With the bulk of 2017 behind us, artificial intelligence (AI) has clearly staked its claim as the buzzword of the year, although the conversation around AI has evolved over the months. It started with Elon Musk and Stephen Hawking warning against an AI arms race (think Terminator) in the Asilomar AI Principles. From there, companies ranging from the smallest startups to the Big 5 – Apple, Alphabet (Google), Microsoft, Amazon, Facebook – tested the AI waters, discovering ways to incorporate AI into their everyday processes through machine learning, natural language processing, chatbots, etc. Now, AI functionalities are becoming an industry standard, and the conversation for the future revolves around how many jobs AI will replace. Among the numerous reports and surveys, a Gartner report predicts that by the year 2020, 85 percent of customer relationships will be through AI-powered services.
Before the customer service AI robots take over, AI stands to augment current customer service practices, especially when it comes to service desks and field service technicians. Despite the rapid advancement in today’s technology, service technicians still need to continually answer three age-old questions:
- What are the common problems the customer faces with the broken equipment?
- What tools does the technician need to effectively fix the equipment?
- What level of expertise is required for the technician to complete the task?
As age-old as these questions are, the technicians’ methods of answering them are just as outdated. Paper service tickets, unproductive phone calls and the one-technician-fits-all mentality create bottlenecks of information and workflow. This problem is exacerbated by the acceleration in customer expectations, brought on by the near-instant gratification that the Internet of Things (IoT) brings.
So, how do service desks and technicians keep up with modern times? By befriending robots, or at least the AI infrastructure behind them. A key opportunity to implement AI functionalities resides in processes that are routine, repeated often and require little-to-no creativity. Service ticket fulfillment and technician dispatch are two such tasks that stand to greatly benefit from AI.
Taking a deeper dive, here are two specific ways AI bridges the gap between the service technician and the customer:
When your device or appliance malfunctioned or broke in the pre-AI era, you had a couple of options to get the service request process started. The first would be to get on the phone with a service technician and attempt to describe the issue and provide adequate context on relevant details. Another would be to fill out a service ticket, which may leave you guessing on where to find certain product specifications and/or how to properly categorize your issue. Both of these “solutions” carry inherent risk, as any miscommunication will most likely force the service technician to make multiple visits.
AI chatbots, powered by machine learning, alleviate the pressure of these issue detection processes. Chatbots can understand customer intent in a chat session, determining if the customer needs help with an issue or wants more information about a piece of equipment. Logistic regression capabilities enable the chatbot to walk the customer through the equipment problem. This ensures that all the necessary details are captured in the “service ticket” that the chatbot fills out.
The dispatch process for service technicians has long been a contentious pain point for customers. Customers often have to wait for the technician to arrive within a window of time, or often a few hours at the very least – and that’s assuming that the technician arrives on time. Furthermore, the nearest technician is occasionally too inexperienced to complete the task in one visit, which adds to the total project duration. Or, the nearest technician can sometimes be too experienced for the task at hand, driving up the cost of the project.
Using information retrieval processes, the AI system provides dispatchers a list of technicians who are readily available and qualified to solve a customer issue. Similar approaches are used in search engines to provide relevant information in real time. Heuristic search algorithms (i.e. the AI algorithms used to beat humans in chess) also ease the search for possible solutions.
The AI not only lists out the technicians, but also prioritizes them based on a number of criteria. Knowledge and availability are the two main criteria by which the AI can sort technicians. However, the AI can go a few steps further from that. The AI can factor in subtler details and restrictions as well when searching for available technicians, such as such as work time hours, legally mandatory lunch breaks, minimum travel time and distance required, etc. These criteria narrow the scope of feasible assignment scenarios, allowing service technicians teams to optimize their dispatch processes. Furthermore, these AI capabilities go beyond the dispatch phase, as they can notify technicians about which tools are necessary for the job before the technicians head into the field. Overall, AI drastically reduces the time and effort needed for a human dispatcher to factor in all these conditions, which reduces project duration and increases overall customer (and technician) satisfaction.
Eventually, AI programs will evolve and automate many of the customer service processes that are handled manually today. However, until we reach the aforementioned 85 percent of “robot takeover” by 2020, service technicians can make the most of currently available AI solutions to continue delivering quality and timely service to their customers.