Estimated reading time: 4 minutes.
In some ways, document capture is old hat. At first glance, it’s not sexy or exciting because it’s been done in various ways since at least the mid-1980s. There’s no doubt that over time things have changed. Originally, the focus was on scanning paper to create electronic document archives which closely mimicked their physical counterparts. This would allow organisations to repurpose their paper stores, meaning the space could be used for other purposes. There was nothing “cognitive” about this data capture!
Over time, capture evolved. It became more transactional in nature, integrating with business processes and expanding the repertoire of document sources. Nowadays, it’s possible to ingest and process feeds from social media. One of the most interesting developments for me has been the emergence of so-called Cognitive Capture. Rather than just capturing and archiving documents, or processing invoices, we can now extract information which can feed into and influence downstream business processes.
This article looks at three ways your commercial operations can be streamlined using cognitive capture.
ONE – Cognitive Capture can answer questions about your documents, such as:
• What is this email about?
• What is the underlying sentiment of this document? Was the author angry? Were they sad? Were they happy?
• What information does this document contain?
• What should be done with the document and information?This information is surfaced using a variety of techniques such as Natural Language Processing (NLP) and machine learning which classifies, separates and understands documents. NLP extends the capability of capture systems by understanding text-rich documents like correspondence or loan applications. Relevant information such as addresses, amounts or dates is extracted. NLP can also surface the sentiment of a document. This can be useful in environments such as customer service where incoming complaints can be identified and prioritised.
Machine learning can streamline the configuration and maintenance of a modern Capture system. From relatively few samples, a modern system can classify and extract data without the need for rules or extensive coding.
TWO – Extracted data can be cross-referenced or augmented using your existing systems of record to ensure accuracy and formatting rules can be applied so that data is correctly presented to downstream systems. E.g., does the date of birth extracted from a document correctly correspond with the extracted policy number? If there are doubts concerning the quality of the data, human-in-the-loop capabilities make it possible to route documents to a human for validation.
Once data validation and formatting have taken place, modern capture systems integrate with downstream business processes or systems through either connectors or APIs. If no connector or API exists, these can either be written, or a Robotic Process Automation (RPA) process built to provide the “bridge” between the capture system and the system of record.
THREE – One of the biggest challenges in document capture is ensuring that documents and data are of sufficient quality when they are delivered to downstream processes. There’s a well-known I.T. proverb: Garbage in, garbage out. That is very true in the capture space also. If the incoming document is poor, the resulting image will also be poor. High quality images are essential for cognitive capture to function optimally. Poor images lead to illegible documents and broken business processes, which consequently result in reduction in service quality.
One of our partners, Kofax, patented an image enhancement technology called Virtual Rescan (VRS), which actively monitors and improves the quality of captured images, ensuring that the best possible images are always produced regardless of whether they are being physically scanned, or electronically delivered. Although VRS has been available for several years, it has constantly evolved and in its latest guise, is even present in Kofax’s mobile capture application which enables high-quality document capture on smartphones and tablets.
Conclusion
Advances in cognitive capture allow us to take unstructured documents from any source and transform their content into structured, actionable data which downstream processes and business systems will consume. The possibilities for this are exciting, opening potential use cases in customer service, healthcare, insurance and banking.
Ready to learn more about Cognitive Capture? We partner with two industry-leaders in the capture space – Kofax and IBM.
Check out our Cognitive Capture page or Contact Us to get started on the journey of unlocking the value in your business documents.
Other Blog articles related to Cognitive Data Capture:
IBM Datacap 9.1.9
Kofax Capture 11.1
Implementing IBM Datacap Web Services in Azure
Intelligent Invoice Processing
Kofax Cognitive Document Automation
Open Connections is an innovative provider of Business Process Automation solutions. We have a wealth of expertise and have helped many organisations find solutions to their document-centred problems.
