Whatever industry you work in, or in whichever interest you may have, you will almost certainly have come across a story about how “data” is changing the face of our world, particularly “big data”. You may have heard this term as part of a study helping to cure a disease, boost a company’s revenue, improve customer service, make a building more efficient or be responsible for those targeted ads we keep seeing.
But we don’t mean THAT “data”!
Despite what term is commonly used, data is simply another word for information. But in computing and business, data refers to information that is machine-readable as opposed to human-readable.
In business, we receive masses of data in human readable form such as contracts, invoices, orders or HR records etc. These documents need to be converted to a machine-readable form so that technology, like RPA, can be used to automate the process end-to-end.
The challenge is to firstly have the creator of the document produce it in a digital format that is also human readable, so that further downstream this can be read, data extracted and passed to a robotic process for downstream automation. Data extraction can be achieved at 100% accuracy if produced in a digital format (if the format contains a text layer).
Images causing havoc
But, where the sender chooses to create an image file, you must rely on Optical Character Recognition (OCR) to convert the text to a machine-readable format. The problem with OCR is that as the receiver has no control over the image quality or how data is presented, the net result is you can never guarantee accuracy and it’s these data errors that cause havoc with the RPA process.
Ensuring the best data for your robots
To make sure your bots do not go awry, the first challenge is getting the sender to create a digital document. To do this, we need to remove any barriers, ensure there is no cost or resource requirement and ideally no process change for the sender. The second challenge is to remove paper or image files that require OCR.
Bad data, big problems
Let’s consider the consequences of bad data for a minute. The impact of misreading a measurement or value could mean an engine part is manufactured to the incorrect size or an order gets processed with the wrong amount, a -10 becomes 100 and so on. Data without context delivers a second layer of complexity, as ‘ea’ could be read as ‘each box’ and not ‘each unit’ etc. There is a clear and obvious need to not only read data accurately but also to understand the context of a data element.
Now consider these challenges at scale and the impact of such errors on ‘big data’ as more of the world’s business processes become digital and move online, the need to process data at scale accurately has never been more important.
RPA for business process automation
In the world of shared services, we have looked to deploy RPA in areas such as invoice and order processing to increase automation and drive efficiencies. Through the implementation of innovative technologies, such as RPA, the human task is rapidly moving from the mundane and repetitive to those of quality control and cognitive value creation. The theory is great, but the reality is that unless the right technology and business process is deployed to convert human readable documents to that of a machine readable format, the data for the RPA bots will always contain errors. You can read more about RPA integration and CloudTrade here.
Technology for data perfection
There is a solution to read digital documents and process that data into a format a machine can read to give bots the right tools for the job.
We’re running a webinar focusing on this integration for RPA, sign up is available here and will address how this proven approach works for RPA, provide a live demonstration of delivering 100% accurate data, and how to automate business processes that will eliminate human intervention.