Improving Speed and Accuracy in Benefit Verification at Genentech

Genetech

For Genentech, the benefit verification process was fragmented across clearinghouse APIs, payer portals, and plan documents. Manual lookups stretched cycle times to 3–5 days, introduced frequent eligibility errors, and placed heavy demand on support staff. To address this, we implemented an Electronic Benefit Verification (EBV) solution that combined real-time API checks with GenAI-powered parsing of payer PDFs, FAQs, and plan documents, while also tapping into proprietary formulary data repositories for up-to-date benefit and coverage rules. The system provided instant visibility into eligibility, copays, deductibles, and coverage limits, and proactively surfaced prior authorization requirements dramatically improving speed and accuracy.

Business Background

Genentech’s Benefit Verification process was fragmented across clearinghouse APIs, payer portals, and plan documents. Manual lookups increased cycle time, created hand‑offs, and introduced avoidable eligibility/coverage errors. Annual Reverification was a tedious process across their specialty disease segment with close to 800+ patients across various programs. Time to therapy for new enrollments were spanning up to 3-5 days and times over a week to successfully onboard a patient leading to patient dissatisfaction.

Our Solution

A GenAI‑augmented EBV capability: deterministic API checks enriched with LLM‑based plan document parsing and guard-railed RAG over payer PDFs/FAQs.

  • Instant eligibility & benefits: Pulls eligibility, copay, deductible, coinsurance, coverage limits-in real time-so hub teams don’t chase portals or wait on calls.
  • Gen-AI accuracy: Uses intelligent data extraction/validation to eliminate manual entry errors and reduce rework.
  • PA awareness: Surfaces prior-auth requirements early so teams can adjust plans and shorten time-to-therapy.
  • Human in the loop: Our offshore Gen AI enablers were also involved to collect necessary information for outliers

Business Outcomes