AI/ML Model Development for Autonomous Medical Coding

The contract seeks the development and training of advanced NLP-based machine learning models capable of autonomously assigning ICD-10, CPT, and other standardized medical codes directly from clinical documentation with a minimum accuracy threshold of 95 percent. The models must not only achieve high precision in code assignment but also provide full explainability, ensuring that every coded output is traceable, interpretable, and audit-ready to support clinical, billing, and regulatory compliance needs. The technology must be robust enough to handle diverse clinical documentation styles, including physician notes, discharge summaries, and diagnostic reports, across a broad spectrum of medical specialties. This subcontract, issued by the Defense Health Agency under the Department of Defense, falls under NAICS code 541512 for computer systems design services and is open for responses until August 19, 2026. The performance location is unspecified beyond a generic placeholder, indicating potential flexibility in execution but requiring close alignment with DHA’s operational and security requirements. The solicitation emphasizes both technical excellence and operational reliability, targeting contractors with proven experience in healthcare AI, natural language processing, and medical coding standards. While no set-aside provisions are specified, the project is critical to modernizing military healthcare administrative workflows and improving efficiency in clinical data processing across the defense medical system.