Should you build or buy Payer Core Administrative Processing System?
Payer core administrative processing systems handle the system-of-record work for a health plan: claims adjudication, member enrollment, provider contracts, billing and appeals. Health insurers, Medicaid managed-care plans and third-party administrators buy or build this backbone to process every claim and premium transaction they run.
Buy or retain the core. Oscar proves a payer can build its own claims platform, but it did so with a large engineering staff over many years and still uses AI only to assist staff, not to adjudicate. For a mainstream plan the core carries benefits, contracts, enrollment, billing, appeals and regulatory change on one accountable system, and the vendor market serves that at fixed PMPM. Build the workflow, analytics and member-facing layers around it.
Build it, buy it, or bridge?
When building makes sense
Oscar Health is the clearest example of a payer building its own core administrative system, and Oscar itself describes the choice as owning its entire technology stack, including a custom claims platform its engineering and design teams built and continue to iterate on, not a side project bolted onto a legacy system. That took years of sustained investment from a company organized around being technology-first from its founding, with a scale of engineering staff most health plans do not carry. Other technology-forward payers have taken a different path: Clover outsources core operations to UST HealthProof rather than building, and Alignment Healthcare runs on TriZetto Facets. No AI-native adjudicator exists anywhere in the public record; even Oscar's own AI work assists its claims-support staff rather than replacing the adjudication logic itself. Build only if your organization is prepared to treat the core system as a permanent, well-funded engineering function for the life of the company, the way Oscar has.
When buying makes sense
Buy or retain a core administrative platform for anything short of a technology-first insurer's engineering commitment. TriZetto Facets processes claims, enrollment and billing for a large share of the payer market, and HealthEdge HealthRules and HealthAxis HealthOS serve the same role for plans that want a different vendor relationship. These systems carry benefit configuration, provider contracts, enrollment, billing, appeals and the constant stream of regulatory change that comes with running a health plan, on one accountable system priced predictably on a fixed PMPM plus transaction fees. A custom build has to reproduce all of that before it even matches what a mainstream vendor already does, and the only company that has pulled it off did so with a scale of engineering investment most plans will never fund. License the core, and put your differentiation into the workflow, analytics and member-facing tools you build around it instead.
The desk read
Build-versus-buy analysis for Payer Core Administrative Processing System is being written. In the meantime, the framework that drives every B4 call is on the B4 Index page.
Frequently asked
Has any health plan actually built its own claims platform successfully?
Oscar Health is the clearest case, running a custom claims platform it built as part of owning its full technology stack. That happened at a company founded to be technology-first, with a large sustained engineering investment, which is a different starting point than a mainstream plan considering a build today.
Does AI make building a claims adjudication engine realistic now?
Not yet. No AI-native adjudicator is documented anywhere, including at Oscar, where AI assists claims-support staff rather than replacing the adjudication logic. The hard part of a core system, benefit configuration, regulatory compliance and provider contract logic, is not the part AI coding tools currently shorten.
What's the real cost comparison between building and buying a core system?
Directional estimates for a custom build run from the high hundreds of thousands to several million dollars over 18 to 30 months, plus 18 to 25 percent of that in annual maintenance, against a vendor's fixed PMPM plus per-claim fees. The build estimate typically excludes migration and dual-running costs, so the real gap is likely wider than the headline numbers suggest.
Why do some payers outsource core operations entirely instead of licensing software?
Clover Health outsources its core administrative operations to UST HealthProof rather than running the software itself, which trades platform control for avoiding both the build and the in-house operational burden of running Facets or HealthRules directly. It is a third option between building and licensing that suits a plan that wants to focus entirely elsewhere.