Nara Health Raises $14M — AI Is Moving Into the Claims Administration Layer

The former Avant Health has rebranded as Nara Health and raised $14 million to scale an AI-native third-party administrator. The funding matters. The bigger RCM signal is where the AI is being deployed: inside the infrastructure that administers benefits, adjudicates claims, coordinates care, and answers members before provider revenue cycle teams ever see the claim.

What Nara Is Building

Nara Health is not positioning itself as another point solution for denial prediction, claim-status checks, or appeal drafting. It is rebuilding the third-party administrator layer for self-funded and alternative employer health plans.

The platform combines benefits administration, claims processing, care orchestration, and member support. Nara says it uses AI to make alternative plan designs — including direct provider contracts, cash-pay arrangements, reference-based pricing, and direct primary care — easier to operate at scale.

That is a different place in the healthcare transaction than most of the AI companies RevCycleAI tracks. Provider-side RCM vendors generally automate work after a patient has been scheduled, treated, coded, or billed. Nara is operating on the health-plan side, where the rules governing payment, authorization, navigation, and adjudication are applied.

The RCM Signal: AI Is Moving Upstream

Most of the first wave of RCM AI focused on labor substitution: eligibility calls, coding review, claim-status follow-up, denial letters, and patient collections. The second wave is increasingly about changing the underlying workflow itself.

Nara is a useful example. The company says its technology supports same-day claims adjudication and same-day prior authorization turnaround, while synthesizing signals from claims, pharmacy information, electronic medical records, calls, texts, and emails. If that model scales, providers interacting with these plans should experience a materially different administrative environment than they do with a traditional TPA.

• Faster adjudication changes follow-up economics. If clean claims can be adjudicated the same day, traditional aging-based work queues become less relevant for that payer population.
• Prior authorization becomes an integrated data problem. Nara is connecting authorization with care navigation and plan administration rather than treating PA as a separate fax-and-portal workflow.
• Alternative payment paths become operationally viable. Direct contracts, cash-pay routing, and reference-based pricing create different reimbursement mechanics than conventional network claims.
• The payer-provider data boundary gets blurrier. A TPA with near-real-time clinical, claims, pharmacy, and member-interaction data can make decisions earlier than systems relying primarily on lagging claims feeds.

A New Competitive Pressure on Legacy TPAs

Nara enters a market where self-funded employers increasingly want more control over healthcare spend but still depend on TPAs to make complex plan designs operational. The company says some employer health plans on its platform have reduced costs by more than 50% versus the prior-year plan, while one customer cited a 55% year-over-year reduction after moving from level funding to self-insurance.

Those are company-reported results, not a market-wide benchmark. But they show the economic thesis: a modern TPA is not just competing on claims-processing accuracy. It is being asked to actively route members toward lower-cost care, support alternative reimbursement arrangements, and make administrative decisions fast enough to influence utilization before the bill arrives.

That puts pressure on legacy administrators whose technology stacks were built around batch claims processing, fragmented point solutions, and delayed reporting. It also creates a new buyer category for AI infrastructure vendors: the administrator itself.

What Revenue Cycle Leaders Should Watch

For providers, Nara is still small relative to national commercial payers. But the model is worth watching because self-funded employer plans can introduce reimbursement variation that is easy to miss inside standard payer buckets.

Revenue cycle teams should watch four things as AI-native administrators grow: whether authorization turnaround actually translates to fewer care delays; whether same-day adjudication produces faster and more predictable payment; how reference-based pricing and direct contracts appear in contract-management and underpayment workflows; and whether denials become less frequent or simply more automated.

The larger takeaway is that AI is no longer confined to automating the work surrounding a claim. It is beginning to reshape the systems that decide how the claim is priced, authorized, routed, and paid.

That is a much bigger shift than another denial bot.