What engineering leaders are struggling with that vendors aren't solving
Blog post from Upsun
Engineering leaders adopting AI are often unable to verify promised efficiency gains because many organizations lack consistent baseline measures for cycle time, review turnaround, incidents, and related delivery outcomes. Increased pull request volume or faster code output may not indicate better performance if it produces poor-quality features, additional support work, or more incidents. Jonas Kröger of Upsun argues that established DORA metrics—deployment frequency, lead time, change failure rate, and recovery time—provide a more useful framework for evaluating AI’s effects than creating untested AI-specific measures. Leaders also face uncertainty over AI spending, particularly when API costs lack caps or visibility and when organizations cannot determine whether higher expenditure improves results. While many teams are already pursuing agentic workflows, their primary challenge is scaling successful use beyond isolated teams, which Kröger says requires enablement, onboarding, workflow support, cost controls, and human oversight in addition to model capabilities.
No tracked trend matches for this post yet.
Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.