Multi-tenancy is an architectural pattern in which a single deployed instance of a platform serves multiple independent organisations — called tenants — while keeping each tenant's data, configuration, and user base strictly isolated from all others. In learning video platforms, multi-tenancy is essential for SaaS (Software as a Service) business models and for MOOC platforms that host courses from many institutions: rather than deploying a separate application instance per client, one shared infrastructure serves all tenants. That shared infrastructure uses row-level or schema-level database isolation, per-tenant CDN path prefixes for video assets, and tenant-scoped xAPI LRS partitioning to enforce boundaries. The economic benefit is infrastructure efficiency — shared compute, shared CDN, shared transcoding pipeline — but the engineering cost is that every feature must be designed with tenant isolation in mind from the start: a data leak between tenants is a critical security failure. Common implementation approaches range from a shared database with a tenant_id column on every table (simpler but higher cross-contamination risk) to separate database schemas per tenant (stronger isolation, higher operational overhead) to fully separate database instances (maximum isolation, scales poorly beyond a few hundred tenants). For learning video specifically, per-tenant branding (logo, colour scheme, custom domain), per-tenant course catalogues, and per-tenant analytics dashboards must all work correctly without any data crossing tenant boundaries. Multi-tenancy complexity is one of the key drivers of build-vs-buy decisions for learning platforms.
