SAS code guru: Beyond the hype in hyperautomation

This is a guest post for the Computer Weekly Developer Network written by David Shannon in his role as head of hyperautomation at SAS UK & Ireland.

This article explores the evolution of automation from the origins of computing to modern-day hyperautomation, with respect to the impact on people that create and consume services.

In Shannon’s view, computing is automation – and equally – automation is computing.

Shannon writes in full as follows…

Automation is the principle for which computers exist.  

Whether you consider the Turing Machine to be the first computer or otherwise, this and everything that followed allowed repeatable and consistent processes. Performed at speeds and complexity which exceeded the ability of people who would otherwise calculate by hand, computing is valuable automation.

From the late 1960s fourth-generation scripting languages (4GL) allowed database programmers and statisticians to automate many of the compiler-level tasks that would otherwise be required to instruct a computer to do something complex. This made repeated use of data and complex mathematics available to those with relatively modest computing skills.  

This brought automation to a larger workforce.  

Automation elevation

Today, digital transformation programmes are pervasive as organisations digitalise their processes; many already have.  

Digitalisation allows the interaction between stakeholders and the process to be fully automated. Digitalised processes may interact with multiple systems and react according to the input of a stakeholder. 

For example, requesting a replacement debit card can be completed by answering questions on a banking app. The card can be requested, produced and fulfilled potentially interacting with CRM, fraud, risk and ERP systems along the way.  

This allows organisations to scale by making services always available without needing an equivalent scaling of the workforce. Simultaneously, it allows more efficient and frictionless interaction between that process and its end consumer, whoever and whatever that end consumer is. We often discuss automation in the context of interacting with customers, when in fact, consumers of automated processes may be the workforce, suppliers, partners and customers.  

This level of automation moves beyond programming. Software provides graphical interfaces allowing workflows to be constructed of actions, commonly referred to as low-code or no-code software.  

Those actions may be trivial, such as responding to an email. However, automated actions are increasingly advanced. For example, Artificial Intelligence (AI) is routinely applied to determine if the interaction with a service is likely to be fraudulent. The resulting decision will trigger business rules and processes to handle the scenario appropriately.  

Hyperautomation defined

Hyperautomation is the convergence of multiple advanced capabilities, composed of artifacts requiring different skills. Workflows are created by Robotic Process Automation (RPA) specialists. Risk and fraud AI models will be trained by data scientists with specialist skills in the domain, who continue to monitor their models ensuring they behave responsibly, as well as making effective decisions on behalf of the organisation.  

SAS code guru Shannon: Hyperautomated processes are always available for consumers.

Hyperautomated processes are always available for consumers.

This is commonly achieved thanks to highly resilient cloud environments. Cloud and DevOps specialists deploy and update processes as quickly as modifications are made and approved. Cloud and DevOps are further skillsets, but require no technical knowledge for the RPA or AI used in the process being deployed  

Organisations who use automation to interact with people – be it customers, employees, partners or suppliers – are concerned with ensuring a human intervention is made at the right time. Described as human-in-the-loop, it ensures decisions can be made not only with AI but acted on with empathy. In this scenario, people are informed by AI, bots and decision-making, but the action is performed by humans.

Displacement dilemmas

The aim of organisations applying automation and hyperautomation is all of cost benefits, productivity and innovation. One of the major concerns is [human] displacement because of AI and automation. In practice, roles evolve rather than being displaced. Anecdotal evidence suggests that organisations who succeed with hyperautomation increase their workforce because they can innovate and offer a greater breadth of products and services.

Consumers also benefit from hyperautomation.  

By interacting with services at a time and frequency of their own choosing it creates a frictionless customer experience. At SAS we have introduced the concept of the ‘AI Bank’ where we demonstrate how a consumer might, for example, arrange a loan to pay for a car. The entire process is completed online in a simple, efficient and speedy way.

However, it is important for organisations to allow interaction with people in their organisation through channels of the customer’s choosing. Those connections maintain trust and are most valued in times of distress and crisis, such as insurance claims and dealings between citizens and government services.  

Automation should elevate everyone by allowing the workforce to be more productive and innovative.  It should create better experiences for customers and improve productivity between partners and suppliers.  

About the author

David Shannon has been a director and consultant in data, analytics, AI and automation for over 20 years.  He provides strategic and tactical advice delivering cost benefits, productivity and innovation. Today, Shannon works for SAS and leads the UK & Ireland’s hyperautomation agenda helping organisations drive digital transformation through AI and automation. Outside of SAS, he is the IT director for The MG Car Club.

Data Center
Data Management