BITComputer launches AI-powered insurance claims search service for hospitals

An AI agent service that consolidates scattered hospital insurance review and claims functions into a single search has been launched.

BITComputer, a medical information company, launched Medigent AI — an AI-based integrated search service for insurance review and claims — on Tuesday.

The service combines AI with Medigent, BITComputer's existing insurance review and claims platform, which 44 of the country's 47 tertiary hospitals, or 94 percent, have adopted. The service allows insurance review staff, who previously had to hunt through scattered review criteria, to complete their work through a single search, the company said.

A single query — whether in natural language or using an EDI code — retrieves benefit criteria, official notices and review cases all at once, the company added. Verifying relevant information and supporting documentation based on the EDI codes that medical institutions actually use for claims has traditionally been difficult, and searching for a specific fee item often returned loosely related notices, limiting efficiency. Medigent AI addresses both shortcomings.

"Medigent AI turns the entire process into a single question," BITComputer said. "Staff can enter the EDI code directly from the claims screen, or ask in natural language using a drug name, test name, treatment material name, or the kind of phrasing they use in day-to-day work."

The company added that the service "identifies the intent of the query and conducts an integrated search across related benefit criteria, official notices, review guidelines, administrative interpretations, deliberation cases, electronic review records, fee schedules and billing methods, presenting all supporting materials on a single screen. Source documents are displayed alongside their original text and citations, so review staff can cross-check AI responses against the primary sources."

BITComputer is rolling out the service starting with hospital-level medical institutions, with plans to expand further. The company said it will continue refining response quality using its proprietary medical-domain large language model, tailoring it to the terminology, context and workflow of Korea's insurance review environment.