Robotic Process Automation: Powering Intelligent Records Management
Our guest blogger K. David Quackenbush, Chief Executive Officer of Gimmal has been kind enough to let us share with you his three articles on ‘Intelligent Records Management’. Gimmal, a Concept Searching partner, is recognized as a leader in the field of records management and content governance. The software is known for establishing a new standard for records management by providing the unique ability to manage information no matter where it exists. This ensures important information is managed across all systems with audit-ready policies in SharePoint and beyond. The below, ‘Robotic Process Automation: Powering Intelligent Records Management’ is the second article in the three-part series. I think you will find it interesting.
It’s amazing what is being done today with robotic process automation (RPA) tools. Once confined to rote processes that were easy to automate due to the repetitive nature of the work, RPA is now tackling more complex business processes and activities. This evolution allows “knowledge workers” to spend more time focusing on actual knowledge work such as critical thinking and real problem solving – one of the key benefits of implementing an intelligent records management program.
When you think about the problem of runaway data growth that plagues most enterprises today, the first thoughts are typically overwhelming: that these information governance problems are not solvable, that effective solutions are too expensive, and so on. Efforts to manage runaway data are usually stalled by the first question organizations ask themselves when getting started: “where do we begin?”
The enormity of this task can rob enterprises of the opportunity to build a day-forward strategy for truly effective records management. Organizations need to get in front of the problem and then take a more moderated and time-based approach to dealing with the avalanche of previously created (and ever-growing) data.
This is where the proper strategy, coupled with the right choice of automated processes, can achieve real results that are cost effective, don’t require an army of resources, and ultimately improve the quality, accessibility, and usability of your information assets. We will save strategy for the next part of this series. First, let’s take a look at how automation can help.
Applying layers of intelligence to process automation
First, imagine if there were a way to continuously “look” at all of the unstructured data you manage and make this information a little smarter. This is an iterative exercise where “robots” have a particular job. Assume that we need five layers of intelligence, each addressing different levels of abstraction and different types of information, added to unstructured information to make it more valuable, findable, and usable.
Tasking one robot or automated process to handle this whole task is not feasible as competing priorities and complex situations would make the logic far too complex to maintain. The separation of concerns is an important principle in any information architecture or IT initiative, because the more problems you try to solve with a single process, the less effective that process will be. By separating the generated knowledge into multiple layers, your system will be able to compare that information and prioritize these goals without introducing ambiguity.
Second, as each process adds layers of intelligence, it also starts to make decisions (or at least suggestions) regarding how that information is managed and where it should be stored. Most enterprises have an idea of where they would like the majority of unstructured information to live, but struggle with making that movement happen without intervention. But there will always be unique situations that process automation cannot yet address.
Exceptions: a job for humans
This is where humans, specifically records managers, factor in. To solve for unique exceptions, each layer discards records that do not fit into typical automated processes into queues that require human intervention. This, perhaps, is the most complex function needed for intelligent records management, as defining the number of different exception queue criteria is important.
Concepts that can drive these exceptions include guessing the most likely content owner, the relative value of content, and data type. In determining the relative value of a record, an organization can assess the date the file was last accessed. Where data type is concerned, audio or visual media may need a different level of attention than more easily indexed content such as Word documents.
Even in RPA, technology does not stand alone
RPA, and automation in general, is the path forward to solving the runaway data management challenge faced by most enterprises today. However, like all technology-based solutions, this is not about just technology, but also the people and process that will yield better results. The next part of this series will address the people and process elements as part of a broader discussion of strategy.