Businesses are looking to expand self-service options for customers and equip agents with AI support, including generative AI. This is great except they face the imperative of organizing their data effectively. This entails dismantling data silos and establishing a cohesive data infrastructure to ensure the consistent provision of accurate and contextually appropriate information across all touchpoints.
By consolidating disparate data sources into a unified framework, companies can enhance operational efficiency, optimize customer interactions and empower AI systems to deliver personalized and effective assistance, thereby elevating the overall quality of service delivery and fostering stronger customer relationships.
Beth Schultz, vice president of research and principal analyst, Metrigy, led a panel discussion at Future of CX Expo, featuring Whitney Meer, applied AI product manager, NICE; and John Chmaj, Sr. practice director, chief KM strategist, Verint.
Schultz raised the point that a third of companies said they have a common data platform – a platform that gathers data and then sets their AI on that. According to the data Schultz shared, only 13% say they don’t need work on knowledge management systems, and 30% said they already upgraded. More than 50% recognize they do need upgrades.
Then she showed a slide that asked, “does your AI-enabled apps pull content from you KM system?” Almost three-fourths of companies say yes. What about if GenAI pulls from a company’s knowledge base, which is managed and accurate, is human oversight needed on the content created? The majority say yes.
To hand it off to the panel, Schultz showed one more graph, “Is keeping a knowledge base up to date for GAI too high a barrier for entry?” Among organizations, 44% say yes, 55% say no.
“ChatGPT made GenAI more compelling,” said Meer. “The tech is not new though.”
“AI is the wild west right now,” said Chmaj. “Everyone is excited for it, but it is extremely unsettled. Yes, it is renewing people’s interest in knowledge management. No people do not know what to do with it.”
Schultz then asked the panel about closing a gap between GenAI and knowledge management.
“There are two kinds of gaps. The fist is that they don’t have enough content. The other gap is that they have too much content,” said Meer. “And it depends on the type of customer.
“A lot of people need to fail, not to sound cynical,” said Chmaj. “Be willing to take the hits of failure. It’s not really about failure, but how to create value. You have to define what you watn to use and by who. The market will pull this forward.”
Schultz then asked the panel to describe traditional data knowledge management versus GenAI data knowledge management?
“Traditional is people writing for people. People writing for an audience,” said Meer. “AI models create more personalized content.
“In a college/university setting, AI can write for you, and professors cannot tell,” said Chmaj. Maybe college changes from creating an argument by yourself, to having all these data points brought together and have content generated with AI. You need someone to oversee content so you can take advantage of these new capabilities.”
Meer added a final comment saying, “you will never have a group of agents you never watched over. Same with chatbot. It will be the same with any GenAI solution. GenAI solution can be seen as smart interns.”
The key insight from this session is to explore GenAI solutions for problem-solving and ensure that the customer experience aligns with desired outcomes, thereby unlocking more opportunities to address the specific problem the organization aims to solve.
Edited by
Greg Tavarez
