The company that was founded in 2013 is known as the Uber of hotels in India and operates a network of 6,500 hotels and 70,000 rooms in over 200 Indian cities. This bot accomplishes the task of searching and booking hotels seamlessly through a chat enabled interface.
Niki.ai’s bot uses natural language processing and machine learning and automates tasks, such as food ordering, cab booking, bill payment, through a simple chat. During the launch, Oyo Rooms’ product head, Anurag Gaggar, said, the rapid adoption of social and messaging apps has driven user preference for chat based interfaces.
There are also other banks like Kotak Mahindra Bank, Axis Bank and UBS Bank, testing chatbots. Many startups like Lenskart and Freshdesk are also leveraging the benefit of chatbots in their nimble infrastructures.
In June this year, Meru Cabs became the first Indian cab services provider to integrate their cab booking system with Facebook Messenger to provide automated cab booking facility. The Indian taxi aggregator’s entire range of taxi services will now be accessible to Facebook users through a service, called Meru Messenger Bot, which allows them to book cabs through keywords, intelligent chatbot and conversational questions and answers. The bot also enables the user locate a cab and check the fare details.
While India plays catch-up, global organizations are taking this bot business to the next level. From this year February onwards, an user in the US can simply ask Alexa to order an Uber, and she will request a ride right to your door. This bot developed by Amazon is a personal assistant that listens to you through a number of physical devices that use her (Alexa) as their brain. This was made possible as a result of integrating Uber API with the bot.
If hailing a taxi has become easier, ordering your favorite pizza is just a conversation away. Pizza delivery service, Domino’s, has launched a new artificially intelligent chatbot, called Dom on Facebook Messenger, using text or even a pizza emoji!
Domino’s biggest rival in the business, Pizza Hut, is still playing catch up. However, it is launching a bot with more interactive features soon.
Looks like these businesses are using bots as an alternative rather than a replacement, and companies like Microsoft, Facebook and a number of artificial intelligence startups are making it possible for them to do so.
Leading from the front
In March this year, Microsoft CEO, Satya Nadella, declared that “Bots are the new apps.” “People-to-people conversations, people-to-digital assistants, people-to bots and even digital assistants-to-bots, that is the world you are going to get to see in the years to come,” he added.
In September, Google came up with a virtual personal assistant and a messaging app, called Allo. Google will also announce tools for software developers allowing them to make chatbots that work on its own messaging services, like Messenger and Hangouts, as well as other social platforms, such as Facebook Messenger. Google will offer developers access to its voice and natural-language recognition software for both typed and spoken interfaces with their bots. Bot makers will also get access to Google Now, which will act as a conduit to data from people’s Google Calendar, Google Searches and browsing activity on Chrome.
Microsoft Bot Connector, on the other hand, helps bots connect with other communication channels, like Skype, SMS, and email. It provides a framework to develop bots. It also aims to forward messages from the Bot to a user, and vice versa.
In contrast to Google and Microsoft, Facebook has offered more than 11,000 chat bots but has kept it limited to its messenger application. This race of VPAs or simply put, bots, is worth billions of dollars and is expanding very fast globally. According to the Grand View research, the Intelligent Virtual Assistant (IVA) market was worth USD750 million in 2015.
According to Kamal Karnatak, Group CIO, RJ Corp (an Indian company which specializes in the food, beverages, medical, and education business), says that the VPA is not exactly going to become the primary web searching app in very near future, but also admitted that it will definitely be in long run as the human touch supersedes.
A sentiment verified by market research company, Technavio, who predicted that the market would grow at CAGR 54% between 2015 and 2019. Transparency Market Research predicts a CAGR of 31.80% from 2015-2022 and global market share worth USD5.1 billion by 2022. Another review by research company, ReportLinker, estimates a CAGR of over 38% from 2016 to 2024 to reach USD12.28 billion by 2024. However, all researches unanimously agree to the fact that APAC has the highest growth, with its root in India and China. As per Technavio, APAC’s will see a growth rate close to 68% from 2015-19.
According to Aneesh Nair, CoFounder & CIO of NDTV E-Commerce, “VPA is the future of self service and it is terrifyingly convenient.”
It is terrifying, both in terms of convenience, and in terms of security.
Automated bots of hijacked computers will drive cyber attack innovation in 2016, according to a report by security firm, Radware. Radware believes that in 2016, security will become less about people and more about machines, as bots become the attack weapon of choice. This is based on the fact that over 90% of 300 companies surveyed, experienced a cyber attack, and half of those attacked, experienced short, intensive automated attacks – up from 27% in 2014.
There are also serious concerns about what and how these bots are learning.
In a recent incident, Tay, a bot by Microsoft, was programmed to behave like a millennial, learning from tweets it interacted with on Twitter. However, in less than 24 hours, it started tweeting objectionable texts and was taken offline.
What if the technology suffers from an outage? Remember the movie, Her, where the protagonist was so dependent on its AI assistant that the AIs’ unavailability for an hour or so left him in a state of complete confusion, emptiness, and lack of understanding.
Essentially, chatbots collect information from different users and that is how they gain the knowledge to provide relevant information. However, this information needs to be constantly updated and checked for its authenticity.
Verticals, such as finance and healthcare that are highly regulated, and collect sensitive information from users, must be extremely cautious while using bots when it comes to collecting data from their customers.
Additionally, where is information collected by bots, stored? Is it secure enough? Is it hacker-proof? These are some of the important questions that organizations must ask.
Until then, let’s bot and watch!