FINEOS

AdminSuite runs new business, policy, billing, claims and absence management without any model, and the generative and predictive features assist case managers inside that workflow. Remove them and the administration platform carries on, slower to read documents and prioritize claims.

Thresholds for when automation acts without a case manager are not published. FINEOS describes AI governance policies built on transparency, decision ownership and a human in the loop, and each insurer can raise or lower the level of AI assistance, automation and validation to match its own AI policy and risk management.

Accuracy, validation and monitoring figures for the claims triage, transition, duration and settlement models are not published. Their scores are visible on dashboards, portfolio views, case summaries and the case record, so a reviewer can see what each model said about a claim.

Customers are named and scale is specific, though no outcome figure is attached to the AI. New York Life Group Benefit Solutions, Beneva, Guardian, OneAmerica Financial, MAIB, Aviva, Sun Life Financial, Principal Financial Group and TAL, a customer since 2012, are named, and 7 of the 10 largest employee benefits insurers in the United States run on FINEOS.

Scale and a migration program, rather than a performance result, are what the Accident Compensation Corporation reports: about 2 million accepted injury claims a year on the platform, after a cloud migration that by its own account took over 2 years and was delivered on schedule.

Whether one insurer's claims data trains predictive models offered to another, and which cloud AI platforms the generative features use, is not stated. The predictive models are described as purpose built for life, accident and health and ready out of the box, which implies training on data beyond any single customer.

Retention, processing and subprocessor terms for the claimant, employee and medical information the platform holds are not set out in its product material, a gap for a system carrying disability, leave and accident claims. The privacy posture rests on the insurer's own controls inside a cloud platform described in general terms.

A service organization control report is signaled without its type, scope or auditor, and no ISO certification or trust center is offered for review.

FINEOS supplies software to licensed insurers and holds no license of its own. Its absence management material tracks state paid family and medical leave programs, so leave rules are encoded in the product rather than asserted as a regulatory standing.

Fairness or bias testing of the claims models is not published, though predicting which disability claims will transition, last longer or settle shapes how claimants are handled. Governance appears as principles, transparency, decision ownership and human oversight, rather than as published testing.

A guarantee, indemnity or correction commitment behind a model's prediction is absent. Decision ownership stays with the insurer's case manager, and model scores sit on the case record where they can be reviewed and challenged.

The cloud AI platforms behind the generative features and the hosting provider go unnamed. The predictive models are FINEOS's own, built for life, accident and health claims, and Limelight Health, acquired in 2020, supplies quoting, rating and underwriting.

FINEOS is the administration core itself, from quote to claim, and connects to the insurer's customer relationship management, accounts payable and receivable, general ledger, document management, communications, human resources, enrollment and data warehouse systems. Components can also run beside an existing policy administration system before a full replacement.

Neither the hosting provider nor the regions it uses are named. FINEOS runs as software as a service and has moved long standing on premises customers such as the Accident Compensation Corporation onto its cloud platform, with offices in ten locations serving markets from the United States to Australia and New Zealand.

Price, tiers and the unit of charge are unpublished, and buying starts with a contact request. Component implementation beside an existing system is offered, which shapes how a purchase can be phased, but not what it costs.

Life, accident and health insurers and employee benefits carriers are the buyers, across group, voluntary and individual business, with separate material for absence management for employers, new business and underwriting, and public accident insurers such as New Zealand's.

The customer list reaches further, to FBD, an Irish property and casualty insurer, Manitoba Public Insurance and the claims administrators Gallagher Bassett and McLarens, though its market solutions are employee benefits, accident compensation, individual insurance and global health.