Identifying collaborators in large codebases
This provides management with tools to observe organic team structures for better talent management, though it is incremental in applying existing embedding and clustering methods to codebase data.
The paper tackled the problem of identifying actual developer collaborations in large codebases by analyzing commit activity, programming languages, and code topics, and showed it could restore GitLab's engineering structure and reveal hidden collaborations using 117 open-source projects.
The way developers collaborate inside and particularly across teams often escapes management's attention, despite a formal organization with designated teams being defined. Observability of the actual, organically formed engineering structure provides decision makers invaluable additional tools to manage their talent pool. To identify existing inter and intra-team interactions - and suggest relevant opportunities for suitable collaborations - this paper studies contributors' commit activity, usage of programming languages, and code identifier topics by embedding and clustering them. We evaluate our findings collaborating with the GitLab organization, analyzing 117 of their open source projects. We show that we are able to restore their engineering organization in broad strokes, and also reveal hidden coding collaborations as well as justify in-house technical decisions.