NLP Chatbots - Rakebots

Rakebots Vs Other Platforms: A Deep Dive

The stats predict that by the end of 2020 about 80% of businesses will adopt chatbot to increase their Business revenue.. So this is a new market and you see lot of chatbot building platforms emerging from various parts of the world. Some of these are generic platforms, using which you can build bots for any business use case and then there are niche based chatbot builders whose focus is on certain process of a business.

Rule Based Vs Natural language processing

Rule based chatbots are limited by functionality. They tend to work on defined set of rules. If a question is asked which is not defined, then the bot will not know what to answer. Natural language processing bots, as the name says are more intelligent. They answer based on the language and intent of the question. You can provide them lot of training data, and they understand and respond to queries if there are variations of same questions or if there are any grammatical errors.

They can also extract certain words from sentences which they are trained for e.g date, time, location, phone, email etc. and these can be used for various business purposes as required.

Check out the below video to understand better:

Manychat, Mobilemonkey, OctaneAI are the famous names we hear when we talk about chatbots. All these platforms specialise in marketing and lead generation and heavily depends on facebook messenger to get you leads and broadcast messages.

These are flow based bots and no much intelligence is involved and to overcome this, most of the time bots built using these platforms don’t allow users to enter text unless they specifically ask for it (e.g contact details). Hence the conversation is completely controlled by the bot.

These bots actually work well for businesses because they can broadcast messages and discounts to users who have already messaged you. But what if the user has queries they want to ask? bam! live chat to the rescue easy-peasy.

oh wait! now I have to hire Virtual assistants who can monitor the chat and take over when required. So now you pay for chatbots and assistants. hmmm…. So what happens when your VA is not available? yeah put up a contact form and respond later. Couldn’t you do this before itself instead of spending so much?

This does not mean NLP bots can answer all the questions. It depends on the training data you provide. Which means your bot becomes better over time based on the user queries. But hey! atleast it can respond not be completely dumb 🙂

Here is a chatbot video built using Rakebots for a client in action:

But I already have a chatbot on manychat and don't want to switch

If you have a chatbot on Manychat or any other platform, you can still make them intelligent. Once you build the training data model with Rakebots we provide you API’s to integrate to your current bot. using Manychat’s dynamic blocks or if your current platform provides such support, you can integrate Rakebots to provide better experience!

Can't I use DialogFlow or Wit.ai for the same?

Technically yes, you can. We already have integrations to facebook messenger, Shopify store, Woocommerce store and we will soon be deploying integrations to live chat softwares like intercom and others who provide such support. Additionaly we have an easy interface to build your training data. In short build once and deploy to various platforms. You may have to spend $$ to hire someone and build these integrations.

What about language support

I bet you have searched all over to find a platform using which you can build bot in your native language. Any luck? With Rakebots you can train your bot in any language!

Again watch this below video of Arabic bot in action:

oh wait! we are not done yet...

As previously mentioned you can build the training data and we provide API’s to integrate to your existing bots. But let’s talk beyond bots! Bot is just a widget interface, it’s all about data!

Let us consider a scenario:

You have a CRM and frequently receive such SMS or Emails

“Can you schedule a call today and contact me at xxxxx”

So now you manually set an appointment and add contact details in the CRM. But I have already shown in the previous video that we can extract such words like “today”, “tomorrow” and convert them into date and time, email, phone number etc.. and send it to your CRM (provided you CRM has the support) using our API’s. Isn’t it cool?

The power of Automation!

So using Rakebot’s powerful intent classification and entity extraction tool, you can think beyond Chatbots and the possibilities are endless!