Introducing DeepGraph

Start with a question, not a ranking

A report is most useful when you know what you want to understand. Instead of asking who has the highest total, try a question you can investigate:

• Who has worked on this project? Explore its contributors, then open the changes behind their activity.
• Where does a teammate contribute? Open their profile and follow their work across the projects in your report.
• What happened during a busy period? Choose a date range and inspect the commits that explain the change in activity.

Project history can point you toward someone with relevant experience. It cannot tell you everything they know, whether they are available to help, or who should own the work next. Use it to start a conversation.

Follow the overview down to the changes

Imagine a project shows a sharp increase in lines added during one week. That is a useful starting point, but it is not yet an explanation.

Open the contributors active in that period and review their changes. The increase might come from a new feature, generated files, or a large import. The same total can describe very different kinds of work.

DeepGraph connects activity trends, contributor profiles, and project insights, so you can move from a pattern to the evidence behind it.

Give each person a coherent history

People change email addresses, use different names on different projects, and contribute through more than one account. Their work can appear as several separate profiles.

DeepGraph lets you link profiles that belong to the same person. Check the details before combining them; similar names alone are not enough. You can also exclude bots or automated accounts when you want to focus on the people contributing to a project.

A clearer report starts with understanding whose work you are looking at.

Bring related projects together

DeepGraph supports GitHub, GitLab.com, Bitbucket.org, and local Git projects. You can explore one project on its own or group related projects into a folder to see contributions across them.

That can help when a feature spans several projects or a teammate works across different parts of a product. You can follow their recorded contributions without switching between separate histories for every question.

For projects on your computer, local analysis uses the saved Git history. Files you are editing outside that history are not uploaded, but earlier saved versions of files are included in it.

Keep the limits of the numbers in view

Commits and lines changed describe what was recorded. They do not capture the whole contribution someone makes to a team.

• A small fix can resolve a difficult problem.
• A large change can include generated or copied content.
• Design discussions, mentoring, and many reviews may leave no direct trace in the commit history.
• Different projects and ways of working produce different patterns, even when the effort is similar.

Use the report to understand context and ask better questions. It is not a productivity score or a substitute for talking with the people doing the work.

Try it with a project you know

Choose a project you recognize and a question you can check against your own experience. Run a report, open a contributor, and inspect a few changes behind their totals.

Our Getting started guide walks through the steps. You can try a public project from the main page, or compare plans when you are ready to bring in more work.

Follow What's new for product updates, and contact us with questions you would like future articles to explore.