Echo Dot (3rd Gen) - Smart speaker with Alexa - Charcoal

Use your voice to play a song, artist, or genre through Amazon Music, Apple Music, Spotify, Pandora, and others. With compatible Echo devices in different rooms, you can fill your whole home with music.

Buy Now

Wireless Rechargeable Battery Powered WiFi Camera.

Wireless Rechargeable Battery Powered WiFi Camera is home security camera system lets you listen in and talk back through the built in speaker and microphone that work directly through your iPhone or Android Mic.

Buy Now

Bots with brains isn’t an oxymoron!

0
151


Nikhil John

I could start this off by calling out chatbots and dissing them about how incompetent they are and how much better they could be with a solution at the end of the article and a link to go and check out our offering. The article would then go on to be titled as a sales pitch with a bit of trivia in the middle and we could all forget about it the next morning. I will still leave a link at the end to check out our offerings but let me start off by commending chatbots for what they are and how diverse its application is in today’s world.

The Chatbot market was valued at 17.17 Billion in 2019 and is forecasted to grow to a whopping 102.29 Billion by 2025. Now that is saying something! There’s a big need to eliminate mundane tasks where manpower could be put to more productive use. Chatbots are all the rage today and with so many entrants in the market, each addressing or specializing in solving a narrow problem for a specific industry there’s no shortage of finding one based out of your neighborhood.

Having a chat bot in your customer or employee experience arsenal is table stakes today and enterprises have jumped on the digital transformation bandwagon with both feet.. But that isn’t necessarily a bad thing. Remember cloud adoption and how tough it was for organizations to wrap their heads around it? Even though cloud adoption hasn’t taken over the world yet, it sure is heading in that direction and even getting here took well over a decade. My point is, it’s necessary! From both the customer experience front and for the businesses deploying them. The only catch is that there is a huge gap between what we expected from these chat bots.

I’ve come across stats that state, chatbots decrease the waiting times for customers to get a solution and that they are great tools for lead generation for a website which is absolutely true. There are far more that claim improved ROI and better customer experience which leaves me thinking, does it really? Having encountered many more bots that contradict these statements and many others like them, I decided to get to the root of this. For that we must go back in time, to the first chatbot and why it was created.

The curious case of the Chat Bot!

The first chatbot ELIZA was created by Joseph Weizenbaum in 1966 with the intention of it mimicking human conversation. The chatbot worked by feeding words in a computer which was then paired with a list of possible scripted responses. The bot used a script that simulated a psychotherapist. That script proved to have a big impact on Natural Language Processing and Artificial Intelligence and saw the rise of multiple copies of the same popping up at various universities around the United States.

ELIZA set the stage for a tonne of other bots to follow. There was:

PARRY

Jabberwacky

Dr. Sbaitso

A.L.I.C.E. (Artificial Linguistic Internet Computer Entity) all of them with a specific use case in mind and each a little better than before. Here’s a link if you’re interested in the history of chatbots.

But a common thread amongst all these chatbots is that they soon started to disappear. Where? Well that’s not clear! Think of it, this rings true with even the chatbots of today. How many chatbot companies have you come across and how many iterations of them do you see every now and then?

Think of it this way. Every chatbot is deployed with a certain use case in mind. For example ELIZA was programmed to simulate a psychotherapist, some chatbots to be customer facing, others to spot visitors on your website and convert them to leads. And then you see something new pop up either doing a task better or claiming to improve a certain area of your business and then again we watch this cycle all over again. Now I’ll leave with you a question that you might find an answer to by the end of this article.

What do you think about this curious case of these missing chatbots? Are they a classic sign of technology being advanced and at the same time too early for its liking or a case of them disappearing to usher in a new wave of better more capable technology?

So when did it’s business applications start? Why did companies start using them? Were chatbots ready to take on such a responsibility?. The answers are uncertain. No one can really pinpoint the first chatbot deployed for businesses but what is clear is that the rage started when We Chat deployed its first chatbot back in 2013. The story behind the Chinese giant and the applications of it’s chatbot is quite interesting and this article is worth a read if you’re interested. Three years later it was Facebook that came out with a similar offering giving its customers an opportunity to deploy chatbots on messenger to interact with their clients. This then led to the rise of multiple chatbots popping up from every nook and cranny of the world. All claiming to be good at what they do some even specific to a business function or industry. These chatbots were clearly not ready or as polished as they are today. But that gives you a rough idea on how ELIZA led to the business bots that are deployed today.

Let’s dig a little deeper and get some things straight! There are two types of chatbots as of today. And this is something you should probably be aware of if you’ve made an investment in them or are planning to.

The Basic Bot:

This is what the first bots probably looked like. With a basic script at its foundation and a bunch of keywords that were necessary to generate a response, these monolithic bots were hardwired this way and were quite easy to set up the first time around but any changes to the script and addition of keywords would most likely require a developer to write a new set of commands for it. And this was revolutionary? Yes, it was! Perfect for scenarios where organizations could predict the flow of a conversation and script responses in that flow. But like most cases in today’s businesses things tend to get a bit more complicated than that. If the bot was ever posed with a question that did not contain a keyword it would usually tell you that it was unable to understand what you just said and ask you to rephrase the question. You see this happening a lot today and in most cases where you’ve opted for a free bot.

The Smart Bot:

These bots, on the other hand, rely on Artificial Intelligence to make responses more conversational and to make use of the conversational history to get better with time. With disciplines like Machine Learning and Natural Language Processing, these bots are better equipped to understand the randomness of human conversation and give fitting replies to questions asked. The Smart bot like the basic bot uses a sort of script or a knowledge base/FAQ to fetch most of its answers. Training these bots is a bit easier now as the knowledge base can be updated with new data and the bot can be trained to make sense of new questions that may arise. What’s more, is that it learns from past conversations and tries to answer questions more efficiently in the future. But how often do organizations update their FAQ’s? Apart from the organization, there’s probably no chance that the customers are always up to date about the changes that have taken place or the additions made to the product or service offering. And even if that is not completely true you can be sure of the fact that there will always be a delay in updating the bot itself about the changes made. The bot in such cases will utilize the FAQ or knowledge base that is already existent to respond to your customers. I’m sure you’ve witnessed it before. A bot or even a human representative pitching a great deal to you. You’re convinced and as you decide to take action you’re told that there was a misunderstanding and that the plan on offer was for the previous month. Bad customer experience right there! But hey, it’s a small price to pay for the actual benefits that the chatbot brings, isn’t it? Now the price of deploying a smart chatbot could vary depending on the set-up you opt for and where in your business you plan to deploy them. Here’s a chart that could give you a better understanding of it:

Source: Mobile Monkey

As you can see here even the smartest of chatbots require continuous maintenance to carry that image of a smart chatbot. And depending on where you are on the price scale it’s safe to debate about the claims made on the ROI they offer. What’s more, is that stuffing tonnes of information and scripts into a bot isn’t necessarily making it smarter. Your chatbot will now have to be trained to understand the context of that information and the good old trial and error method is going to take time. Then comes in the maintenance and depending on the type of work that needs to go in, you can be sure to see these costs increasing.

Known widely to the world, these are the two types of chatbots that exist today. But what if I said there exists a chatbot a lot more advanced and a lot more capable than the 2 mentioned here? What if said it isn’t even a chatbot, just a plugin that could work straight on top of the bot that you have deployed?

We call it the Augmented Bot!

So what’s the catch? Think of it this way, you know how Peter Parker from the Spider Man franchise whose a complete nerd with no athletic prowess gets bitten by a spider and gains super strength and the ability to climb walls. Similarly, think of iEngage as the spider who bit your bot. Now instead of crawling all over your PC and leaving cobwebs on your keyboard, it heightens your bots reception towards conversations and the ability to make sense of what’s being said, by whom and why. Lining conversations precisely with webs to connect multiple conversations with context, making it easy for your bot to just follow each string and make sense of the information stored. Each conversation is wrapped and thereon connected to the next. Now queries are easily identifiable and answers are that much easier for your bot to spin. In reality though just a bunch of APIs that plug-in seamlessly into the chatbot framework you have deployed. At iEngage, we believe that chatbots could perform way better if they were augmented by some Human Intelligence, instead of doing all the heavy lifting themselves. We believe that a bot can only behave more like a human if they are provided with the right information and if it’s able to make sense of all the information coming in.

So what’s the benefit of an Augmented Chatbot?

The augmented chatbot is capable of keeping track of a conversation across multiple platforms whether it is your website, your social media handle or an in-app chat interface, thereby providing a more personalized chat experience. The bot can pick up the conversation from where it last left off. There’s no need of updating FAQ’s as all the information is stored centrally in a knowledge graph format ensuring the bot is able to make sense of the conversation and retain the context in which questions are asked or things are said. It also listens in on the conversation from sources such as emails, slack or other platforms of your choice to stay updated. It could also communicate if you choose to, with bots deployed in other areas of your business to give your customers a seamless experience and you some peace of mind. Irrespective of the number of ways in which a question is asked, the iEngage module will enable the bot to give the right response and if the situation arises that it isn’t able to answer a question it will listen in on its human counterpart handling the situation and add that to its list of skills, should the question arise in the future.

We take things a step further and enable the user to create scenarios in which a chatbot should directly transfer the interaction to a human. For example, an angry customer deciphered through our sentiment analysis modules should be transferred to an executive or maybe for the chance of up-selling a product or a service the bot can be trained to pass on the conversation so that it is better handled by a human. All of this ensures your bot breaks down less and is able to answer most of the questions thrown its way. With the capability of listening and making sense of conversation it even eliminates the need of constantly updating FAQs.

Like to know more or just test the capabilities of your existing bot? Feel free to reach out to me here and go ahead and ask your questions in the comment section. Or you could email me at nikhil@aikonlabs.com.



Read More

LEAVE A REPLY

Please enter your comment!
Please enter your name here