How Generative AI Is Enhancing Customer Service and the Contact Center
Invest more, spending less. That's one of the opportunities that generative artificial intelligence brings to customer service and to contact centers in particular: to invest in more empathetic customer service using better AI alongside humans who are available and trained to deal with important queries.
The past few years have been marked by a significant investment in ensuring that conversational AI development becomes more accessible and that CX can be maintained with minimal expertise and less investment.
In a recent survey, Gartner found that despite the economic pressures facing many organizations, "only 7 percent of chief financial officers plan to decrease customer service spending over the next 12 months, with 21 percent planning to increase their spending and 72 percent planning to maintain their current investments.
With the dawn of generative AI, the industry can reduce even further the need for expertise in a number of areas (natural language understanding, voice user experience, custom integrations, and other technical roles). This change has cut costs significantly while flattening the learning curve for conversational UX development. We are finally entering a time when the business need will drive what is done with AI as opposed to the technology defining the limits.
Nonetheless, there's a lot of room to explore, and it goes without saying that generative AI will become instrumental in fundamentally changing how contact centers are operationally designed and managed. Let's take a look at how generative AI will enhance customer service applications and roles within the contact center to get a sense for what organizations can expect in the future.
Gen AI Is impacting Customer Service Applications
As a result of generative AI, a lot of new functionalities, including Q&A bots and self-service, are now ready to be bootstrapped to applications. Information can now be compiled and consumed through voice and chatbots in a fraction of the time.
Conversational AI bots have improved their context awareness, ability to reply in context, and ability to create rapport. As a simple example, it is perfectly reasonable to expect a conversational AI bot to be able to effectively upsell and cross-sell without the need for heavy fine-tuning on case-specific models.
There has been contention in the past toward the efficacy of AI to manage the entire end-to-end customer service life cycle with minimal human intervention, but generative AI has now proven that there's less justification for this hesitancy and that an AI-first approach can be superior to the current human-first call center default.
Emerging Roles in the Contact Center
Generative AI will also play a significant role in shaping the job of customer service agents and business analysts. It will make it easier for relevant information to be extracted from structured and unstructured data, and it will reduce the technical barriers to leveraging AI to promote efficiency within contact centers around the world, not to mention accelerating the development of conversational AI applications by an order of magnitude.
It will become easier to train workforces; they'll become more productive and achieve better results in a shorter period of time. Churn rates in contact centers have been reported on average as 30 percent to 40 percent and even sometimes as high as 60 percent, so making training more effective and consumable via generative AI will be paramount for the evolution of the contact center.
Knowledge bases can now provide answers to customer queries faster than before and help support and train agents. This data will also become more valuable as it feeds generative AI systems for continual learning. We will see AI and humans in a shared environment constantly generating content that simultaneously improves the performance of both.
Generative AI will also impact the roles of agents and supervisors within the contact center. We are expecting to see both agents and supervisors become orchestrators of AI and expect contact centers to move closer to a human-in-the-loop AI-first approach.
A productive agent will be able to orchestrate several AI tools and produce training data (and/or feedback) while performing regular tasks; this will iteratively allow customers to either self-serve or be served with the assistance of a conversational AI solution. Supervisors will become responsible for curating data and ensuring the appropriate monitoring and barriers to hallucination are being respected.
This Is Just the Beginning for Contact Centers
There was a certain feeling after the release of ChatGPT that data quality was not relevant anymore (i.e. you could just trust the internet and make things work), which is a misconception.
This dawn of generative AI does not mean we have achieved artificial general intelligence (AGI) or that we are anywhere close to it. Specialization of large language models (LLMs), for example, is paramount for cost efficiency and upholding task-specific accurate behavior. Efficiency—either computational or operational—will always be a concern for contact centers.
Generative AI is giving us the tools to generate integrations, generate conversations, and create functionality at a fraction of the price, with less resources involved and maximum fulfillment. However, to fully reap the benefits of generative AI, we have to define control, governance, and explainability.
Non-deterministic does not need to equal unpredictable when talking about generative AI. Managing generative AI responsibly and securely will allow organizations to grow their AI strategies with confidence. Vendors that can create and manage their own specialized large language models will be adding a powerful tool to their technology stack and enabling contact centers to do exactly that.
Claudio Rodrigues is chief product officer at Omilia.