Skip to main content

Version v2.5.0

· 3 min read
Rafał Marczuk
Rafał Marczuk
Senior Business Analyst @ Q247

Version 2.5.0 introduces a key feature for tracking AI adoption in the daily work of development teams — an important capability given the dynamic changes artificial intelligence is bringing to the software development process. Organizations gain visibility into the real scale of AI-assisted tool usage in code creation, with data presented at the organization, team, project, and individual participant level. We have also clarified how documentation content volume is calculated regardless of its source, and expanded visibility of project data for selected administrators.

Below you will find a detailed description of the most important changes.

New Features​

Tracking AI Adoption in Development Teams' Work​

Q247 gains a new, comprehensive view showing to what extent teams are using AI-assisted tools — such as Claude Code, GitHub Copilot, Cursor, Aider, Devin, or OpenAI Codex — when writing code.

AI adoption indicator on the Overview page (collapsed view)

The view presents the AI adoption indicator along with its trend over time. It also shows the development effort attributed to AI-assisted code (AI calories) against the team's total effort (Coding Calories). This is complemented by token usage and the dollar cost resulting from the use of AI tools. It is also visible how well AI-assisted code holds up over time compared to code written without it.

Expanded AI adoption view with a per-project breakdown of calories, tokens, and cost

The data is available on the organization's home page, in the Teams view, and in the detailed views of every project and team, always with the option to expand into more granular data.

AI adoption widget on the Participant view (OTIF – Retention)

Administrators can link accounts used in AI tools to Q247 users (matched automatically by email address by default, with the option for manual correction).

Users and teams list with the option to edit AI user mapping Edit AI user mapping dialog for linking an AI tool account to a Q247 user

A commit is recognized as AI-assisted when its commit message contains a characteristic phrase — set by default for each supported tool (e.g. Co-Authored-By: Claude for Claude Code), with the option to add custom detection patterns tailored to the organization's specifics.

AI adoption pattern configuration listing detection patterns for supported AI tools

In the current version, the feature collects token usage and cost data from Claude — we are also ready for further integrations with other AI tools, based on customer-reported needs. The feature requires updating the Enterprise Plugin to the latest version, but does not involve any additional installation on users' computers.

Improvements​

  • Consistent documentation content volume counting — the way documentation content volume is calculated is now unified regardless of its source.
  • Broader visibility of Project Managers for organization-level admins — administrators with the Team Member role at the organization level can now see the full list of Project Managers in the organization management view.