Health information exchanges play a critical role in aggregating and sharing clinical data across providers, payers, and care teams. Yet the quality of that data is often assumed rather than verified. While format-level validation has become standard practice, content-level validation remains largely absent across the industry. A diagnosis code may be syntactically valid but clinically implausible, or a lab result may be present but lack the units or reference ranges needed to be actionable. These gaps are often not identified with format checks and represent a significant negative impact on data quality. Manifest MedEx (MX), a nonprofit Health Information Exchange serving California and the first HIE in the state certified by the National Committee for Quality Assurance (NCQA) under the Data Aggregator Validation DAV) program, recognized this gap through its work with NCQA and auditors during the PSV process. This session will walk attendees through MX’s experience in designing a process that incorporates AI and staff to identify and close data quality gaps. Attendees will hear practical examples of the types of content errors the system has identified, how those findings are surfaced to data submitters, and what feedback loops have been established to drive improvement over time.