In recent years, Web APIs have become a de facto standard for exchanging machine-readable data on the Web. Despite this success though, they often fail in making resource descriptions interoperable due to the fact that they rely on proprietary vocabularies that lack formal semantics. The Linked Data principles similarly seek the massive publication of data on the Web, yet with the specific goal of ensuring semantic interoperability. Given their complementary goals, it is commonly admitted that cross-fertilization could stem from the automatic combination of Linked Data and Web APIs. Towards this goal, in this paper we leverage the micro-service architectural principles to define a SPARQL Micro-Service architecture, aimed at querying Web APIs using SPARQL. A SPARQL micro-service is a lightweight SPARQL endpoint that provides access to a small, resource-centric, virtual graph. In this context, we argue that full SPARQL Query expressiveness can be supported efficiently without jeopardizing servers availability. Furthermore, we demonstrate how this architecture can be used to dynamically assign dereferenceable URIs to Web API resources that do not have URIs beforehand, thus literally ``bringing'' Web APIs into the Web of Data. We believe that the emergence of an ecosystem of SPARQL micro-services published by independent providers would enable Linked Data-based applications to easily glean pieces of data from a wealth of distributed, scalable and reliable services. We describe a working prototype implementation and we finally illustrate the use of SPARQL micro-services in the context of two real-life use cases related to the biodiversity domain, developed in collaboration with the French National Museum of Natural History.