30 pts
Code Quality
Architecture decisions, maintainability, defect risk, review signal, and whether AI-assisted code remains safe to ship.
Developer Impact Index
Measure developer impact after AI changed the meaning of output.
DII is Deveval's 0-100 score for understanding who is shipping valuable, high-quality software — not just generating more commits, pull requests, or AI-assisted code.
30 pts
Delivery Capability
Merged product work, business logic shipped, feature completeness, and consistency of real output over time.
25 pts
Engineering Efficiency
PR cycle time, first-pass merge rate, review friction, and whether AI usage speeds delivery instead of creating noise.
15 pts
Contribution Breadth
Cross-repository ownership, review contributions, collaboration leverage, and maturity beyond isolated code volume.
Why it exists
AI made activity metrics easier to inflate and harder to trust.
Commit count, pull request volume, and lines of code used to be rough proxies for effort. AI coding tools changed that. A developer can now create more code without creating more durable product value.
DII shifts the question from “who produced the most activity?” to “whose work improved delivery, quality, and team leverage?” That is the measurement engineering leaders need when AI usage is rising faster than shipped outcomes.
Signals behind the score
Merged pull requests and shipped output patterns
Code quality and maintainability signals from changed code
Cycle time, review rounds, and first-pass merge behavior
AI usage patterns compared with downstream delivery
Repository ownership, review help, and team leverage
What is the Developer Impact Index?
The Developer Impact Index, or DII, is a 0-100 score that summarizes engineering impact across code quality, delivery capability, engineering efficiency, and contribution breadth.
Why is DII better than commit count?
Commit count measures activity. DII measures whether engineering work turns into useful, high-quality shipped output, which matters more in the AI coding era.
Does DII punish AI usage?
No. DII rewards effective AI use when it improves output, quality, and efficiency. Heavy AI usage without delivery impact is treated as a warning signal.
Who should use DII?
DII is designed for CTOs, VP Engineering leaders, engineering managers, founders, and non-technical stakeholders who need a readable view of developer impact.
See who is creating real engineering impact.
Connect GitHub, run an evaluation, and get a board-ready view of AI usage, shipped output, and developer impact in minutes.
Start free