The growing popularity of voice assistants in the home is quickly extending itself to the enterprise environment, where a growing emphasis on lean operations is helping establish the technology. The rise of voice assistants in the enterprise context is also being fueled by the advances in natural language processing and machine learning. This is demonstrated by the fact that word recognition error rates have gone from 50% in 1995 to less than 5% in 2019.

But for many, voice assistants are still a new and relatively untested technology. Although tremendous advances have been made in digital assistant technology, it’s important to note that they have only been part of the mainstream for less than a decade and will likely see several iterations before being accepted as an indispensable business tool.

A brief history of voice assistants

Contrary to popular belief, voice assistants predate smartphones by several decades. In fact, the first true voice assistant, although very rudimentary, was launched by IBM way back in 1961. Here’s a short timeline of the voice assistant revolution through the past century.

Voice assistants in the enterprise

Currently, not many enterprises use voice assistants for business beyond deploying them in customer support and lead generation roles. Furthermore, voice assistants may not translate well from the home to the office, given current operational structures. For example, making an online purchase via a voice interface may be easy for a consumer, but in the business context the process usually involves a purchase form to be filled out and an invoice to be signed.

However, as the digitization of business systems continues to grow, we will likely see these hurdles demolished.

A great example of how voice assistants are helping boost business performance is the integration of Alexa with Salesforce - this helps sales teams quickly dictate simple administrative tasks to the voice interface (like logging client interactions or setting up meeting reminders). Similarly, the A.ware platform is an enterprise voice assistant development framework that helps businesses build custom business bots to handle specific business functions and integrates into your business systems via API.

• Voice assistants in sales

  As mentioned earlier, voice assistants can make sales teams more productive, by streamlining administrative tasks and enabling more focus on lead conversion. But they can also add value to the sales funnel by nurturing and acquiring warm leads on their own. For example, with a suitable Artificial Intelligence engine, a voice assistant could be assigned to send out specific emails to prospects within a particular stage of the buyer journey or could be deployed as sales AI assistants, helping enterprise purchasers find the right service package for their needs.

• Voice assistants in administration

  Tasks like ordering fresh office supplies from a pre-approved list of vendors, organizing meetings, sending out internal memos, recording minutes and accessing data or reports, are all suited to voice-enabled personal assistants. Not only do they offer a faster resolution for all these tasks, while reducing the scope of error, but they also help your transition to a leaner operational model by eliminating the need for low-tier administrative staff.

• Voice assistant in hospitality

  The idea of digital concierge is slowly catching on in hospitality circles. Imagine a hotel where every room is equipped with a digital assistant to instantly convey instructions from guests to the hotel staff. Similarly, the different departments of a hospitality concern could be seamlessly connected via an intelligent voice assistant, helping streamline operations and deliver high service standards.

• Voice assistants in healthcare

  In the healthcare context, patient registration and diagnostic assistance is one area where AI voice assistants can immediately be deployed. Not only will this eliminate long queues in hospitals as staff struggle to keep up with patient intake but can also help doctors make a more accurate diagnosis. In fact, IBM Watson for healthcare is already leveraging its AI algorithms to help doctors determine the most effective treatment routes for cancer patients.

How do enterprise voice assistants work?

As the bleeding edge in man-machine communication, intelligent voice assistants can do everything from opening the garage door to booking a vacation flight for your entire team. Broadly speaking, most voice assistants function on the basis of a 3-step process. Step one is where speech-to-text converts your voice commands into a text input that the AI can process more easily. This is accomplished by analyzing the phonemes or component sounds of your speech, the context, and the syntax of your sentence to quickly match the command to the voice assistant’s databases of text inputs.

Once your spoken command has been matched to a text input, the process of determining intent can be started. For instance, when you ask your voice assistant ‘What’s happening in New York?’, Are you talking about the weather? Are you asking for the latest news from the city? Or are you enquiring about popular events in New York? A search engine will typically display all these results in order of relevance, but a voice assistant has to work harder to determine your exact meaning. Most voice assistants leverage machine learning algorithms, natural language processing engines and data from previous interactions to determine the correct intent.

The last part of this cycle is translating intent to action. Digital voice assistants today, can be integrated into virtually every aspect of enterprise operations and are especially useful when streamlining administrative tasks. Using the previous example, if you ask for the weather in New York City, and you have recently booked tickets to a conference there, the voice assistant might follow up on your query by offering to book your flight tickets or display a list of hotels near the conference venue.

Are there any cons to enterprise voice assistants?

Before you dive right into adopting voice assistants across your enterprise, it’s important to note that there are a few drawbacks to this technology.

• Privacy - The idea that a smart speaker is always in earshot of your conversations can be disquieting for many people. This is especially true to the enterprise context where sensitive data may be under discussion at any given point in time. To negate this issue, custom devices and privacy-oriented voice assistants will likely be launched in the near future.
• Compliance & Security - Although your interactions with every device may be encrypted, it’s possible that smart speakers may act as potential entry points to a business system for bad actors and hackers. But as the technology matures and becomes more popular within businesses, we can fully expect enterprise data governance models to be implemented within voice assistant architecture, as will access management and identification systems.

The future of voice assistants

This article lists just a few examples of how voice assistants are already changing the enterprise paradigm. And this change is not going unnoticed. According to an Adobe survey, 91% of business decision-makers are investing in voice technologies and over 88% plan to support several voice assistants in the near future.

This is hardly surprising given that machine learning algorithms are helping voice assistants become smarter and more intuitive with every interaction. It does not take a huge leap of logic to surmise that digital assistants will replace the administrative workers over the next decade, just as the smartphone replaced the Rolodex and the personal planner. And with the rapid growth of AI technology, it’s only fair to say that we can expect voice interfaces to disrupt the GUI as the next big enterprise communication interface.