The insurance industry is awash in documents of various types, but one category stands out: ACORD forms. Any insurance company that succeeds in implementing intelligent document processing for ACORD forms will be ahead of the game. But companies that automate ACORD form processing along with other types of unstructured content – well, they’d be dominating the field.
ACORD forms come from the Association for Operations Research and Development, a non-profit that develops standards for the insurance industry, including standardized forms. The index of available ACORD forms runs 13 pages, many of them listing 80 or more forms each. They cover topics ranging from “Automobile Loss Notice” to “Equine Liability Supplement.” Most states have many forms that apply only to them.
Standardized forms – but varied
These days, ACORD forms are completed electronically, but they don’t have to be. Insurers work with forms completed by hand, or at least checkboxes marked in pen or pencil. And given the sheer number of ACORD forms, variation is a given.
The forms are standardized and structured; that’s the whole point. You could automate the processing of ACORD forms using a templated approach, where you define the field(s) from which you want to extract data for input into a downstream processing system. Similarly, optical character recognition (OCR) would work, perhaps with a robotic process automation (RPA) tool.
But any of those approaches require labeling the forms to define the fields for extraction. Given the sheer number of ACORD forms available, that would be a painstaking task.
Related Article: Reduce P&C Insurance Underwriting Risk with Intelligent Automation
Automation for unstructured content
It would also address only part of the battle regarding insurance document automation because ACORD forms come with other content that may or may not be in a structured format. An auto claim, for example, would likely be submitted along with photos, an estimate from an auto body company, the adjuster’s notes, and more. None of that content is structured, so a templated automation approach or RPA tool would be out of scope.
A more effective approach to automating ACORD document processing is to use intelligent document processing (which falls under Gartner’s hyperautomation category). This intelligent automation approach involves using artificial intelligence technologies, including natural language processing, to “read” unstructured documents, including images, much like a human would. Such an approach would enable an insurance company to automate the processing of ACORD documents and nearly any content.
One of the keys to success in document automation efforts is choosing an automation tool built on an extensive database of labeled data points. That’s what gives the tool the ability to identify the context behind documents and images.
Intelligent Process Automation covers all documents and forms
Indico’s Intelligent Process Automation (IPA) platform includes a database of some 500 million labeled data points. AI technology known as transfer learning enables customers to build that database and create models to automate their document processes. With just a few dozen of the actual documents involved in the process, you can create a working model that will have a high degree of accuracy.
So, don’t settle for an automation approach that can only handle highly structured content like ACORD forms. Take the IPA approach, and you’ll realize gains in process cycle times and capacity that will have you far ahead of the field.