Companies are increasingly turning to conversational AI to automate customer engagements at scale. This technology allows digital systems to “speak human,” so users can independently and rapidly locate information and resolve issues in the way and manner in which they communicate in everyday life.
The ROI on any conversational AI deployment is directly tied to the size of the user base – the more automated engagements, the more services the company will be able to provide. However, in order to deliver a quality experience for a large pool of users, companies will need to anticipate the wide spectrum of human engagements and experiences contained therein.
When beginning an AI journey, companies should gather a team as diverse as the customer base they hope to serve. Those that rely solely on IT staff from similar backgrounds will find themselves at a competitive disadvantage when their AI experience fails to connect with a wide enough audience.
You Don’t Know What You Don’t Know
A few years ago, a Nigerian man caused a viral sensation when he uploaded a video of an automated soap dispenser whose optical sensor didn’t recognize users with dark skin. Surely the manufacturer did not purposefully design a product that would not work for certain users. We can also assume that the manufacturer was not aware of this problem before going to market – which might be the larger issue. In all likelihood, the problem stemmed from a lack of diversity among the engineers (or, at the very least, testers). In this case, the lack of diversity on the development team led to an embarrassing design flaw that was magnified in the social media age.
While diversity in a company has many inherent benefits, it is particularly important when developing a conversational AI to address the needs of a large and likely diverse pool of users, with a wide spectrum of communication styles. Diversity of developers doesn’t only refer to demographic backgrounds, but professional backgrounds as well.
Diversity in Skillsets
When the head of a major multinational investment bank hired our industry-leading Virtual Agent (VA) Amelia to automate many functions of its internal IT helpdesk, the executive in charge of the implementation noted that one of the keys to success was including a mix of skillsets. This included technologists and “humanistic” people who understand how human beings work and relate to one another.
IT skillsets will, of course, be pivotal to developing an AI-powered solution. However, in order to make a solution usable to a large diverse group of people, many AI pioneers have noted the importance of tapping non-IT skillsets as well. Otherwise, you might end up with a flawlessly working solution that only caters to other IT workers.
When it comes to building an in-house Cognitive Center of Excellence (CCoE) to develop your AI solution, IPsoft’s Director of Enterprise Solutions Allan Anderson recommends adding humanities-focused team members to the core group. This includes a well-rounded Conversational Experience Designer (CED), as opposed to relying purely on the work of linguists, because “when implementing and training Amelia, we’ve found the overall conversational journey is closer to a design process than a traditional linguistic exercise.”
To further this point, when describing the role of CEDs, IPsoft’s Director of Cognitive Experience Christopher Reardon says “our CEDs come from a diverse set of backgrounds including advertising and design agencies, consultancies, publishing and service design companies. These skillsets are critical as conversation is far more of an art than a science.”
At this year’s Digital Workforce Summit (DWS) on May 8 in New York City, attendees will have the opportunity to hear directly from global enterprise executives from Bank of America, BNP Paribas, Telefonica, Becton Dickinson and more. Visitors will hear a variety of insights from these executives, including how to choose the right team for the best AI deployment. Click on the link below to reserve your spot.