Medallion

AI agents do the credentialing work, but the company is unusually candid that they do not do all of it. It describes purpose built AI agents for credentialing overseen by industry experts, paired with a proprietary payer clearinghouse to deliver real time results. It then states plainly that AI is not perfect yet, which is why credentialing expertise is embedded directly into every workflow and specialists step in when automation encounters an exception. That is a hybrid model honestly labeled rather than an autonomy claim, and a buyer should size it as automation plus a services layer.

The oversight design is explicit, structural, and stated without spin. Experts review critical stages across credentialing and enrollment before work moves forward to ensure accuracy, completeness, and readiness for regulatory review, and from primary source verifications through committee ready packets that review is built into the process. The company also states its experts continuously train and configure the AI agents, which makes human review a feedback mechanism rather than only a gate. For work where an error means an improperly credentialed clinician treating patients, embedding review rather than sampling is the right architecture.

The company describes an AI agent workforce combined with a proprietary payer clearinghouse and direct integration with primary source databases, which explains where the data comes from and how results are produced at a functional level. No model provider, architecture, accuracy measurement, or exception rate is published, and exception rate is the figure that would tell a buyer how much of the work the AI actually completes unaided.

Operational rather than clinical, appropriately, and backed by an unusual commercial commitment. The company states it stands behind credentialing operations with performance backed guarantees, which transfers some delivery risk to the vendor and is rare in this index. Third party analysis cites credentialing time cut by up to 80 percent with committee ready files in days rather than weeks. The company reports saving more than 250,000 administrative hours across named customers including several large digital health and primary care organizations. Figures are vendor or third party reported without independent audit.

No published data governance framework was located. Worth noting the data here is provider rather than patient information, meaning licenses, sanctions, education, and malpractice history, which is sensitive personal data about clinicians but sits outside PHI, so the usual patient privacy framing applies only indirectly.

No HIPAA or BAA commitment was located, and the requirement is weaker here than for most vendors in this index because credentialing operates on provider credentials rather than patient records. A buyer should still confirm the posture where the platform touches any patient linked data.

No SOC 2, HITRUST, or ISO 27001 attestation was located, and no trust center was found. The company does reference NCQA certification for its credentials verification organization, which is an accreditation of credentialing process quality rather than an information security attestation, and the two should not be conflated.

No FDA pathway applies, but this vendor operates under a real and specific accreditation regime that functions as its equivalent. The company describes an NCQA certified credentials verification organization and states it generates NCQA compliant credentialing files in real time, with workflows aligned to both NCQA and Joint Commission standards for privileging. Those accreditations are what allow a health plan or health system to rely on delegated credentialing, so they are the credential that matters commercially, and the company holds them.

No formal governance framework was located, but the stated design of expert review at critical stages before work advances, plus specialists handling automation exceptions, is a functioning control on AI error even though it is not framed as governance. No bias analysis was located, and the relevant question is whether verification performance varies by provider type, state licensing board, or credential pathway, since uneven automation could slow credentialing for clinicians from less common backgrounds.

Integration is with credentialing infrastructure rather than the clinical record, which is the correct surface. The company integrates directly with primary sources including the federation of state medical boards, the national practitioner data bank, federal exclusion and entity databases, and the national provider registry to verify and monitor credentials automatically, and operates its own payer clearinghouse for enrollment submission. Continuous monitoring against those sources with real time alerts on credential and sanction changes is a materially harder integration than periodic manual checks. No EHR integration applies.

No hosting, tenancy, or data residency disclosure was located.

No published rates, but two disclosures give a buyer more than most vendors here provide. The company states it offers performance backed guarantees on credentialing operations, which is a commercial term rather than a marketing claim and shifts delivery risk. Third party review notes the absence of a strong self service option for organizations wanting to manage their own enrollments, which implies the model is oriented toward managed service rather than software licensing. Neither structure nor price is published.

Broad on every dimension that matters for this function. Provider type coverage spans all major categories including physicians, nurses, nurse practitioners, physician assistants, and licensed clinical social workers. Geographic coverage spans all United States states and territories for licensing. Functional coverage runs the full lifecycle: licensing, primary source verification, credentialing, payer enrollment for both commercial and public programs, delegated credentialing, privileging, and continuous monitoring. Buyer types span health systems, digital health companies, payers, and provider groups.