Bots – Changing The
Customer Experience

Author: Sunil Kapadia


Tech Lead – Innovations

How do conversational bots make a sensitive & rational conversation to a customer?

Artificial Intelligence is not a new term for all of us as the first movie on the same topic was released in 1927 so almost 90 years from now. The great director Steven Spielberg had also directed a movie called A.I. in 2001. Why am I discussing the movies all of a sudden? The reason is the concepts of AI were made known to a lot of people through these kinds of sci-fi movies. Since then the AI as a platform is improvised a lot. One of the latest implementations of AI is the rise of Smart Bots.

The meaning of Bot is different for everyone. Like for a game programmer bot is an artificial character in a game. While in terms of cybersecurity, botnets are a group of hijacked computers which is used by hackers to attack websites or may be something else. For us, the BOT is a virtual assistant like Siri or Cortana or Google Assistant.

Does that mean Bots are new?

Certainly not, in 1966 Joseph Weizenbaum wrote a ‘bot’ called ELIZA which was very famous and it was able to give answers like a therapist by typing to it. nowadays bots can do the same but with a pinch of NLP (Natural Language Processing) based on LUIS (Language Understanding Intelligent Service) from Microsoft or LUIS from Google, Microsoft Cognitive APIs or Google Cognitive APIs, they are much smarter and most importantly they can mimic the human touch while talking to the user. As many of us don’t like to chat to Bots.

Powerful insights from NLP were a major barrier in the past but there has been a major change in technologies in last 5 years. Because of distributed computing and intensive research by the academicians and professionals, NLP now can provide a very details information and with granularity and accuracy. Let me explain how NLP can give us a better insight like a human brain. Let’s say: I have taken two bananas in the morning as a breakfast. Now for us, it provides information like the guy is health conscious and has taken two bananas which are a fruit.

Now let us see how NLP will provide information about his sentence:

I – Entity – User of the app

Two – Number of entity which is food

Banana – Fruit entity a part of Complex Entity Food

Morning – Because of time it is breakfast

Overall Personality – Health Conscious

So, now if we compare the results are pretty same. So conversational bots which are empowered by properly trained NLP engines can give us a better experience in terms of chatting experience.

The essence of human communication is an effective analysis of communication and natural language. The moment NLP engines are capable of mimicking the same human behavior they can take the conversational BOTs to a whole new level of accuracy while providing the most human-like experience to its users. Up to now, we have seen how NLP can help the users in the most humane way. But if the user sends a photograph or speaks a sentence how would the BOT understand those things. With the help of cognitive services like Vision, Speech to Text, Text to Speech, Speech analysis, Speaker Recognition, Knowledge Graph API, Entity Linking, etc. can help getting useful and vital information from Images, Voices, and Texts. This entire plethora of services can remove the limitations of BOTs and give them the edge of human communication and analysis skills.

The latest edition to these services is the object detection service released by Tensor Flow which can detect multiple objects from a single image which can open new doors of smartness for the BOTs. As now the BOTs can be fine-tuned based on the human response in all the three senses which are utilized by humans to analyze and process information (Vision, Speech, Image, and Text). In order to understand it better, a video would do the trick.

In the above video, there is an artificial intelligent headset which allows the user who is blind to see the world and analyze the information of surroundings

More Perspectives