Invention and Innovation

Overmatch correction on a national identity platform

An overmatch is a set of linked records that may not belong to the same person. It can put one patient's record in front of a clinician treating someone else. Since October 2024, Saiteja invented an explode-and-rematch method for the enterprise patient-identity platform: break a bad link apart, rematch each record on its own evidence, and push the correction downstream in real time. A high-confidence match completes automatically. A low-confidence match goes to a person.

He also designed the real-time demographic events service and the enterprise event taxonomy so a demographic change has one meaning across the platform. The design went to the architecture review board and became the required enterprise standard.

• Manual reconciliation fell by about 40 percent
• Corrections that took hours now reach 14 lines of business in seconds
• About 150 transactions per second on the path; about 70 percent faster queries; about 40 times faster uploads
• More than ten teams build on the taxonomy

Platform context (not personal output): the organisation serves more than 185 million people. Black Book Research estimated that inaccurate patient identification costs more than 6.7 billion dollars a year.