Trellis AI

Overview

Trellis AI is a Stanford AI Lab spinout that runs AI agents to automate prior authorizations, intake, appeals, benefits verification, and eligibility checks for specialty providers, health systems, and pharma companies across all 50 states, with 98%+ accuracy reported in PA form completion. The agent is trained on millions of clinical data points and uses computer-use agents to operate payer portals and EHR systems where API integrations are unavailable.

Trellis AI is best read as an AI-native PA-and-appeals automation engine for specialty therapies, not as a payer UM tool, an outsourced hub, a benefits-verification operations vendor, or a specialty pharmacy. The relevant manufacturer wedge is end-to-end automation of pre-service administrative work, including specialty pharmacy onboarding for high-cost therapies.

Reimbursement and PA Capability Model

The framework below standardizes how Rx Almanac evaluates reimbursement-prior-auth capabilities, so buyers can compare vendors like-for-like while the readout column stays vendor-specific. For this table, Trellis AI is evaluated as an AI-native PA-and-appeals automation engine for specialty therapies, not as a payer UM tool, an outsourced hub, a benefits-verification operations vendor, or a specialty pharmacy.

Capability	Buyer should compare	Trellis AI readout
Electronic PA initiation and payer connectivity	ePA submission, payer integrations, plan-specific requirements, real-time decisioning, and transaction reach.	Core PA casework. PA form completion reports 98%+ accuracy; computer-use agents operate payer portals where API connectivity is absent.
Benefits verification and coverage discovery	Eligibility, pharmacy/medical benefit checks, policy criteria, coverage route, and reimbursement path identification.	Core coverage workflow. Real-time eligibility and coverage checks run alongside PA workflows.
Clinical documentation and appeals support	Document collection, criteria matching, appeal packets, peer-to-peer prep, and denial management.	Core PA casework. Agent ingests unstructured documents and assembles PA forms, justifications, and appeals.
Provider workflow integration	EHR, portal, fax, call, pharmacy, and practice-management workflows that reduce office burden.	Core network reach. Specialty-pharmacy onboarding for high-cost therapies and computer-use agents for portal/EHR work where direct integration is unavailable.
Automation, AI, and queue prioritization	AI agents, rules engines, form completion, status retrieval, queue triage, and exception handling.	Core automation workflow. End-to-end automation across intake, PA submission, and appeals.
Reporting, status visibility, and policy intelligence	Case status, payer trend reporting, denial reasons, turnaround metrics, and policy-change intelligence.	Confirm reporting layer in the RFP. Confirm dashboards, status fields, transcripts, and audit trails available to manufacturer and hub teams.

Buyer Fit

• Evaluation trigger: Pilot Trellis AI when high-cost specialty PA volume, complex documentation, and appeals are the bottleneck and a computer-use agent can act where electronic integrations are absent.
• Best-fit buyers: Pharma manufacturers, providers, and health systems running specialty programs across all 50 states; reported deployment with Fortune 500 healthcare customers.
• Less ideal fit: Programs whose primary need is voice-based payer follow-up, outsourced hub case management, fulfillment, or dispensing.
• Commercial fit: Subscription pricing; scope appeal types, payer reach, clinical-review boundaries, and SLA expectations in the SOW.
• Scope diligence: Confirm payer plan and EHR reach, provider burden, clinical-review boundaries, appeals depth, and whether the engagement is provider-side, hub-side, or manufacturer-side.

Differentiators

• Stanford AI Lab pedigree: Agent trained on millions of clinical data points; academic-grade ML applied to PA automation.
• 98%+ PA accuracy: Reported accuracy in PA form completion and justifications is the headline diligence anchor.
• Computer-use agents: End-to-end automation including portal/EHR action where direct API integration is unavailable — extends reach beyond API-only PA vendors.
• Specialty-pharmacy onboarding: Workflow handles the high-cost therapy onboarding path, not just front-end PA submission.
• 50-state coverage at scale: Processes billions in therapies annually with 10x revenue growth reported late-2025 to early-2026.

RFP Questions

• Which payer plans and EHR systems are live for the product’s expected patient mix?
• Which responsibilities do computer-use agents operate versus direct API integrations, and what governance applies to portal automation?
• What clinician review, transcript QA, and audit trails govern AI-generated PA submissions and appeals?
• What status, denial-reason, and turnaround data flows back to hub, specialty pharmacy, and manufacturer teams?
• How does the platform handle peer-to-peer prep, reauthorizations, and specialty-pharmacy handoff?
• Which Fortune 500 references and specialty therapy areas are available for comparable workflows?

Recent Activity

• 2025-2026 - 10x revenue growth reported; ~25 employees as of early 2026.
• 2025 - Deployed with Fortune 500 healthcare customers (names undisclosed); processing billions in therapies annually across 50 states.
• 2024-Q1 - Y Combinator Winter 2024 (W24) cohort; raised $500K accelerator round with minority backing from Carya Venture Partners.
• 2023 - Company founded by Mac Klinkachorn (CEO) and Jacky Lin in San Francisco.