Document Intelligence Analyst II at GHX in Hyderabad, Telangana, India

Job description

• GHX connects healthcare suppliers and hospitals across North America, processing over 2.2 million Purchase Order (PO) documents per year through its ADM (Automated Document Management) platform. The platform processes the majority of documents automatically; documents the system cannot process with sufficient confidence are routed to a human review queue.
• The Document Review Specialist verifies those flagged documents and corrects extraction errors at the field level. Every correction feeds directly into model retraining — you are not maintaining a status quo, you are actively improving the platform’s accuracy over time. This is a precision operations role, not conventional data entry.

Responsibilities

• Document Verification & Field Correction — 85%
• Verify documents like purchase orders, invoices, etc. flagged by the platform as requiring validation
• Compare extracted field values against the source document image and correct every discrepancy at the field level
• Verify and correct core fields, including vendor name, PO/Invoice number, line item descriptions, quantities, unit prices, delivery dates, and ship-to addresses
• Sustain throughput of ~30 documents per hour at a field accuracy rate of 99% or above
• Self-audit daily output before shift-end submission to catch and correct errors before they enter the system
• Flag ambiguous, damaged, or edge-case documents per the team escalation protocol; do not attempt to resolve documents outside defined parameters
• Process Documentation & Reporting — 10%
• Create and maintain Standard Operating Procedures (SOPs) for document review workflows, and update them as document types and field standards change
• Build and maintain process maps for end-to-end document review flows, including exception and escalation paths
• Perform basic analytics — collate production and quality data, and pull routine and ad-hoc reports on throughput, accuracy, and error trends as required
• Calibration & Continuous Improvement — 5%
• Participate in team calibration sessions led by the Supervisor to maintain consistent field interpretation standards
• Surface recurring error patterns to the Supervisor (e.g., consistent misread of a specific supplier’s PO template) through the team escalation channel
• Complete structured onboarding and participate in refresher sessions as document types and field standards evolve

Requirements

• Bachelor’s degree in any discipline
• 2–4 years of experience in a fast-paced document review/processing, data verification, quality review, or back-office operations — BPO, KPO, or shared services background is required; prior experience in a document review or data verification role is essential
• Excellent communication skills in English, including reading, writing, and comprehension — all documents are US-format business records written in English
• Ability to multi-task with minimum 35 WPM typing with high accuracy; comfort with keyboard-driven, browser-based workflows
• Working knowledge of business and procurement documents such as purchase orders, invoices, and order acknowledgements
• Demonstrated attention to detail — proven ability to detect numerical and textual discrepancies in structured documents
• Quick learner on new enterprise applications
• Prior exposure to procurement documents: purchase orders, invoices, order acknowledgements, or similar supply chain records
• Experience in AI-assisted document review, data annotation, or output quality assurance
• Familiarity with US healthcare supply chain or hospital procurement terminology (vendors, line items, ship-to addresses)
• Experience meeting and sustaining daily productivity quotas in a measured, SLA-driven operations environment
• Basic understanding of why field-level accuracy matters in downstream automated systems and suggest improvements

Skills

Required

• Attention to detail
• Quick learner
• Ability to multi-task
• 35 WPM typing
• Keyboard-driven, browser-based workflows
• Basic analytics
• Process mapping