7A20--AI Medical Coding (VA-25-00075960) (VA-25-00104841)

Other / Unclassified

This document serves as an update regarding the previously posted Request for Information (RFI) with Solicitation Number 36C10B26Q0028, concerning AI Medical Coding.

After reviewing industry responses and further assessing the Department of Veterans Affairs' (VA) requirements, the VA has decided not to proceed with further action on AI Medical Coding at this time. Bidders are advised to monitor SAM.gov for future announcements. The VA expresses gratitude to the companies that submitted responses and information for the RFI.

Request for Information (RFI)

This RFI is for planning purposes to gather information regarding AI Medical Coding solutions. It is not a solicitation and does not obligate the government to acquire any products or services. Responses will be used for market research only.

Vendors are requested to provide information on their capabilities to meet the requirements outlined in the draft Performance Work Statement (PWS). Specific areas to address include:

• Integration with ICD-10 and CPT coding databases.

• How their AI ensures coding accuracy and consistency across multiple clinical encounter types.

• Methods for collecting user feedback during beta testing or post-deployment.

• A Rough Order of Magnitude (ROM) for the effort and the methodology used.

• Corporate experience and expertise with specific examples/references (agency, POC, dollar value, contract number).

• Whether the draft PWS provides sufficient detail for software development and production operations support.

• Technical comments/recommendations on the draft PWS if insufficient detail is provided.

• Page Limit: 10 pages. No marketing materials or excess information will be reviewed.

• Required Initial Information: Company Name, Address, POC, Phone, Fax, Email, Business Size/Status, SBA VETCert verification (for VOSB/SDVOSBs), NAICS code(s), Socioeconomic data, DUNS Number, Existing Contractual Vehicles.

• Format: Responses must address capabilities specific to the draft PWS.

• Questions: Technical questions must be submitted in a separate document titled "AI MEDICAL CODING_RFI_Questions".

• Marking: Mark responses as "Proprietary Information" if business sensitive.

• Email Size Limit: 5 MB.

• Due Date: No later than 1:00 PM EST, January 09, 2026.

• Submission Method: Via email to Joshua Fitzmaurice (Joshua.Fitzmaurice@va.gov) and Mina Awad (Mina.Awad@va.gov).

• Subject Line: "AI MEDICAL CODING"

• NAICS Code: 541519 (Size Standard: 150 Employees).

• Vendors are requested to provide Company Business Size and Status, and for VOSB and SDVOSBs, proof of verification in SBA VETCert.

• A draft PWS titled "AI MEDICAL CODING_DRAFT_PWS" is attached.

• The government will not pay for information submitted.

• Generic capability statements will not be accepted.

Request for Information (RFI)

This RFI is for planning purposes to gather information regarding AI Medical Coding solutions. The Department of Veterans Affairs (VA) is seeking information on vendors' capabilities to integrate with ICD-10 and CPT coding databases, ensure AI coding accuracy, collect user feedback, and provide Rough Order of Magnitude (ROM) estimates. This RFI is not a solicitation and does not obligate the government to award a contract.

Vendors are asked to provide:

• Company Name, Address, POC, Phone, Fax, Email

• Company Business Size and Status (including VOSB/SDVOSB verification if applicable)

• NAICS Code(s), Socioeconomic Data, DUNS Number, Existing Contractual Vehicles

• Summary of capability to meet draft PWS requirements, specifically addressing:

  • Integration with ICD-10 and CPT coding databases.
  • AI accuracy and consistency across clinical encounter types.
  • User feedback collection methods.

• Rough Order of Magnitude (ROM) and methodology.

• Corporate experience/expertise with specific examples/references (agency, POC, dollar value, contract number).

• Assessment of the draft PWS sufficiency for software development and production operations support.

• Technical comments/recommendations on the draft PWS if deemed insufficient.

• NAICS Code: 541519 (Size Standard: 150 Employees)

• Product Service Code: 7A20

• Responses are due by December 22, 2025, 1:00 PM EST.

• Page limitation: 10 pages. No marketing materials allowed.

• Responses must address capabilities specific to the draft PWS.

• Submit responses via email to Joshua Fitzmaurice (Joshua.Fitzmaurice@va.gov) and Mina Awad (Mina.Awad@va.gov) with "AI MEDICAL CODING" in the subject line.

• Email file size limit: 5 MB.

• Proprietary information should be marked.

• Technical questions must be submitted in a separate Word document titled "AI MEDICAL CODING_RFI_Questions".

• Not specified in this RFI, but vendors should provide their Company Business Size and Status, including VOSB and SDVOSB verification if applicable.

• Attached documents: AI MEDICAL CODING_DRAFT_PWS and VA-25-00075960 - Performance Work Statement Template - AI Medical Coding_RFI.

• The government will not pay for information submitted.

• Generic capability statements will not be accepted.

Statement of Work (SOW) / Performance Work Statement (PWS)

The Department of Veterans Affairs (VA) aims to leverage Artificial Intelligence (AI) to enhance clinical documentation and coding accuracy, aligning with federal guidance on trustworthy AI adoption.

This Performance Work Statement (PWS) outlines the requirements for designing, developing, deploying, and supporting a real-time, browser-based AI chatbot. This chatbot will assist in improving the accuracy of coding for CPT, ICD-10, CC/MCC, and DRG-like groupings. Key functionalities include:

• Analyzing clinical notes (up to 2000 words) with a large language model (LLM).

• Highlighting key clinical terms and recommending terminology improvements.

• Providing rationale for suggestions and supporting structured note creation.

• Enabling version control for edits.

• Displaying assigned and suggested codes.

• Generating lists of current and additional recommended codes.

• Supporting VA-specific terminology, reimbursement logic, and multiple clinical disciplines.

• Fine-tuning the LLM for VA terminology and updating it regularly.

• Implementing continuous monitoring for performance and bias.

• Supporting human-in-the-loop review and feedback workflows.

• Tracking utilization and edits for auditability.

• Architecting for future integration with the Computerized Patient Record System (CPRS).

• Developing training modules and documentation.

• Rapid response times of five seconds or less for standard clinical notes.

• Continuous monitoring for performance degradation or bias.

• Tracking utilization, text edits, and coding suggestions.

• Deliverables include a Monthly Progress Report, Systems Report, Project Plan, Data Acquisition Report, Trained AI Model, Beta Testing Implementation Plan, and Discrepancy Analysis Report.

• Performance Period: 12 months from the date of award.

• Place of Performance: Primarily at the Contractor's site; no anticipated travel.

• Compliance with numerous federal laws, VA directives, and NIST publications regarding information security, privacy (HIPAA), data management, and IT standards (e.g., FIPS, FISMA, Zero Trust, FICAM, IPv6, TIC).

• Adherence to the VA Technical Reference Model (TRM).

• Personal Identity Verification (PIV) card enablement for user authentication.

• Support for VA's Veteran-Focused Integration Process (VIP) and Product Line Management (PLM).

• All personnel must complete mandatory VA training and sign a VA Systems Access Agreement.

• Business Associate Agreement (BAA) required due to potential access to Protected Health Information (PHI).