REQUEST FOR INFORMATION (RFI) - Intelligent Coding Assistance Tool

The Centers for Medicare & Medicaid Services is conducting market research through a Request for Information to assess vendor capability to procure an AI-powered Intelligent Coding Assistance Tool integrated with Salesforce that consolidates medical chart and record data for diagnosis abstraction, with responses due by August 20, 2026.

The tool will support the HHS-Operated Risk Adjustment Data Validation Program's Second Validation Audit phase and must be deployed within the CCIIO Customer Service Response Management System Salesforce instance. CMS intends to deploy an initial solution within six months of award and expects a developed product requiring limited customization.

Distinctive requirements include FedRAMP authorization (High, Medium, or Low) and FISMA compliance; strict conflict of interest standards prohibiting involvement with health insurance issuers, IVA entities, or TPAs; and support for data scraping from multiple sources in PDF, Excel, Word, JSON, and XML formats. The tool must restrict first-level review with senior reviewer confirmation required for error discovery under 45 CFR 153.630, support template design and customization, map diagnosis codes to HCC categories, measure error rates, validate medical records, and generate reports.

Primary objectives include reducing time for medical coders to reach accurate conclusions, reviewing medical charts faster and as accurately as human coders, reducing CMS hiring costs, improving efficiency through progressive use, and achieving demonstrable return on investment. Respondents should address technical feasibility (deployment phases, architecture, security certifications), operations and maintenance (adoption factors, training, one-to-five-year ROI projections including implementation and licensing costs), and acquisition strategy options. Responses are limited to ten pages total with 2,000–3,000 characters per question and are not required to address all questions. This RFI does not commit the Government to procurement.