Improve accuracy of rates and speed-to-market.
Eliminate data entry and focus on higher value work.
Reduce wait time for policies, assistance, and claims.
Foundation AI has fueled and expanded the scope of Automation use cases across the Insurance value chain, providing significant capabilities that enable business transformation and improve insured and employee experience.
AVP of Innovation and Strategic Integration - The Hanover Insurance Group
Foundation Extract automates document classification and content extraction, enabling end-to-end automation across insurance operations. It combines best-of-breed computer vision, natural language processing, and machine learning models to process complex unstructured documents widely used in insurance.
The Foundation Extract Platform provides a number of Intelligent Document Processing solutions that can be delivered through a customizable and configurable low-touch automation workbench. AI automates the majority of the work, and your operations team steps in only as necessary to keep processes moving. Validation and Exception handling is handled through an intuitive user interface designed to learn from every human interaction, continuously increasing the AI accuracy and throughput automation with each page processed. With its robust API and suite of customizable connectors, Foundation Extract seamlessly integrates with most enterprise software, iBPM platforms, and RPA bots to automate manual tasks across the enterprise.
Underwriters must quickly gather and analyze information across disparate sources to appropriately price risk and write new business. Critical information spread across submission documents must be manually keyed and then rekeyed into multiple downstream systems to rate and generate quotes. By the time carriers complete this labor-intensive task, the customer may have taken their business elsewhere. Most carriers rely on outsourced labor to gain efficiency, but, with the cost of outsourced labor steadily increasing and the quality of output continuing to decline, many are realizing that outsourcing is no longer an effective solution. Unfortunately, because traditional OCR cannot handle the multitude of document types and formats necessary to write new business, most automation initiatives have failed to remove the data entry bottleneck, and carriers have been left with little alternative to manual labor.
Foundation Extract processes all types of submission documents. It can intake submissions directly from your designated mailboxes or content management systems, sort and categorize the contents, and extract the requisite data directly from unstructured emails, ACORD applications, loss runs, and other submission documents. Extract applies your context specific underwriting rules to identify the best information from across submission documents, and integrates seamlessly to route clean, perfect data into your rating, quoting, and content management systems.
With coverage needs constantly changing, carriers must deal with an increasingly high volume of endorsement requests, cancellations, and renewals. Altering coverage and processing renewals requires agents, raters, and underwriters to key information from emails and attached documents into multiple downstream systems to re-rate and renew policies, as well as reflect endorsements.
Foundation Extract can identify endorsement requests, automatically capture pertinent information from email bodies, ACORD forms, and attachments, and apply your specific business logic to gather correct information for processing. With the Foundation Extract workbench, information is validated as necessary by raters and other operations employees, then automatically entered into downstream systems, reducing more than 80% of the manual labor involved with processing inbound requests. This allows your employees to spend time on more profitable work like sales and underwriting new business.
Carriers and Third-Party Administrators (TPAs) are overwhelmed with incoming emails from agents and insureds, ranging from customer service inquiries to coverage questions and endorsement requests. Today’s customers demand immediate answers, but before inbound emails can even be responded to, they must be reviewed, indexed, and routed to the appropriate department. This manual triage process can add days to response time. Many carriers have outsourced this manual sorting and routing to cut this time delay incrementally, but poor-quality classification and improper indexing has led to even more delays and dissatisfied customers.
The Foundation Extract platform instantly classifies and indexes your incoming emails, then routes them to the correct work-queue with better-than-human accuracy. Extract processes thousands of emails a day for a major US P&C carrier, taking a previously manual process with a 24-hour SLA and reducing it to under 24 seconds. Not only does Foundation Extract classify and route emails, but it also automatically extracts policy numbers and other requisite data to complete routine workflows like cancellations, renewals, and endorsements with little to no manual intervention
Claims departments are overwhelmed. Correspondence floods in from insureds, external claimants, attorneys, physicians, and other case participants, all of which must be reviewed, matched to the correct claim file, and indexed before it can even make its way to an adjuster to be acted on. Because most of these incoming documents do not specify the claim number, claims representatives must manually search claims management systems using identifiers like names, addresses, DOBs, dates of loss, and VIN numbers to locate the correct claim file. This labor-intensive process increases the cost of claims handling, delays the speed with which carriers and TPAs can respond, and unnecessarily increases combined ratios due to increased operating expenses.
Extract Filer intakes inbound correspondence directly from electronic sources, like eFax and scanners, and applies natural language processing and machine learning models to automatically classify documents according to your document management rules, identify the correct case file, and index the document directly into your downstream systems with a descriptive title. Extract cuts the labor cost and cycle time associated with indexing incoming claims documents by more than 80%. Extract Filer can even be configured to apply complex business rules to extract specific information from different types of documents and to trigger context specific alerts and workflows, like auto-updating claims notes, calendaring events, and alerting adjusters whenever they receive specific documents, like settlement requests.
The First Notice of Loss (FNOL) – the first step in claims processing – is one of the most crucial touchpoints for an insurance provider or TPA. Despite the trend toward digital FNOL processes, most insurance providers and TPAs still receive a high percentage of FNOL documentation through emails, faxes, and physical letters from customers, brokers, agents, claimants, and others involved in the process. After claims representatives identify, classify, and index the FNOL from the thousands of documents received on a daily basis, they must complete the manual, time-consuming process of entering information from these emails, ACORD forms, PDFs, Word and Excel Files, and images into claims and policy management systems to launch downstream claims adjudication workflows.
Foundation Extract automatically identifies FNOL reports, indexes each document, using extracted data like the insured’s name and address and policy number, and captures the information necessary to initiate a claim, like the date, description and cause of the loss, and enters it directly into your claims management system. By using Foundation Extract to process FNOL documents, you can substantially reduce labor costs, while also decreasing cycle time for claims initiation by more than 80%, resulting in more responsive claims handling, lower loss ratios, and higher customer satisfaction and NPS.