Hyland IDP
by Hyland
SaaS / CloudOn-PremisesHybridHIPAA BAA
Overview
Hyland IDP is an AI-native document processing capability built on the Content Innovation Cloud platform. Combines LLM-based recognition engines (OCR, ICR, OMR), prompt-based low-code configuration, and agentic workflow automation for classification, extraction, validation, and enrichment. Purpose-positioned for healthcare payers on the CMS 2026 prior authorization decision timeline requirements (72 hr expedited, 7 day standard). Extends Hyland's established OnBase footprint in payer operations with documented integrations into TriZetto Facets/QNXT/CareAdvance and Epic via SMART on FHIR.
Workflow Coverage
Click any row to expand capabilities. Hover rows for form examples.
| Workflow | Maturity | Accuracy |
|---|---|---|
▶Intelligent Mailroom | Available | Not published; validate on your document corpus |
▶Claims Intake & Validation | Preview | Not published; evaluate on your corpus |
▶Prior Authorization | Available | Not published |
EOB & Remittance Processing Commercial only | Unknown | — |
Risk Adjustment / HCC Coding | Unknown | — |
▶Appeals & Grievances | Available | — |
Provider Enrollment MAC only | Unknown | — |
Provider Credentialing Commercial only | Unknown | — |
▶Member Enrollment Commercial only | Available | Valeris: 40% time savings; processing time 25-30 min → ~5 min per document packet |
▶Medical Records & Clinical Documentation | Preview | — |
Integrations
Payer Core Systems
- TriZetto Facets/QNXT/CareAdvance — documented point-and-click integration; surfaces documents in payer core UI
- Epic via SMART on FHIR — OnBase connector; tested across 450+ test cases for 64 workflows
- OnBase — native content services and workflow layer (system of record for extracted documents)
- Salesforce — member records linkage (connector pattern details not published)
Known Limitations
- ✗No standardized payer-form accuracy benchmarks published for CMS-1500 or UB-04
- ✗Clearinghouse integrations not documented in public sources
- ✗HealthEdge payer core integration not documented (only TriZetto confirmed)
- ✗Model internals and training datasets not disclosed publicly
- ✗Claims 'no training data required' is a vendor marketing statement; technical basis not published
- ✗Legacy Brainware (older capture product) may be present in some existing Hyland deployments