Engineering metrics in the AI era
Building a Metrics Framework for Human-AI Engineering Teams
A practical framework for measuring engineering performance across four pillars: velocity, quality, oversight, and team health — with AI attribution built in.
Human Oversight Metrics for Agent-Assisted Development
As AI agents take on more coding tasks, oversight becomes a safety-critical metric. Track stale PR age, review depth, and reviewer coverage to maintain control.
Cycle Time and Throughput in the Era of AI Agents
More pull requests don't mean faster delivery. Learn how to decompose cycle time in hybrid AI-human workflows and why percentiles matter more than averages.
Measuring Code Review Effectiveness When AI Writes the Code
AI-generated pull requests demand different review strategies. Learn which metrics reveal whether your team is genuinely reviewing or just rubber-stamping.
Why Engineering Metrics Need Rethinking in the AI Age
Traditional engineering metrics assumed human effort. AI-assisted development inflates output metrics, demanding a shift toward outcome-oriented measurement.