A global academic medical system needed to protect proprietary clinical IP across Microsoft 365, but conventional pattern matching missed nuanced research content and labels were inconsistent. Lighthouse solved this by mapping IP categories to Purview labels, combining Sensitive Info Types, Exact Data Match, and trainable classifiers into a blended strategy, and piloting models tied to retention, legal hold, and eDiscovery with auto-labeling and user prompts. The result was consistent protection with policy-driven controls, high-confidence detection of custom IP, and a governed workflow that reduced false positives and accelerated eDiscovery.

Join for free to read