Chat Bots for the Next Generation Enterprise
June 2, 2020
Eduard Sorescu

The Artificial Intelligence enables machines to simulate the human intelligence, to learn from experience, and to perform many types of tasks that are, in general, done by a human being.

Artificial Intelligence is used directly or indirectly in more and more industries, from healthcare (a traditional industry based on human to human interaction) to car construction (early advocates of using machines and robots within the production flow). AI is now a necessity, which should be adopted on fast track by the companies because it can enable them to remain competitive by automating and optimizing processes, reducing costs and the time that is required to get things done.

The rise of Artificial Intelligence is changing the way users are interacting with applications as well. Application developers should no longer rely solely on standard controls but should embrace a more rich, natural and free form of interacting with their users. Chat bots are just such tools.

What is a chat bot?

A chat bot is a software module able to carry out a coherent conversation with a human, providing relevant information or translating end-user requirements in actions. It is a software that can simulate a conversation with a human being through several communication channels – web applications, mobile applications, social media and many more, even the telephone.

In, when we are implementing chat bots, we take care of the process of conveying business logic within the chat bot context. We are adding artificial intelligence to it and we train our bots using ‘Question and Answer’ and`Language Understanding` systems in order to best fit our customers’ needs.

When should you use a chat bot?

A chat bot is a perfect tool that can complement a support team which is answering end-user requirements, or within an enterprise application, it can act as the intelligent assistant for a specific business department.

Whether you plan to off-load your business users or complement your support teams, you should really think about integrating a chatbot in your existing applications if your answer to the following questions is yes:

·      Do my support team/business users receive a lot of requests through ticketing systems or emails?

·      Can I group those requests by specific areas and actions executed by the team?

·      Can a significant number of those requests be automated?

·      Do I have APIs in my backend systems through which I can automate this?

·      Is it smarter to train bots than humans for recurrent activities?

·    Would I have higher ROI if my team engagement and creativity is focused on new features and business growth?

Chat bot example

There are multiple technologies to implement a chat bot. There are open-source stacks and enterprise class solutions. You can either opt-in for having on a premise’s installation or for a cloud based solution. When choosing a chat bot you should consider multiple aspects from functionalities, security requirements, budgets, to technology stack and deployment options.  

Below is a sample chat bot flow, showing how we use Azure Bot Service to implement reliable chatbots, in such way that they are capable to provide valuable information for users that need help going through any kind of process.

If you are a company that already uses Microsoft technologies, you should definitely consider Azure Bot Service as an option. It comes with a lot of goodies, such as:

·      Microsoft Azure QnA System is a way to train a sophisticated bot using FAQ pages, product manuals, SharePoint documents through the UI or REST API.

·      Microsoft Azure LUIS is a Machine Learning system that builds natural language into apps, bots and IoT devices.

·      Azure Bot also permits to hand the conversation off to a human when there are topics/questions the bot cannot find a reliable answer for and it needs human interaction.

Recently, we developed such a Chatbot for one of our customers, in order to facilitate the interaction with its vendors, to automate the time-consuming tasks, and to minimize the occurrence of possible errors.

After the welcoming message, the bot asks the user to choose from a list of suggested topics:

After the user selects the topic, the bot asks him to fill-in the textbox with some details about what he needs to know:

Then, the bot determines what exactly to show the user and it asks for a confirmation:

If the user confirms, the bot provides the instructions he needs, otherwise, the list of topics is displayed again.

How the future looks like?

Whether you plan to integrate chatbots in your IT architecture for customer service needs or you envision its use for various support functions, the chat bots are here to stay.

Applications will become more interactive, allowing for agile companies and employees to rely on a large army of intelligent assistants always available to help them in their work.

A business process started in an application chat will instantly be transferred to a mobile channel while the user is on the run and will be resumed in another application to carry on specific tasks.

The future of the business looks very promising. From the improved customer service (which is already happening in a lot of organizations) to enhanced RPA and cross-applications interoperability (between various organizational applications like ERPs, CRMs, Enterprise Apps) to a smoother and more efficient collaboration between different business departments.

Talk to the team