Microservice Catalog Generation and Inference Based Selection of Microservices

Mechanisms are provided for indexing microservices for optimized querying based on microservice attributes. A plurality of application graph data structures are generated with nodes representing microservices and edges representing functionality of microservices. A data transformation is performed on the graphs to generate, for each node, a corresponding microservice document specifying microservice attributes of the corresponding microservice. A machine learning training operation is executed on an embedding computer model based on a plurality of the microservice documents to train the embedding computer model to learn a representation vector space for representing microservices as vector representations. The trained embedding computer model is executed on the microservice documents to generate corresponding vector representations and compile them into entries of a microservice index data structure which is used to process queries for microservices.