Chatbots for Work

chat bubble with a headset signifying a chat bot

Executive assistants are a luxury that people only experience if and when they reach the highest levels of the org chart—hence, the “executive” tag. With the inception of the enterprise chatbot, however, this rarified exclusivity is beginning to wane. Work bots are well on their way to delivering a future where every employee enjoys the executive experience of today.

Of course, chatbots at work won’t be able to and should never replace actual EA’s. In practice, the reality is quite the opposite. For example, if Capacity is any indicator, work chatbots will simply boost executive assistants’ productivity and ability to be that much more effective. And that’s the whole point: chatbots at work support the business as a whole by empowering everyone in the org—EA and CEO alike—to do their best work. 

What’s a chatbot?

an illustration of capacity's knowledge base

Let’s break down what we mean by chatbot before we dive into the features that one should have. A chatbot by definition is a computer program designed to simulate conversation with human users, especially online. Since we’re at the forefront of the 4th Industrial Revolution, we’re going to be specifically talking about an AI-powered chatbot. 

An AI chatbot uses natural language processing, gets smarter with time, and can facilitate much more than question-and-answer exchanges. To be successful in an organization, an AI bot needs to integrate with people, documents, and apps. 

Reaching people where they are

For an AI chatbot to work at an enterprise level, it needs to connect with everyone where they are already working, instead of adding a new place for them to do work. The great thing about Capacity’s chatbot is that it can be added to any communication channel, in addition to any internal or external web page.

The simplicity of a chatbot makes it easy for an organization to implement the solution across the company. A chatbot also allows a user to continue operating in their preferred channels without learning a new system or creating a new window. In short, it can follow you where your workflow takes you.

The average employee switches between different applications 1100 times a day. With a chatbot, this number can dramatically decline. Rather than switching between different apps throughout the day to find specific information, a user can simply ask the chatbot in their preferred communication channel to locate specific details without toggling themselves. 

Connecting to disparate systems

Information silos are hard to break, especially those that are born out of disparate systems. Gathering data often requires time and expertise that not all organizations can willingly spare. It’s important to make sure any enterprise chatbot can connect to disparate data systems like legacy systems. 

When chatbots can connect to information silos, a user should be able to access them quickly, which means saved time—and ultimately saved money. The less time a user spends searching for information that is difficult to access, the more time they can spend focusing on meaningful tasks that drive revenue or thought-provoking ideas that lead to future efficiencies. 

Even if an organization outsources disparate systems, an enterprise chatbot should be able to connect to the system and securely transfer accurate information to the person requesting the details so there are zero concerns with data integrity. 

Built-in learning

It’s common knowledge at this point that machine learning and the AI chatbots it powers require a lot of data. The question that is raised in many organizations revolves around the type of data that will live in the enterprise chatbot. 

A chatbot shouldn’t need to rely on storing sensitive information to work. Find models that work great out of the box, and get better over time by learning what each user likes and doesn’t like—not by taking in unnecessary information.

For example, imagine asking a question once, and receiving an undesired answer. Instead of getting the same canned response to that question, a user can rate the answer they received with a thumbs up or thumbs down. AI bots like Capacity are created with machine learning so they can take cues and sharpen answers as time progresses, and responses are received and stored. 

Facilitating conversations

Chatbots are often known to spit out one standard answer to a question, but sometimes a question has more than one standard answer. Whether a broad question is asked, or the answer requires clarification, an enterprise chatbot should be able to get to the bottom of a question by responding with clarifying questions. 

For example, if a new employee asks you about adding someone to their insurance, you’d follow up with something like, “Do you need to add a spouse or a child?” to ensure an accurate answer. The same goes for a chatbot. When AI is built into a chatbot, it can facilitate conversations that ensure each user receives the most accurate responses that actually answer their questions. 

Leveraging natural language processing (NLP)

Rather than asking a user what they mean, or leaving a question unanswered, an enterprise chatbot should use natural language processing to understand and respond to the user’s questions, just like a coworker would. 

In every organization, there’s industry jargon that can seem like a foreign language to someone outside of that industry. But an AI bot with NLP baked into its system should be able to break down the context around a word or phrase, account for any acronyms or other industry terminology, and find different ways that word or phrase was used in the past, to provide a relevant answer. 

In addition to the example of industry terms and phrases, a chatbot should get what every user is trying to say and take the right matters to do so if it doesn’t. There are chatbots that can facilitate spell check, word importance, word order, term frequency, acronym translation, phrasal matching, word vectorization and more—so no user is ever left feeling misunderstood. 

A human-in-the-loop

New information is always coming up in an organization, especially at an enterprise level. When this new information lives in the minds of a team, or in new documents that have not been widely shared, it’s impossible for an enterprise chatbot to access it and use it to answer incoming inquiries. 

When a chatbot doesn’t know the answer to a particular question, it should be able to find the person who does. Because the enterprise chatbot is powered by AI, a backup plan often includes quickly finding the right person to contact when this situation arises. By looking at factors such as job title, years of employment, past projects, and past conversations, an AI bot should know the exact person to reach out to when they need a specific answer. 

Additionally, once the chatbot gets an accurate answer from the designated expert, it shouldn’t have to ask again. This ensures that different members of an organization can also ask this question and get the accurate answer—that is until an update occurs, and the process continues. 

A joy to use

With all the capabilities listed above, it would be a crime if an enterprise chatbot lacked personality. While jokes and witty banter are not the main concern, it definitely can be a form of stress relief for a team that’s knocking tasks out of the park. 

AI chatbots are supposed to simulate a human experience and feel like a living and breathing person is on the other end. Users should be able to add fun and quirky information about teammates and the organization to their chatbot’s knowledge base to bolster the culture and boost morale across departments. Why not encompass work and play into chatbots at work?

See an enterprise chatbot at work by watching the award-winning live demo we presented at In|Vest 2019. If you’d like a bespoke demo that speaks exactly to your organization, sign up for a 15-minute discovery call.

If you'd like a full overview of what Capacity can do for your productivity, just click the button below!

Request a demo

Check out these related articles