Practically every organization that operates with a number of complex processes is excited about the concept of intelligent automation and why not? The capability to take manual processes and automate them to reduce effort and time can only be a good thing. And a lot of these processes are so straightforward and repeatable that automation can be implemented with a high degree of accuracy and what “straight through processing” or STP. Straight through processing means that a certain percentage of tasks don’t need any human involvement. The goal is for that percentage to be as close to 100% as possible.
The reality is that there are a lot of processes that CAN be automated very accurately at near 100%. Those processes are often secondary or tertiary processes that, while needed, are generally not core to an organization’s primary business. Think of provisioning a new virtual server, or answering a basic list of questions. Very few businesses have these types of processes as central to an organization’s raison d’être.
STP and Your Core Processes
When it comes to the core processes that provide and support revenue or are required due to compliance, the level of complexity increases significantly, often to the level that a high level of STP is not really possible. Take, for instance, the process of selling the servicing rights of a home mortgage. This involves a lot of documentation review and preparation of key loan facts in order to present to the buyer. However, the complexity of the documents involved typically hinders the ability to achieve high levels of STP.
Consider the process of auditing healthcare providers for quality of care that is performed annually by healthcare plans. Selected medical charts must be separated into individual records and then reviewed by skilled staff, who must spend a lot of time hunting for specific information. Or, take the process of reviewing health plan contracts in order to convert them into structured information. Agreements of varying lengths and structures must be read and specific information must be located and entered into a rules engine used for plan administration of claims.
In each of these cases, the information is complex and highly varied. This means that the need for human intelligence is a key requirement. Does this mean there is no place for automation? Not at all. The key is to understand the nature of the process and where it can be made more efficient. The first common aspect of these processes is that they are typically handled by knowledge workers who often perform many different tasks, some more important than others. The less important, yet critical tasks of data preparation is where the biggest benefit can be had from automation.
How to Leverage Automation
For instance, the case of data preparation for a loan servicing sale can be highly automated through machine learning-based document identification and data tagging. The result is a set of summary loan data along with the linked supporting documentation. This enables a very efficient transfer of data from the seller to the potential servicer. In the case of medical record review, individual records can be identified within a chart and key medical data such as diagnoses and recommended treatments can be identified and tagged. This significantly reduces the time required for medical reviewers to do their jobs. For the data preparation required for efficient plan administration, specific contract terms including medical service reimbursements can be easily located and tagged for review by staff leaving only a few clicks of a button to transfer complex contract language into structured data.
In each of these processes, automation assists knowledge workers in doing their jobs more efficiently by significantly reducing or eliminating cumbersome and time-consuming data preparation work that can represent up to 60% of overall effort. We call this “assistive automation”. Assistive automation is made possible through the application of new machine learning tools that can examine complex document-based information like a human. This means the software learns how to perform a task and how to optimize it as well. Assistive automation has the potential to dramatically change the front office, not as a replacement for staff, but as a way to augment complex staff work letting each worker, both digital and human, focus on what they do best.