Ada Project Manager (APM) formally defines a language describing a project which uses the Ada programming language. Its purpose is the management, beyond Ada's language specification, of a set of sourcecode and configuration options/specifications.

As an example, the F-22 has multiple microprocessors, but all of these are a part of a larger system. Where the Ada source-code may regard/model the system uniformly. That is to say, we may regard the entire system as a single program. We may also regard what portions are executed on the particular microcontrollers to be a single program as well, it is necessary to compile those portions of the source to execute on the microcontrollers as necessary.

In terms of the Ada programming language this is covered, briefly, by Annex E — the Distributed Systems Annex (DSA) — which refers to a program in terms of partitions and nodes. In such a model, a program at least one processing node (the portion[s] of the program which actually have a thread of control), and zero or more storage nodes (which could be a database-interface, for instance), which are configured/compiled via an external tool-chain so that in most cases there is no difference semantically between a single executible running on a multithreaded maching and multiple executables running on multiple processors.

The goal of APM is to address that configuration management, as well as the case of having multiple targets from the same source code. — While this is similar to a build management system, it is not concerned with finding files, retrieving them, and presenting them to the compiler (as a build system does), but with the managment of configurations for the compiler(s) to use.