The rise of the enterprise virtual assistant
As we stand today, it is really only the so-called C-suite executive tier in any given organization that gets to enjoy the benefits of a personal assistant.
Dedicated to fixing meetings, arranging trips and occasionally ordering lunch, a personal assistant – often gloriously abbreviated to PA – is a power status symbol that is almost redolent of the commercial excesses of the eighties.
But this is the post-millennial digital age and, like everything, personal assistance is becoming digitized and democratized.
No longer the sole preserve of the CEO, we can all now receive help from virtual assistants built from software that is specifically engineered with an ability to learn our preferences in the context of any specific task or scenario.
Virtual assistants are often referred to as ‘bots’, short for software robots and sometimes as virtual ‘agents’.
In software development coding and computer science terms, an agent is a smaller program that serves another program or higher-level system in a ‘relationship of agency’, so the term makes sense.
But there’s a problem. Up until now we have only really embraced virtual assistants suited for consumer-level home use.
Many of us have found ourselves shouting at Alexa, Siri or the Google voice interface in our Android smartphones, but we have failed to carry those actions forward into the office.
In the workplace, enterprise level software usually involves compound interactions combining complex keyboard inputs with additional use of a mouse.
Combine the complexity factor with the challenge of human speech recognition and things start to get even more difficult. Enterprise-grade virtual workers have failed to surface, but things are changing – and it’s due to developments happening in software automation intelligence, Artificial Intelligence (AI) and machine learning.
As virtual assistants start to become truly effective, we will need to bring about a cultural shift on a number of levels.
How many of us are happy to start barking orders at our screens in a packed office environment? How many of us are happy to interact with a ‘Hi, can I help you!?’ conversation window in an application or on a website that isn’t actually a person at all?
We need to understand that completely virtual chatbot windows are learning all the time. When the software brain they are built with fails to be able to resolve our query, it knows it needs to perform a ‘handoff’ to a human operator.
But – and it’s a big but – equally, the software is able to learn from this failed interaction and understand what needs to be done in future once it sees the result of the human agent’s actions.
Deeper still, virtual assistant platforms are developing a level of symbiotic integration in the spirit of open source sharing.
We can reasonably suggest that all apps will feature a voice-driven fully interactive help agent within the next five years.
Natural Language Understanding (NLU)-empowered speech recognition that is capable of parsing human speech and interpreting the idiosyncrasies and nuances or human conversations in the correct context will be widespread. Moreover, it will not be ‘bolted on’ to an app, it will be part of it.
Alongside our cultural shift we will also see companies start to develop a formalized ‘voice strategy’.
Firms will need to set limits around what virtual agents can do as they start to come into contact with employees, partners, and customers so an enterprise voice strategy will need to apply the correct governance in any given situation.
Our next bots will do more than just chat, they will lead us all the way through to a real-time automated resolution.
Enterprise cloud services company ServiceNow’s latest Virtual Agent is a good example of software that promises to be able to resolve conversational requests automatically through more natural interactions. It is built to provide personalized responses in context.
For example, when an employee asks for a new phone, the chat knows what cell phone plan and carrier the employee already has based upon the backbone or corporate intelligence it is programmed with.
“Our Virtual Agent enables a powerful conversation model built natively in the Now Platform,” said CJ Desai, chief product officer, ServiceNow. “This enables our customers to develop a wide range of intelligent service conversations, from a quick question to an entire business action through the messaging platform of their choice.
Rise of the robots?
Desai insists that this development will free up human operatives (that’s you and me) to do higher value work. The robots aren’t replacing us, they’re just doing to repetitive stuff that us creative humans get so bored doing. ServiceNow estimates that 15 percent to 20 percent of routine enterprise interactions can be handled by its Virtual Agent.
Examples of the kinds of tasks that can be carried out by ServiceNow’s Virtual Agent include IT help desk requests such as resetting a password.
YOU MIGHT LIKE
Alexa wants to join your workforce
It can also perform Human Resources (HR) related functions to make it easier for employees to get answers or submit requests such as a leave of absence, research pay discrepancies, check status and update their profile.
Enterprises will now start to use voice agent platforms to design their own conversations that interact with humans in the specific context of the goods and services that they specialize in.
The fact that voice bots and chat interfaces of all kinds have taken longer to impact the enterprise space is a bad thing perhaps, but it’s also a good thing. This latency has allowed us to get used to these tools at home first and that’s great insurance for the longer-term development and survival of any technology.
You may not be the CEO yet, but now you can start to issue orders like one for a number of the tasks you will need at home and at work. Just don’t forget to pick up your own dry cleaning, for now at least.
Editorial disclosure: Adrian Bridgwater has worked for the ServiceNow EMEA content team.
6 June 2023