Gemini Health Partners has launched an AI-focused RCM advisory practice for high-cost specialties including radiation oncology, medical oncology, radiology, urology and orthopedics. The consulting offer is interesting. The underlying asset is more important: Gemini is basing its recommendations on real adjudicated claims and prior-authorization outcomes from revenue cycles it already operates.
Gemini said in July that it had processed more than $180 million in specialized medical claims and prior authorizations since launching in October 2025.
Most AI strategy engagements start with a familiar sequence.
Map the workflow.
Interview stakeholders.
Identify manual work.
Evaluate vendors.
Build a business case.
Recommend a roadmap.
Gemini is adding a different input:
what payers actually authorized, denied, overturned and paid.
The company says its advisory work draws from de-identified adjudicated claims and prior-authorization data across proton therapy, radiation therapy, radiopharmaceutical therapy, medical oncology drug reimbursement, MRI, PET imaging, urology and orthopedics.
The data follows cases through the full reimbursement lifecycle:
The real differentiator is not “AI advisory.” It is operator-owned reimbursement data. Gemini can advise on the same high-cost specialty workflows it runs every day, using the outcomes from those workflows as evidence.
This is an important shift in how to think about RCM services businesses.
Historically, a billing company was valued primarily for:
But every claim a specialized RCM firm touches also creates structured information about payer behavior.
What required prior authorization?
What documentation was accepted?
Which payer criteria mattered?
Which denials were overturned?
How long did reimbursement take?
What was actually paid?
At sufficient scale, the service operation becomes a proprietary data-collection system.
That can create an asset with value beyond the original billing contract.
Gemini's focus is narrow by design.
Radiation oncology.
Medical oncology.
Radiology.
Urology.
Orthopedics.
These are not low-dollar, highly standardized workflows.
A single authorization decision can determine whether a six-figure therapy moves forward.
Payer rules can vary materially by modality, drug, site of care, clinical indication, contract and geography.
New therapies may have limited payer precedent.
That makes generic industry averages less useful.
A provider deciding whether to launch a radiopharmaceutical therapy program, add a PET scanner or expand medical oncology does not need a national average.
It needs to know:
What are payers actually doing for organizations like mine?
Gemini is also making an important commercial choice.
Organizations can buy the advisory work without outsourcing the billing department.
That opens the addressable market.
A hospital may not want to replace its internal RCM team.
But it may still need help deciding:
Gemini can advise on those questions while the provider keeps its existing operating structure.
If the client later wants execution support, Gemini already operates the same specialty cycles.
That creates a natural path from advisory into managed services without requiring it.
Operate the revenue cycle → collect real reimbursement outcomes → improve the advisory model → help clients redesign workflows → optionally operate more cycles → collect more outcomes.
One of the more useful parts of Gemini's positioning is that it does not start with the model.
It starts with the reimbursement workflow.
That is how RCM AI should be evaluated.
For example:
A prior-authorization automation project is valuable only if the targeted workflow creates enough avoidable delay, labor, denial risk or lost volume to justify automation.
An oncology drug reimbursement tool is valuable only if it materially improves payment accuracy, speed or recovery.
An MRI medical-necessity check is valuable only if it reduces real denial or rework risk.
Gemini says its engagements model expected impact using the client's payer mix, contracted rates, volumes and real adjudication patterns.
That is a much more mature approach than attaching an AI use case to a generic industry benchmark.
Claims data alone tells you what happened financially.
Prior-auth data tells you what happened before the claim even existed.
Gemini's dataset includes the submitted request, payer criteria, denial reason, peer-to-peer review, appeal, overturn and final outcome.
That can reveal:
For high-cost therapies, that information has clinical, operational and financial value at the same time.
The advisory use case is only one application of the data.
The same dataset can potentially support:
Gemini already markets feasibility work for radiopharmaceutical therapy programs, PET and advanced imaging, urology expansion and de novo centers.
That is a meaningful expansion of the traditional RCM role.
The billing company is no longer only asking:
How do we get this claim paid?
It is also asking:
Should this clinical program exist, and what will the reimbursement environment look like if it does?
Gemini has been explicit that it sees itself as a service company first and a technology company second.
That is strategically notable in a market increasingly filled with “autonomous” claims.
High-cost specialty reimbursement contains enough nuance that pure automation can create material risk.
Complex therapies can involve:
Gemini's model is to use AI where it improves throughput and decision quality while keeping clinicians and reimbursement specialists in the loop.
That hybrid model may be particularly durable in specialties where the dollar value and clinical complexity of each decision are high.
There is a broader implication here.
Traditional consulting firms have expertise.
Software companies have technology.
RCM operators have daily workflow exposure and adjudicated outcomes.
AI makes the operator's position more valuable because those real-world outcomes can become training data, benchmarks and decision support.
The advisory market may therefore shift toward firms that can say:
We do not just know the theory. We operate the workflow, see the payer behavior and can show you the resulting economics.
That is a very different consulting product.
Gemini's new advisory practice is interesting because it shows how an RCM services company can move up the value chain without abandoning operations.
The service business produces data.
The data improves insight.
The insight becomes advisory.
The advisory can guide AI implementation.
And the implementation can create more operating data.
That is a much stronger flywheel than selling labor alone.
The defensible asset may not be the AI model. It may be the adjudicated reimbursement history underneath it.
For high-cost specialty RCM, that history is especially valuable because every payer decision contains information about access, denial risk, reimbursement and the economics of care delivery.
The next generation of RCM advisory firms may be built by operators who turn the claims they already touch into intelligence no outside consultant can easily replicate.
Daily coverage of payer policy, denials, deals, AI, and RCM strategy — written for people who live in revenue cycle.
Sources: Gemini Health Partners via PR Newswire / Yahoo Finance · Gemini claims-data methodology · Gemini advisory practice · RevCycleAI analysis · October 8, 2026