In the first three blogs of this 6-part Information Capture series, capture was all about automating the acquisition and processing of documents. But information in documents only represents a subset of all the information required for business processes to be truly automated. Where is the rest? It’s in the form of electronic data trapped in internal and external systems, waiting to be acquired and used by the business process.
Take a loan application process, for example. In addition to any forms of ID, copies of paystubs and tax forms, or proof of address documents that are captured, the application process also requires income and residence addresses to be validated, verification that the person applying is really that person, and other important data about the applicant pulled from internal and external systems and then aggregated. Without automation, bank employees need to log into internal systems and visit numerous external websites to perform these validations, fraud checks and data collection activities, enter the results into the bank’s loan processing system, and manually kick off the next step in the process.
Sophisticated Information Capture systems can automate these activities, spanning both the documents and the electronic data required by business processes. These systems deliver two key components:
- CPA (Cognitive Process Automation) to automate the processing of unstructured data contained in documents and emails – known as the intelligent “head work” of understanding what the document or email is about, what information it contains, and what to do with it
- RPA (Robotic Process Automation) to automate the processing of electronic data from internal and external systems and websites – known as the repetitive “hand work” of acquiring and entering electronic data
What does “processing” mean, exactly, in the context of CPA and RPA? We can break this down into three chunks of work – see graphic.
Acquire:
The previous blog in our series addressed the multi-channel capture of documents coming into the organization. But documents represent only half the equation; how do we obtain the structured electronic data that our process needs, and where does that data live?
This data could reside in the form of invoice information from vendor web portals; pricing data from competitive web sites; data validations from third-party web services; customer databases; Excel spreadsheets; and Citrix systems and other systems of record that run the business (ERPs, AS400s, mainframes, etc.).
Without automation, armies of people must do repetitive, manual work to scrub websites, log in and out of systems and acquire the data through tedious copy/paste or “swivel chair” data entry work between multiple monitors. This is the kind of work software robots should be doing─not people.
Understand:
The robot’s job doesn’t end at data acquisition. Robots can aggregate data from multiple sources into a single source and a single view to free up humans to do what they do best: run the business. The robot can also transform the formatting of the data to match what downstream systems and processes require. Smart robots can even apply AI-based natural language processing technology to understand the content of what has been acquired – more on this in later blogs.
When documents and emails are being acquired from various capture channels, the “Understand” step involves the cognitive transformation of the document into useful business information using document classification/separation and data extraction technologies.
Integrate:
After robots acquire data from various systems and translate that data into a useable form, they can automatically integrate this data to internal and third-party systems. Smart robots can do this even if an API is not exposed, so there’s no more writing and maintaining custom integration code for every system. Robots can also be used to reconcile data between systems, read/write data to multiple systems and fields, and automatically trigger downstream processes.
Sophisticated Information Capture systems can also send captured documents to ERP, ECM and LOB destinations without the need for custom integration work.
CPA and RPA: Better Together
Use cases abound that require the automated processing of both documents/emails (Cognitive Process Automation) and electronic data (Robotic Process Automation) within a business process. Here are just a few real world examples of CPA and RPA in action:
Invoice Management – RPA to Acquire
This use case illustrates an example of good vendor relationship management. RPA downloads invoices and other documents from partner portals, and for non-PDF documents, cognitive process automation uses document capture and transformation to read the documents, categorize, add to workflows and send to employees for approval. This automated process speeds time to payment and improves relationships with vendors. RPA then integrates the documents and data with an ERP.
Customer Verifications & Approvals – RPA to Understand
Know Your Customer (KYC), Customer Due Diligence (CDD), Anti-Money Laundering (AML), and all the other compliance acronyms aren’t just an operational issue—the delay in verifications causes a long application process for the customer, too. In this use case, document capture and transformation extracts, indexes and classifies loan application documents, and then sends identity verification information to RPA to perform KYC checks, resulting in customer approval within minutes instead of days.
New Customer Applications – RPA to Integrate
When a new customer applies for insurance, a paper-driven process can take weeks, while a digital process can be anywhere from a few days to a few minutes. In this use case, an insurance company moved to a completely digital workflow using document capture and transformation to extract and classify application information and supporting documentation, and RPA to feed the content into business applications and additional workflows.
Learn more about how the one-two punch of Robotic Process Automation and Cognitive Process Automation can automate your information-intensive business processes by downloading this eBook here.