Isolating fine-grained business functionalities byboundaries into entities called microservices is a core activityunderlying microservitization. We define microservitization asthe paradigm shift towards microservices. Determining theoptimal microservice boundaries (i.e. microservice granularity) is among the key microservitization design decisions thatinfluence the Quality of Service (QoS) of the microservice applicationat runtime. In this paper, we provide an architecturecentricapproach to model this decision problem. We build onambients — a modelling approach that can explicitly capturefunctional boundaries and their adaptation. We extend the aspect-oriented architectural meta-modelling approach of ambients—AMBIENT-PRISMA — with microservice ambients. A microservice ambient is a modelling concept that treatsmicroservice boundaries as an adaptable first-class entity. Weuse a hypothetical online movie subscription-based systemto capture a microservitization scenario using our aspectorientedmodelling approach. The results show the ability ofmicroservice ambients to express the functional boundary of amicroservice, the concerns of each boundary, the relationshipsacross boundaries and the adaptations of these boundaries. Additionally, we evaluate the expressiveness and effectivenessof microservice ambients using criteria from ArchitectureDescription Language (ADL) classification frameworkssince microservice ambients essentially support architecturedescription for microservices. The evaluation focuses on thefundamental modelling constructs of microservice ambientsand how they support microservitization properties such asutility-driven design, tool heterogeneity and decentralised governance. The evaluation highlights how microservice ambientssupport analysis, evolution and mobility/location awarenesswhich are significant to quality-driven microservice granularityadaptation. The evaluation is general and irrespective of theparticular application domain and the business competenciesin that domain.