Virtual assistant technology

Three essential elements are required in order to create a virtual assistant which can competently assist your users and provide measurable results for your organization.

Our artificial intelligence has some of the most robust natural language processing (NLP) and natural language understanding (NLU) in the world. We also have the machine learning that helps us make continuous improvements to our software.

Certified human operators, who we call AI Trainers, efficiently train, customise and expand the virtual assistant, constantly making it smarter under controlled conditions. We can certify your employees so that they are able to update and maintain your virtual assistant into the future.

The final element is your own domain knowledge, acquired through years of your unique and specific industry experience. This provides the basis for the output of your virtual assistant, and you need to make sure it becomes an accurate representative for your organization and brand.

We will help you achieve this by using our extensive domain knowledge of implementing virtual assistants in your industry.

 

 
 Essential elements that are required in order to create an AI virtual assistant
 
 

How does it work?

A virtual assistant built on artificial intelligence isn't something that just magically understands what your users are talking about; it's all about cutting-edge natural language technology and state of the art machine learning models working together in harmony.

 
 
 AI chatbot technology: User input

1. User input

The user's message is usually received through your website, but the James platform is compatible with the most popular communication platforms, such as Messenger, Slack and Skype.

The message itself can be a simple question, such as asking for the opening hours of your business, or a more complex question that requires secure APIs to process the request, such as making an insurance claim or transferring funds between accounts. 

Whatever the user needs and no matter how the message is received, your new virtual assistant has the brainpower and flexibility to handle it. 

 

 
 AI chatbot technology: pre-processing

2. Text-processing

When the message is received, the virtual assistant starts pre-processing the content. This consists of several ways to "prepare" the message it receives for interpretation and ensures that the virtual assistant will understand it, whether that entails interpreting slang or dialects, or even correcting spelling mistakes.

To achieve this unrivalled language understanding, numerous advanced processes are required. They each play a crucial role in ensuring that your virtual assistant will understand what your users are asking for. Examples of these processes are language detection, spelling correction (tailor made for each domain) and splitting combined words.

When pre-processing is complete, the message is in an optimal state to be correctly predicted to the right user intent.

 

 
 AI chatbot technology: intent prediction

3. Intent prediction

The virtual assistant is now ready to calculate the user's intention, something it does by using our unique multi-level intent hierarchy, which has been developed to enhance both scalability and accuracy. The hierarchy can handle thousands of user intents that each require as little as 10 training messages to be correctly predicted.

We achieve this through an ensemble of prediction models in our intent mapping process, including convolutional neural networks (CNN) and recurring neural networks (RNN). The virtual assistant is able to understand the context of a conversation without forcing the user to keep repeating the same keyword.

Together, these methods give our intent predicitions a level of unparallelled accuracy.

 

 
 AI chatbot technology: dynamic response

4. Dynamic response

When the intent has been correctly predicted, the virtual assistant needs to provide a satisfactory response.

The responses are highly dynamic. They can be pre-defined text replies, which answers simple questions. They can also be complex answers and actions using entity extraction and external APIs that can perform advanced tasks on the user's behalf.

How the virtual assistant responds is up to you. What is certain, though, is that when you add support for advanced responses, such as making the virtual assistant able to interact with authenticated users, you can provide the personalised virtual assistance of the future - today.

 
 AI chatbot technology: anonymization

5. Anonymization and privacy

The final step of the message's journey is the removal of personal data. Privacy, and the protection of a user's personal data, has never been more topical or more important than it is right now.

Boost.ai takes privacy seriously. When using a virtual assistant on the James platform, you can rest assured that your personal data will be processed in accordance with your own instructions. We implement all the adequate technical and organizational measures to secure your data, including features such as full anonymization.  

We act in accordance with the GDPR principles and we are constantly developing and implementing better ways to secure your privacy, and that of your users.

 
 AI chatbot technology: analysis and optimization

6. Analysis and optimization

When a dialogue with the user has been concluded it is time to analyses and optimize the data.

Some of your content may need to be expanded or user intents covering subjects you did not anticipate may need to be created. Some replies could be sub-optimal, and some could be even more valuable if you connect APIs to provide your users with  a richer and more personalized response. 

These are just some of the issues you might consider. Your users will notice the difference when your virtual assistant steadily improves. We, and our trusted partners, will guide you through this vital process and ensure that your virtual assistant continues to absorb knowledge and evolve.

 
 AI chatbot technology

Not all questions should be answered by a virtual assistant

A virtual assistant is able to help with the majority of users' issues, but there will always be certain cases where human assistance remains the best course of action. These could be very specific problems that need human intervention, for example if a user becomes agitated or distressed. In these cases, the James platform's sentiment analysis is able to detect the mood of the user and respond accordingly. 

This means that when the virtual assistant realizes that it can't adequately help a user, it will gently guide them towards a human operator who can be standing by, contactable through the same chat window, ready to provide the necessary human touch. 

To help smooth this transition, we have developed a human chat panel which you can implement at the same time as the virtual assistant. The James platform is also compatible with the majority of the most popular chat software.

Features overview and comparison

Compare the James platform with other solutions for virtual assistants.