Skip to content

Improve speed, cost and time of building Virtual Assistants (Chatbots and Voice Bots)

 Building a Virtual assistant for your enterprise can be tedious, expensive and resource intensive. As any enterprise that has gone down this road has realized by now that it wasn't as easy or as fast as was promised and many times most enterprises walk away from (CAIP- Conversational AI Platforms) wondering where is the AI. So let's investigate how bots are built today and how you can do it in a smarter, more efficient way.

Historically there have been two ways that companies have built these Virtual Assistants

1. Flow Builder: This was very popular at the start of bot making and remains a great way to build bots for small use cases but difficult and not user friendly if you want to cover multiple use cases and queries

Source: Master of Code Playbook Conversational Design Series #2

2. Intents structures: Intent-based bot building became all the rage about 5 years ago. This is now extremely popular in most conversational AI platforms: This is a great way to get started with AI but expensive and requires significant training to make it usable. Even with these pitfalls it remains the most common option offered by CAIP providers. 

 

This is the reason that as per Gartner reports chatbots can costs over $250k to setup and then their maintenance costs for AI based bots being close to 50% of the setup costs. This is one of the highest recurring costs for maintenance. 

That is why a new breed of conversational platforms are emerging that look at solving this problem. This new breed combines three major technologies to provide what could be considered the swiss army knife that could transform the industry and change how we build maintain and analyze bots.

The approach that these platforms are taking is first understanding the most fundamental truth of conversational bots out there. This is that 20% of Intents make up 80% of the traffic. This means that 80% of resource time and effort of your organization is being spent into building and maintaining Intents or knowledge bases that make up less than 20% of the requests. So what is the smarter way?

1. Personalize and Automate: For the Top 20% use cases its imperative to create the best possible CX. This could mean smart workflow journeys, integrations or just extremely well written responses to clients. Many times organizations try to cover all use cases and miss truly hitting the top requests that will generate the highest ROI and increase in NPS.

 

2. Smart Automation: We all agree that the current methods and process to automate are expensive. This has led for two major changes that are slowly being see in the industry. The first is where CAIP platforms are no more just conversational AI platforms but quickly moving to becoming conversational AI platforms. The reason is quite simple. No one wakes up in the morning and says hey I want to chat with a chatbot today. They are looking for a solution to their problem and by embedding automation into the platform the idea of companies like ours is to make it faster, easier and cheaper to automate the most frequent requests to enable true CX and ROI. We provide a tool called Blitzico as a part of our offering to enable exactly this piece, This is not going one step further though with the introduction of NLQ. With NLQ capabilities you  can connect to systems like salesforce once and re-leverage that for multiple use cases without writing multiple connectors.  and automate them faster with lesser effort. 

3.Autnomous Document and FAQ extractor: This is a one of the new features that has also come up in the past 12 months where conversational AI vendor tools are offering a way to not just help with the top 20% but the rest 80% as well. Using this tool you can utilize AI to create knowledge for the AI directly from your Website (Autonomous Learning), Wiki's, Documents or any other knowledge repository you might have without any need for intent creation. The big advantage here is that this removes the need to build any manual intents that can take a significant time and is completely Autonomous. This is a great way of ensuring that you do not need resources to continuously update your Virtual assistant or chatbot knowledge as these are automatically understood from your data sources and automatically updated every time you you update your website or wiki. This is a great way to cover the long tail with almost no effort, high compliance and coverage without any of the effort that was previously required. 

We at  Enterprise Bot offer all these capabilities in one single platform and expect this to become the new standard very quickly. It might not be the best idea to switch to this platforms for simple quick 1 or 2 use cases but for production ready Enterprise Virtual assistants we believe this will become the new standard that will allow the industry to achiever higher ROI and NPS with significantly lower investment and finally start showing Enterprises where is the AI in CAIP.