Have you ever wondered how Meta moderates billions of posts, ads, comments, photos, and videos shared every day? How technology empowers our reviewers so they can remove election misinformation, hate speech, or health hoaxes at scale? Media Match Service (MMS) is the system at the heart of automatic moderation and our first line of defense. MMS is one of the 32 Meta Shared Systems, with ~70 internal teams using it to automatically handle harmful or illegal content. As a result, MMS accounts for >30% of making Meta products safe and inclusive by finding accurately and quickly all instances and derivatives of a given piece of content (photo, video, audio, text, post, ad) across all Meta apps. See MMS Product Overview and MMS Technical Primer for more details. MMS is built by the Similarity Detection (SD) org in Integrity (XI) using Hack, C++, and Python. SD is based in London, UK, and spans 3 teams with 29 engineers (E4 to E6) and 9 XFN (PM, TPM, SPM, DS, DE, PD). The SD Match Quality (SDMQ) team consists of 7 MLEs (E4-6) and has partial product and data science support. The team owns the algorithms that power MMS. Specifically, we develop and optimise uni- and multi-modal similarity classifiers, embeddings, and indexing algorithms for performance, scale, and wide applicability (see 3-yr MMS Product Strategy.) In our quest, we collaborate extensively with partner teams of Meta scientists and applied researchers (see Similarity Research Group).