CombineHealth

CombineHealth fields an AI Workforce of seven named agents across the revenue cycle: Amy (autonomous medical coder), Mark (medical biller handling eligibility verification and claim validation), Adam (denial manager automating payer follow up), Rachel (appeals manager), Penny (policy reviewer), Taylor (data analyst feeding Amy's learning), and Jessica (medical scribe). Amy reads encounter notes from the EMR, assigns ICD-10, CPT, HCPCS, E/M, and modifier codes with line by line rationale citing the documentation and payer specific reasoning behind every decision, and proactively escalates roughly 15 percent of ambiguous charts for human review while teams choose their automation depth (CDI support, pre bill review, or autonomous). Its architecture uses retrieval augmented reasoning over payer policies, medical necessity criteria, and similar past encounters. An ED deployment automated about 85 percent of coding workflows at 98.5 percent E/M accuracy, and a published head to head matched expert coders across 1,000 ED charts.