In our increasingly ‘smart’ age, the business world remains bogged down by administrative tasks. That’s according to a 2016 Accenture study, which found that managers ranging from the frontlines to executive level spend an average of 54% of their working time on administrative duties like scheduling meetings and writing reports.
When asked by the study’s authors if they’d be in favor of automating some or all of these tasks, a whopping 84% of business leaders said they would welcome the help of artificial intelligence. In response, a Harvard Business Review article on technological trends in enterprise declared, “Leave Administration to AI.” This can and should become the era of automating administration.
Since 2014, the New York City-based startup, x.ai, has sought to deliver on one piece of this ambition: offering autonomous AI assistants to help individuals and businesses schedule meetings more efficiently. The company’s two AI assistants, Amy and Andrew, offer a remarkably simple enterprise solution, acting as virtual intermediaries in email exchanges and conversations on platforms like Slack. By syncing with users’ calendars, Amy and Andrew facilitate the entire scheduling process, obviating the need for the usual back and forth in picking the right time and place.
Unlike many enterprise AI solutions, Amy and Andrew are accessible to both businesses and individuals, starting at $17 per month for a personal plan. And they’re remarkably simple to use, as I found out during a free 7-day trial. Once I connected them to my Google calendar, activation was as simple as CC’ing firstname.lastname@example.org in an email exchange. Here’s an example from the x.ai website showing Amy intermediating a request for a meeting:
Image source: https://x.ai/how-it-works/
The Basis of x.ai’s Assistants in Action
Amy and Andrew’s apparent simplicity belies the achievements in AI that have made them possible. The assistants work via natural language processing (NLP), the subfield of artificial intelligence dedicated to building machines that can understand human language.
Indeed, behind Amy or Andrew’s interpretation of even the simplest email lies a multi-part process. First, they must preprocess an email’s text, separating elements like salutations and signatures from the actual body of the message. Next, in a step called ‘entity extraction,’ they must parse the relevant text for keywords and phrases related to the message’s meaning. Finally, they have to place those key entities in context to know precisely what the message is asking – a process known in NLP as ‘intent classification.’ Adding to the complexity, x.ai’s assistants must be able to understand things like meeting requests among a multitude of participants, or situations where scheduling involves working across different time zones.
A demonstration of entity extraction in Amy. (Image source: Rakesh Chada/YouTube)
“The Future of Work” and Automating Administration in x.ai’s Vision
x’ai’s CEO and co-founder, Dennis Mortensen, has led the company through $44M worth of funding rounds with blue-chip investors such as Two Sigma Ventures and Firstmark. Meanwhile, x.ai has placed on CB Insights’ AI 100 and found mentions in platforms like The Financial Times and WIRED. Mortensen, a serial entrepreneur and data scientist originally from Denmark, claims that the idea for x.ai came from a long-term vision to “kill the inbox.”
“I came up with the idea for x.ai after I went through the enormous pain of manually scheduling 1019 meetings in 2012, of which ~670 had updates or were rescheduled at least once. Anybody who goes through that amount of pain will try to find a way to remove it.” -x.ai CEO Dennis Mortensen
While the death of the inbox seems far-off, x.ai’s Andrew and Amy give us reason to feel optimistic about the future of work – a prime example of ‘narrow’ AI freeing us from the tedium of everyday administration. And of course, personal AI assistants like these are poised to make a significant impact for a wide variety of organizations.
Some experts believe such assistants will become as ubiquitous as chatbots, likely going on to perform many more tasks beyond scheduling. But for now, one can only imagine the time and energy saved if Amy and Andrew—if we created a world where automating administration was a no-brainer—were used for the estimated 11 million business meetings that take place every day in the US.
Learn more at x.ai.
If you’re interested in startups harnessing artificial intelligence, check out the Accelerate AI event series at ODSC. Discover how the most innovative startups and Fortune 500 companies are using AI to disrupt industry at ODSC’s upcoming summits in London and San Francisco.
Stay up to date on all things data science here.