Data Multi-tenancy

Data multi-tenancy is an architecture in which one system serves many customers while each customer's data is kept isolated and separated from the rest.

In a SaaS tool a 'tenant' is usually a customer or a project; multi-tenancy means all these tenants run on shared infrastructure yet none sees another's data. The value of this architecture is economic efficiency, but its heart is a security guarantee: the isolation boundary must be leak-proof, or one customer's data spills into another's report. It has two perennial challenges: the 'noisy neighbor', when one tenant's heavy usage slows the others' performance; and limits and quotas, because each tenant needs its own usage ceiling. For an analyst, the practical point is that metrics always make sense within one tenant's boundary; comparing across tenants is only allowed via deliberate, leak-free aggregation.