Small and mid-sized Class II medical device firms face a structural operating challenge. Regulatory evidence, supplier records, quality files, and launch decisions must be coordinated across markets, but compliance capacity often grows more slowly than operating complexity. This paper develops a dual-loop AI operating discipline for Class II diabetes device manufacturing. The model links a Regulatory Readiness Loop with a Supply Adaptation Loop through human review gates, controlled evidence retrieval, supplier early-warning scoring, and audit-ready quality-system records. The study uses a six-month case pilot at New Horizon Biotech, where a five-FTE AI Business Unit supported FDA 510(k) and EU MDR workflows without replacing expert judgment or formal quality controls. Results show that regulatory drafting cycle time decreased from 6.5 to 4.4 weeks, defect density fell from 2.8 to 0.9 errors per page, review iterations decreased from 4.2 to 2.5 cycles, and evidence retrieval time decreased from 12.5 to 1.5 hours. Supply outcomes also improved. Lead-time recovery decreased from 8.5 to 5.2 weeks, sole-source exposure declined from 42% to 34% of the bill of materials, and critical stock-outs decreased from two events to zero. The framework reached a median groundedness score of 0.94 by Month 6 and generated an estimated 82.4% six-month ROI. The paper contributes an operations management model for applying AI as a governed process-control layer in regulated manufacturing.

Keywords

Artificial Intelligence, Operations Management, Supply Chain Resilience, Quality Control, Medical Devices.