The ability to succeed in any economy requires processes that can sustain growth at scale. To get there, supporting technology and IT governance must be evaluated in the context of present and future states. What’s more, how business processes manage complex activities, while also supporting continuous awareness of business conditions to inform real-time decision making, requires a seamless partnership between humans and the machine environments in which they operate.
By extending process management from process logic to business logic, enterprises can embrace a cognitive approach to business process management (BPM), which offers flexibility, agility and adaptability in evolving and complex business ecosystems.
BPM & Cognitive Computing in a Nutshell
BPM refers to the systematic method to improve business processes. More often, it involves cohesive systems that extend beyond the management of people and information. BPM experts study, recognize, manage, optimize and monitor business processes that support enterprise goals.
Over the years, these activities have evolved significantly to include systems that can learn at scale, employ logic and reason, and interact naturally with human beings. We call this extended form of BPM “cognitive computing.”
Cognitive computing systems are not explicitly programmed, but instead are trained like humans. They gain experience and hone processes over time, and manage both structured and unstructured data using artificial intelligence (AI) and machine learning (ML) algorithms. They can also adapt to new usages and formats in real time, just as humans do.
Defining Cognitive Process
A cognitive process is one that makes BPM more dynamic and probabilistic by enabling decision management systems to understand, evaluate and comprehend business events. For instance, cognitive data management can fuel process automation via machine learning algorithms, offering businesses the potential for massive returns. In conjunction with social, mobile, analytics and cloud (SMAC), smarter processes are more than just a tool for planning and execution; they are an intelligence engine that delivers smart and contextually relevant decision-making based on operational insights.
Combining business processes with machine learning produces cognitive solutions that can enhance customer experiences. These solutions compare the data that business activity generates with data from other sources, enabling dynamic and timely decision-making.
This opens the door to an entirely new level of straight-through processing. For example, in a predictive maintenance process, machine-learning algorithms can be applied to sensor data to identify situations, which indicate that a machine breakdown is imminent.
The evolution of business process management is synonymous with the decades-long evolution of the automobile industry. Just a few years back, driverless cars seemed like fiction, but now they are nearing reality. In the same way, basic workflows are evolving into cognitive automated processes (see Figure 1).
Specifically, these capabilities can be bifurcated by:
Enhanced decision-making and optimization: Cognitive computing enables a business process to make decisions on behalf of humans based on rich experience and large amounts of unstructured data. By digesting a myriad of unstructured information, cognitive systems can inject intelligent insights into the decision-making process. Using predictive and adaptive decision capabilities, they also add value to preventive decision-making.
Advanced intelligent automation: Cognitive agents can process human interactions over any preferred communications channel. Using AI and ML, these systems can effectively capture insights and codify process specifications, revealing new automation opportunities that can be leveraged using robotic process automation (RPA) to augment and mimic human intelligence.
Cognitive Process Transition and Adoption
To be cognitive, the process must think and learn on top of traditional constructs. We break this down into processes that:
Do, by enriching the traditional process with knowledge.
Think, by enhancing the system with decision-making.
Learn, by expanding the business with insights.
The overall approach can be subdivided into four high-level phases:
At a high level, the journey to cognitive processing starts with collaborative discovery to learn and define existing business processes in a launch workshop (the “do” part).
The next phase is to define actionable insights captured from actual process usage and business pain points.
In the design phase, the future cognitive process model is identified along with a strategy to extract insights (“think” and “learn”) from non-structured data.
Finally, the identified, recognized and explored capabilities are implemented using prototypes for testing in real-life use cases.
Cognitive Business Processes Applications
Cognitive business operations can be applied across industries and functional areas of an enterprise. For example:
Healthcare. Hospital care management systems can leverage data from social media to examine the spread of diseases and track the outbreak of epidemics. For example, during the outbreak of dengue fever in a city, hospitals can monitor Twitter feeds to identify symptoms experienced by the public.
Banking. When customers are approved for a loan, they are directed to the bank’s loan-servicing department, which ensures proper payment collection, as well as any changes to the payment plan. This involves inbound and outbound calls that generate call transcripts. By applying cognitive analysis to this process, the bank can then determine whether its employees are asking the right questions, how polite they are, and whether they are working efficiently.
The human touch. Companies can use cognitive technologies to analyze information from customers in the form of letters, email or other communications. For instance, when handling customers with strong negative sentiments, companies can deploy sentiment analysis.
Improved decision-making. In recruiting, managers faced with hundreds of applications for dozens of openings typically spend enormous amounts of time trying to identify the best candidates, using just simple intuition and other limited tools. Cognitive BPM can change all this, as it looks beyond the formal attributes of candidates (such as their degrees or years of work experience) and incorporates more modern techniques of data collection.
Cognitive Process: Goals and Benefits
Cognitive BPM supports self-learning and adaptive BPM systems, enabling seamless interaction between systems and humans to achieve better results. Cognitive BPM achieves this in the following ways:
Connecting customers, contexts and content across a multiplicity of channels.
Implementing an anytime, anywhere and anything (AAA) process model.
Developing automated, adaptive and predictive decision-making capabilities.
Establishing intelligent, connected and contextual interactions with users.
Automating agent-oriented processes.
Traditional workflows based on predefined process logic offer little support in today’s complex and dynamic digital business environment. By infusing cognitive computing, BPM moves beyond traditional process automation and optimization, and can advance the digital business goals of large and small organizations.
The continuous awareness, gradual learning and real-time decision-making of a cognitive approach are essential to managing modern business needs, and can drive any process automation, no matter how complex.
Cognitive technologies make automation possible across all enterprise domains. The organizations that adopt these technologies today will have a distinct competitive advantage once the cognitive BPM revolution spreads.