Photo by Carlos Gonzalez on Unsplash
A January 2026 CMS enforcement memo, surfaced through Google News and analyzed in depth by TechBullion, confirmed what most Medicare Advantage administrators already suspected: RADV audits are now running quarterly across every eligible contract. That single policy shift transformed risk adjustment software from a revenue optimization lever into mission-critical compliance infrastructure — and sent health plans scrambling to evaluate whether their current platforms can withstand the new scrutiny.
The Compliance Earthquake Reshaping This Market
$556 million. That figure — the DOJ settlement secured from Kaiser Permanente affiliates on January 14, 2026 — is the largest Medicare Advantage risk adjustment fraud settlement in history, eclipsing Cigna's $172 million recovery in 2023 and Independent Health's $100 million settlement in 2024. Whistleblower Dr. James M. Taylor received $95 million of the recovery for exposing alleged unsupported diagnosis codes across Kaiser's affiliated plans. The message to every compliance team: this category of enforcement is not cooling off.
CMS moved in parallel on audit infrastructure. As of July 4, 2026, the agency expanded its certified medical coder pool from approximately 40 to roughly 2,000 — a 50x staffing increase — and announced plans to deploy AI-enabled audit tools while maintaining that human analysts must make all final RADV determinations. Per the quarterly cadence confirmed in that January 2026 memo, all 550+ RADV-eligible Medicare Advantage contracts are now subject to three-month audit cycles, with PY2020 audits having launched in February 2026. CMS also restored a five-month medical record submission window in 2026, giving plans additional preparation time amid the accelerated schedule — a meaningful operational relief valve, but not a structural fix.
Adding to the recalibration burden: 2026 marked the full replacement of CMS-HCC Model V24 with V28, the first major model overhaul since 2014. Valid ICD-10 codes dropped from 9,797 to 7,770; HCC categories expanded from 86 to 115; the projected average risk score reduction stands at 3.12%. Every platform reviewed below had to retool for V28. How cleanly a vendor executed that transition is a reasonable proxy for how it will handle the next one.
The Real Job You're Hiring This Software to Do
Frame it precisely: you are not hiring risk adjustment software to capture more HCC codes. You are hiring it to identify legitimate diagnosis gaps, document them with clinically defensible evidence, and survive a quarterly RADV audit without restating revenue. Any platform still pitching raw score maximization is selling a 2020 product into a 2026 compliance environment.
The secondary job is operational throughput. Chart review has historically been a bottleneck — manual, expensive, and inconsistent at scale. As a RAAPID compliance analyst stated in early 2026: "Risk adjustment is no longer a revenue lever. It is a compliance and cost-control function." The workflow automation tools winning on this axis are reducing per-chart labor while building the audit trail that quarterly reviews demand. That dual requirement — speed and defensibility — is what separates the current generation of AI-native platforms from their predecessors.
Arcadia's vendor analysis surfaces a structural observation worth anchoring the procurement conversation around: "Accurate risk adjustment requires the collaboration of both providers and payers, who share responsibility for accuracy and compliance. Rather than treating risk adjustment as a siloed function, organizations must view it as an ecosystem challenge." A point solution that automates coding but misses the provider-payer handoff is solving half the problem.
Photo by Vitaly Gariev on Unsplash
How the Leading Platforms Stack Up
TechBullion's July 2026 review examined seven platforms across AI capability, compliance posture, and pricing. Combined with Arcadia's interoperability framing, here is the editorial read across sources:
RAAPID — TechBullion's top-ranked platform. Uses what the vendor calls Neuro-Symbolic AI — neural pattern recognition combined with symbolic clinical reasoning — to automate MEAT (Monitor, Evaluate, Assess, Treat) evidence validation. Published figures: 92% AI accuracy at the suggestion stage, 98% final accuracy on HCC code suggestions, chart review time under 8 minutes, 5x coder productivity improvement, average revenue capture of $2,500 per member, and a guaranteed 10:1 ROI. Security certifications include HITRUST r2 and SOC 2. Pricing is quote-based. The 10:1 ROI guarantee is aggressive — worth pressure-testing against your own member volume and current coder cost per chart before signing.
Innovaccer — Strong on interoperability across multi-vendor EHR (electronic health record) environments. Organizations using Innovaccer's AI assistant have reported, per available case data, a 70% improvement in coding accuracy and a gross benefit of $27 million. The platform connects risk adjustment into a broader population health stack — a natural fit for health systems that need coding gaps identified alongside care management workflows, not as a standalone function.
Arcadia — Positions itself as an ecosystem connectivity layer between providers and payers, addressing the collaboration gap that point solutions leave unresolved. Less optimized for raw coding throughput; better suited when the documented failure mode is documentation handoff between clinical and administrative teams, not coding accuracy itself.
Episource — Long-established specialist in retrospective and prospective HCC coding services, increasingly AI-augmented. Chart retrieval at scale remains a documented operational bottleneck for large MA plans; Episource's core differentiation is here.
Edifecs — Enterprise-grade platform with strong V28 model compliance and a consistent regulatory update cadence. Commonly deployed across Blue Cross/Blue Shield affiliates and large regional payers. Less differentiated on AI-native features; stronger on process integration within existing payer IT stacks.
nThrive — Revenue cycle management platform with risk adjustment capabilities layered in. Better suited to provider organizations than payers; the Medicare Advantage-specific feature depth is shallower than dedicated platforms.
Cognizant TriZetto — Mature platform with deep payer penetration and a solid track record on V28 compliance updates. The trade-off is AI differentiation: TriZetto wins on process stability and enterprise IT integration, not on autonomous coding intelligence.
Chart: CMS expanded its RADV-certified medical coder workforce approximately 50-fold in 2026 — a structural escalation in audit enforcement capacity, not a temporary surge.
As of July 4, 2026, industry data places the risk adjustment software market at approximately $2.69 billion, projected to grow to $5.2 billion by 2035 at a 6.9% CAGR — with an alternate model projecting $6 billion by 2033 at a 15% CAGR. The spread between projections reflects genuine uncertainty about how fast AI replaces manual coding workflows at scale. My read: the quarterly audit cadence is the forcing function the more aggressive projection needs to prove out.
Before You Commit: The Switching Cost Reality
Risk adjustment implementations are not pilots. Basic software licensing starts at $120 for limited functionality and $800 for advanced features. Enterprise deployments can reach $400,000 in upfront costs before integration, training, or data migration. Unlike many productivity software categories where a failed implementation costs a quarter of productivity, a failed risk adjustment migration during an active RADV audit cycle can create direct legal exposure.
The real lock-in is documentation history: HCC coding decisions, audit trails, and clinical evidence records from prior years. These do not port cleanly between platforms. Attempting a mid-year switch while quarterly audits are running is a risk most compliance teams will not accept. Build at least one full contract cycle — typically two years — into your evaluation timeline before treating platform migration as a realistic option.
The team-size cliff is real here. Platforms like Innovaccer and Arcadia are architected for enterprise payer and health system scale. A regional MA plan with under 50,000 members may find that focused solutions — Episource's chart retrieval services, for example — deliver stronger ROI per dollar than a platform license that assumes data infrastructure you do not yet have.
Industry technology assessments note consistently that AI-powered risk adjustment features can automate time-consuming tasks and identify subtle diagnostic patterns that human analysts miss. But CMS requires human-certified final determinations for all RADV audits — making the "fully automated" vendor pitch a red flag rather than a selling point. The right architecture, as the broader enterprise agentic AI adoption pattern confirms, is AI handling the pre-work while a qualified human retains sign-off authority. CMS has effectively mandated this model for risk adjustment specifically.
Which Platform Fits Your Situation
Adopt RAAPID if your primary failure mode is HCC coding throughput and accuracy, you operate a meaningful-scale MA plan, and you need a defensible audit trail for quarterly RADV review. Run the 10:1 ROI guarantee against your own coder cost per chart and gap rate before committing.
Choose Innovaccer if you operate across multiple EHR environments, need risk adjustment embedded in a population health platform, and the $27 million gross benefit benchmark is directionally relevant to your scale.
Consider Arcadia if the documented failure mode is provider-payer collaboration — where diagnosis codes are clinically valid but poorly documented at the handoff between clinical operations and coding teams.
Episource, Edifecs, nThrive, and Cognizant TriZetto are worth evaluating if you are already inside their ecosystems, prioritize V28 compliance stability over AI differentiation, or operate on the provider side where payer-grade platforms are functionally oversized for your workflows.
In my analysis, the Kaiser Permanente settlement rewrote the central procurement question for this entire category. Plans are no longer asking which platform captures the most codes — they are asking which platform leaves the smallest audit surface for a DOJ investigator. That is a fundamentally different RFP, and the vendors that have retooled their pitch around documentation defensibility rather than revenue maximization are the ones worth shortlisting right now.
Frequently Asked Questions
What is risk adjustment software in healthcare, and how does it differ from standard medical coding tools?
Risk adjustment software specifically identifies and documents Hierarchical Condition Categories (HCCs) — the diagnosis codes CMS uses to calculate per-member payments to Medicare Advantage plans. Standard coding tools handle claims billing; risk adjustment tools focus on the clinical documentation that justifies a risk score, which directly determines plan revenue and audit exposure. The distinction matters because CMS's RADV audits scrutinize risk-coded diagnoses specifically, not claims coding broadly.
Is risk adjustment software worth the investment for smaller Medicare Advantage health plans?
It depends on member volume and current coding gap rates. Enterprise platforms with $400,000 upfront costs make economic sense at large scale; basic software starting at $120 to $800 per user may be more appropriate for regional plans. Published ROI figures — like Innovaccer's reported $27 million gross benefit — reflect enterprise-scale deployments. A realistic ROI model requires your own gap rate and member count, not vendor benchmarks. With quarterly RADV audits now standard, the cost of under-investing in documentation accuracy has increased significantly since 2023.
Why did the CMS V28 model transition matter so much for risk adjustment coding in 2026?
V28 fully replaced V24 in 2026 as the first major HCC model overhaul since 2014. Valid ICD-10 codes dropped from 9,797 to 7,770, and HCC categories expanded from 86 to 115. The projected average risk score reduction is 3.12%, meaning plans that did not recalibrate their coding workflows saw payment reductions with no change in their actual patient population. Every risk adjustment platform had to update code mapping, clinical logic, and audit documentation templates. How quickly and cleanly a vendor executed that transition is a reasonable proxy for its regulatory responsiveness going forward.
Disclaimer: This article is original editorial commentary based on publicly reported facts and is intended for informational purposes only. Tool features, pricing, and regulatory requirements change frequently — verify current details directly with vendors and CMS. No commercial relationships exist between the author and any platforms mentioned. Research based on publicly available sources current as of July 4, 2026.