Building Trust with AI

MRS, Adam, and Prioritizing Customer Experience

While AI or Artificial Intelligence has roots dating back to the 1950s, significant advancements took place in the last two decades which propelled the type of capabilities that we see today. Machine learning became heavily funded, and such backing allowed for the great advancements we see today from Watson to Siri to Alexa. 

With applications of AI embedded in everyday devices like phones, people have started to build rapport with machine learning. However, along with your friendly neighborhood AI, comes its more nefarious counterparts like deep fakes and intrusive data gathering technologies. Building AI isn’t enough anymore—people want AI that they can trust. 

As AI grew and became more agile and seamless, MRS knew there was a place in the collections industry for this kind of technology. Innovation has always been integral to the way MRS operates, so we implemented Adam, our own virtual collection agent, in early 2018. 

Collecting debt creates an environment where it is important to foster trust between agent, client, and customer. Because MRS is experienced in building these kinds of relationships, introducing Adam into the collections process posed a challenge, but one that we knew we could handle. 

In order to build the best AI, MRS looked to its agents. Co-CEO Jeff Freedman said, “We began with trying to understand how our human agents build trust with customers. Humans do this through listening carefully, demonstrating empathy, and showing a willingness to work together with a customer. AI obviously cannot do all of those things, so the first thing we had to recognize was that Adam would not work for many customers, no matter how sophisticated he became.  But there are a sizable number of customers who want to work with the AI because they believe it will allow them to do what they want in a quick and efficient manner and avoid the discomfort of having to deal with a live agent. To build trust with these customers, we had to show them that Adam can allow them to accomplish what they want in a quick and secure manner. We also needed to identify what would cause customers to be wary or untrusting of the technology.”

red circuit board

Oftentimes, traditional Interactive Voice Response, or IVR, features a long winded menu option and that customers recognize as laborious and therefore opt for a live agent. Instead of building Adam with the traditional menu option, MRS engineered its AI to be ready to listen to the natural speaking voice of any customer. Freedman said, “With Adam, we wanted customers to realize right away that they could speak naturally and speed up the process and avoid the lengthy procedure of responding to menu prompts that may or may not get them what they want.”

Other factors that account for AI mistrust were considered and implemented in Adam’s design. Any pause before an answer, background noise, and colloquialisms or typical human lingo diminish a customer’s trust, so Adam was taught to handle and learn from these experiences. Freedman also highlighted that knowing Adam’s limitations is important too, and when those circumstances arise, Adam can transfer the customer to a live agent.

Adam offers a range of solutions and payment plans, while prioritizing compliance and the needs of both the customer and client. One of the main concerns with AI stems from the technology’s placement among everyday household items. Conversely, Adam differs from Alexa and Siri, because it exists solely as a virtual collections agent, to help people pay off their debts. AI with a singular purpose like Adam executes a task to completion, and then rests, eliminating many of the fears that arise from the more ubiquitous AI household items. 

Despite all the challenges that building Adam posed and the mistrust that surrounds more common AI, Adam excels. Freedman said, “Customers have shown us through their behaviors that they trust Adam to handle the most sensitive of interactions—the handling of payment processing and taking their payment information. In 2020 alone, Adam took almost 30,000 customer payments totalling over 6.5 million dollars. All of those 30,000 payments were done completely via self-service with no agent interaction and many of these calls took place outside of normal business hours, late in the evening or on weekends.”

AI surrounds us everyday, sometimes helping, sometimes hindering, but MRS wanted to build technology that assists customers on each call. Adam inspires trust through careful communication and placing customer needs at the forefront of every call. AI is the future, and MRS is ensuring that our AI is the kind that you know you can trust.