AI Driving Digital Customer Engagement: Part II

AI Driving Digital Customer Engagement: Part II

By Setrag Khoshafian, Ph.D.  |  June 15, 2017
Dr. Khoshafian is Pegasystems' Chief Evangelist, IoT strategist, and strategic BPM (digital processes) technology leader

In Part I of “AI Driving Digital Customer Engagement” we focused on the “Insight” part of the insight to action dynamics. In Part II we focus on the “Action” part. We’ll cover digitization of value streams and the digital transformation through AI-Assisted Work.

Digitized Value Streams

The most powerful impact of AI can be realized in the context from the end-to-end digitization of value streams (aka “value chains”) involving multiple participants, business units or partners. A value stream is associated with a business measure (a Key Performance Indicator (KPI)) and typically involves multiple participants across the digital enterprise. Borrowing a compelling perspective from Theory of Constraints: “a chain is no stronger than its weakest link.”

Let’s face it: Most organizations are siloed. Organizations are organized vertically. However, responsiveness to business value generation typically involves multiple business units or participants. Value streams go horizontally across the digital enterprise. It is interesting that this aspect of digital transformation and the use of AI to optimize responsiveness for cases involving multiple business units is often ignored.

The substantive business return and business benefits of AI will come from services that connect the customer to the rest of the extended digital enterprise. What does the digitization of value streams mean for a digital enterprise? How does it impact digital enterprises that are attempting to evolve, improve, and transform the customer experience through AI? Well, it means customer interactions leverage AI for targeted, contextual and optimal interactions towards timely as well as effective and efficient resolution of their cases.

End-to-end value stream
The digital transformation of the customer experience goes beyond the call center or self-serving customer channel interactions (e.g. Web or Mobile or IVR or Bots or Customer Service representatives (CSRs) or increasingly Intelligent Virtual Assistants). Optimizing the front-end channel customer interaction is of course critical and important. AI plays a huge role in optimizing the customer touchpoint. However, the customer promotion scores  will depend on the aggregation of tasks that involve multiple business units to resolve the customer request. The work or task for the customer, needs to be routed to the best resource – in the specific context or situation of the customer. Essentially, managing the end-to-end value stream work needs AI. For instance fixing a broken appliance might involve an intelligent chat-Bot, a CSR, the Service department, the Field Service, the Supplier, and the Warranty department. As noted above, even in this modern era of digitization, most organizations are siloed. It is not enough to just focus on the responsiveness of the Virtual Assistant, the CSR, and the elegance of the Web or Mobile app for the customer. These are of course important. But they are part of the end-to-end value stream that needs to be strong and responsive through all the stages of the value chain. Each business unit needs to leverage AI to optimize the resolution of the customer’s end-to-end case.

Here are some of the AI capabilities that can be leveraged in end-to-end digitized value streams

  • Virtual assistants and increasingly intelligent chat-Bots are becoming one of the most important categories of AI optimizations for the customer experience. There are a number of key AI enablers for this domain including Natural Language Processing. The interaction of an AI-enabled virtual assistant could be text-based or voice-based or both. The often sighted “Turing Test” is particularly relevant here – as virtual assistants become more intelligent, it becomes more difficult to distinguish them from human intelligence.
  • The Best Action in a context: Customer situations, customer context, and customer sentiments are in constant flux. The aggregation and fusion of customer interactions, customer behavior with their connected IoT Devices, and sentiments through social channels need to be mined and then manifested in prioritized list of actions: write-off a dispute, make an offer of a new product or service, offer free upgrade, or offer a discount on a service.
  • Robotic Process Automation: AI techniques can also be used to analyze the performance as well as effectiveness of customer service agents, the applications they are using for customer interactions, and the processes in responding to customer requests. Unlike more daunting process re-engineering efforts that often require extensive changes to underlying IT technologies, RPA leverages the existing enterprise application landscape and uses AI with software robotics to eliminate human involvement in especially repetitive tasks.
  • Intelligent Skill Based Assignment of work: Often a lot of waste and customer frustration emanates from inefficient (aka “dumb”) processes that route work to the wrong resource. Handling a customer case will involve participants who could be humans or intelligent virtual assistants. The AI techniques can also be used to route the work to the best or most skilled resource for the specific customer request.

AI-Assisted Work

Digitization and especially work automation trends are changing or even disrupting entire industries. According to the Economic World Forum nearly 50% of current jobs could be automated by 2055. AI and automation are the main culprits. Even now, automation is having a huge impact on the workforce.

The spectrum of workers begins with clerical or manufacturing workers: repetitive and predictable work that can easily be automated. Increasingly automation with software and robots are taking over this repetitive category of work. This of course includes physical Robots as well as Robotic Process Automation. AI is increasingly playing an important role in optimizing work processing and understanding where are the potential bottlenecks for improvement.

Cognitive worker
At the other end of the spectrum of work and worker types is the knowledge (aka cognitive) worker. Knowledge workers think for a living. They are often the retiring expert workforce that have a lot of the know-how in their “head.” They innovate and often come up with ideas for new products as well as the policies and procedures in the organization. At best, their knowledge needs to be captured and digitized to drive work automation. At the very least, these cognitive workers need to be made readily available and participating in social interactions in the context of digitized processes.

Between these two, you have the most important category that represents the majority of workers: the AI Assisted Worker. The medium of the assistance could be an elegant UX, Voice, Text, or Visual – including Video. It could be in the form of suggestions, recommendations, or observations. It is like having an always on Virtual Assistant or an AI enable UX that tries to help you with your assigned tasks in a particular context. The AI assistant continuously learns and improves itself with a feedback loop from the data that captures the results or the performance of the end-to-end value streams.

Therefore in essence the “Action” part of insight-to-action is the shift in wit AI-Assisted automation from the left to the right. More and more of our routine and not-so-routine work will be automated. The Virtual Assistants will be increasingly sophisticated and will start to take over work - freeing us to do more of what we are very good at: cognitive and creative work.