girl with hand in robot's
PHOTO: Andy Kelly

Artificial intelligence might still sound like a thing of the future, but the truth is AI-driven chatbots are changing the way we look at the workplace today. 

Accenture predicts that by 2035, artificial intelligence “will double economic growth rates in 12 developed countries and boost labor productivity by up to 40 percent.” All around us, we see technologies such as Siri, Alexa and Google Home incorporating natural language conversations between humans and AI into everyday household interactions. Why can’t that happen in the workplace?

Powered by natural language processing (NLP) technology, an enterprise chatbot can enable conversations between humans and computers in everyday business interactions. They bring deeper natural language understanding to enhance search and provide an entirely new way for employees to interact with corporate data. 

In the age of artificial intelligence, natural language processing and semantic search, chatbots can bring natural language understanding into the workplace to improve information discovery and collaboration.

Chatbots for the enterprise, while developing rapidly, are still in their early days. Companies are still navigating and developing the right use cases for the technology, or experimenting with and refining their first bots. If you haven’t had a concrete plan, the thought of implementing a chatbot can be daunting and uncertain. Let’s break things down and try to understand the use cases, benefits and challenges of bots, and then look at how to get started.

Related Article: The Next Generation of Chatbots: Connected, Intelligent, Capable

Chatbots as Coworkers

Today’s most common use cases for workplace chatbots include relatively simple, high-volume tasks. By having the bot rather than an employee handle the tasks, businesses can improve employee productivity and customer experience processes. Here are some examples of tasks that workplace digital assistants could handle:

  • Order fulfillment (status, cancellations, return instructions, tracking numbers).
  • Basic tasks such as reserving conference rooms.
  • Automated customer support/responses.

A workplace chatbot should be able to respond well to natural language queries and be capable of dealing with rare cases or exceptions, such as situations involving language differences. In addition, it should be integrated with mobile apps, business systems and enterprise security measures to fulfill specific user and business requirements.

Related Article: Accelerate Workplace Transformation: From Chatbots to Intelligent Agents

Where Chatbots Don't Work

Remember that chatbots typically cannot handle a large amount of intents (hundreds or thousands), and that they are most effective when used for simple requests. Therefore, while chatbots can be deployed for many business use cases, you always need to consider whether an enterprise chatbot is practical for what you intend to do. Here are some cases for which chatbots are not the right tool:

  • Situations where language is unconstrained or widely varying.
  • User queries that can cover a wide range of possibilities.
  • Highly technical disciplines where complex language is required.
  • Applications and queries without discrete inputs (unless the options can be narrowed down for your customer base), such as house-hunting sites or online dating platforms or decisions about what to have for dinner or what book to read next.

Related Article: 8 Changes Chatbots Will Bring to the Workplace

Building a Chatbot

In general, an enterprise chatbot architecture should support a seamless integration between your business data in different systems, search engines, NLP algorithms, chatbot services and, in some cases, voice recognition, to provide answers or perform tasks.

Among various chatbot technologies available today, Amazon Lambda (an Amazon technology to execute small, quick-running programs in response to events) is a solid platform with which to start a project. There are many others, including Chatfuel, ChatterBot and Semantic Machines. These chatbot services can be securely connected with your business systems via APIs.

With the wide range of chatbot technologies today, it can be overwhelming to fully evaluate your options to ensure that you select the right tools for your needs. Additionally, it’s important to understand the complexities of building a chatbot and to determine if you have the right business systems and resources in place to implement such a system. Building and deploying a chatbot is extremely complex, so a dedicated internal team that understands your businesses processes, along with skilled chatbot implementation experts, will be essential.

It’s also important to remember that chatbot technology is just beginning to emerge as an information discovery and collaboration tool for the enterprise. The future will likely bring exciting advances in the technology that result in better capabilities and functionalities. Possibilities include learned behaviors and more advanced conversations for high-speed customer interactions.

With AI transforming every aspect of life and business, it is worth keeping track of how chatbots might add value to your business operations.