Why does Skan think Process Discovery - the art and science of observing work at scale, and deciphering the process nuances – is a superior method of understanding process permutations and variations as compared to process mining?
The answer is process discovery happens between committed states of data, and this where much of the magic of work occurs as well. Whereas traditional application log analysis-based process mining focuses on analyzing the hard data commits, it misses the bigger picture of what transpires between those recorded events. (Of course, another challenge for process mining is the many applications, beyond the transactional ERP systems, may not necessarily record event data thus rendering process mining of such applications moot.)
How Process discovery happens committed states of data?
Let's take a couple of examples to demonstrate what process discovery does and how it is vital to get an accurate picture of the business processes.
Let's assume as a part of a business process, an analyst opens two different applications and manually copies and pastes information from one to the other. This piece of the process, which seems repetitive and redundant, could potentially be a candidate for automation. But the log files will not show these types of events.
Or imagine as a part of another process, the process agent has to check the social media accounts and postings of the customers through a browser interface. This information is impossible to fathom by event log analysis.
Last but not least, let's imagine the process steps include performing a calculation in Excel or sending an email from Outlook. Now, these two applications are not a part of the transactional processing system and hence go unrecorded.
So, how Process Discovery is a superior method to document business processes? (Please visit here to learn about the key differences between process discovery and process mining.
The premise and promise of process discovery are to observe work – at a vast scale, high precision, and unobtrusively – to record process participants' interactions with digital systems and from these digital traces of human labor, decipher a comprehensive and multi-dimensional business process and its innumerable variants.
Process discovery goes beyond the recorded events and instead focuses on observation. The technologies that power process discovery includes Computer Vision, Machine Learning, and Data Science techniques. And all this human level observation and machine-level precision capture the nuances of work between committed states of data. All this power and accuracy without the need for backend integration or batch file uploads.
The word observation, computer vision, and recording may raise alarms in many executives' minds who may have concerns about personal privacy and information security. These are valid concerns, and Skan, an AI-enabled process discovery platform has adequate safeguards to address the privacy and security concerns.
Skan only records what is on the screen and not who is in front of the screen. Second, if companies chose not to attribute work to an individual, Skan technologies can anonymize the users. Selective screen level masking or field-level redaction allows for protecting confidential information. And Skan's core machine learning and data science algorithms work on metadata and not the contents of the screens. Of course, to stitch together a business process, a field such as Case ID may be an essential data element vital to the training and working of Skan's models.
After processing the images and stitching together the business process, the outputs are verifiable, data-driven evidence that process owners, consultants, and process experts can leverage to conduct various interventions.
The process metamodel and the digital twins allow for efficient automation, effective digital transformation, precision process-specific just-in-time training and help, process conformance, and AI-supervisory.
Contact the Skan team to learn how process discovery can unveil your hidden enterprise and help you unleash the processes.
