DORA 2025 State of AI-assisted Software Development - General information on the report (Devops column)
I’ve talked briefly about the study.DORA Research: 2025"in the past posts. Then I read the report carefully, but then told about its results very briefly. And now I need a more detailed analysis for The Impact of AI on Engineering Culture So I decided to write it for my subscribers.
State of AI-assisted Software Development 2025prepared by the DORA research team (DevOps Research and Assessment)This is part of the Google Cloud. The lead authors were Google Cloud specialists with the participation of invited experts. The research partners were IT Revolution, GitHub, GitLab, SkillBench and Workhelix. The initiative was also supported by a number of sponsors. (Swarmia, Thoughtworks, Deloitte, Atlassian and others.).
The study conducted a global survey of almost 5 000 Technical specialists from different countries and companies. Respondents included development engineers, DevOps/SRE, teamlids, product managers. The survey covered a wide range of topics, from the extent and ways AI is used (What tools and LLM models are used, for what tasks, how much time is spent, how often do they trust AI tips?) before the characteristics of the commands and processes (architecture and development platforms, learning culture, Value Stream Management, version control practices, size of change batches, etc.). Team performance was also evaluated: classic DORA metrics (For example, delivery speed and stability of software release)Code quality, developer productivity, friction frequency and employee burnout. To deepen the analysis, the researchers collected more 100 hours of high-quality data – interviews and practice reviews to supplement the figures with live insights.
Interestingly, six months ago I told How the DORA methodology works to compile such reports 2025 The authors separately described “Measurement Framework"and how it can be used to manage change in your company." If we talk about the kids and their DORA report. 2025Data processing combined statistical and analytical methods. A large-scale survey allowed for correlation analysis and regression models, revealing the relationship between practices and final performance indicators.
As a result, a report consisting of the following parts - "Beyond the tools": It discusses why successful implementation of AI is a system task, not just a tool choice. Analysis of the current status of AI-assisted developmentThe results of a survey on the level of AI adaptation in the industry, usage practices and impact on key indicators are presented. - "Understanding your teams: 7 profiles": In this section, DORA presents the seven archetypes of development commands discovered. **- "DORA AI Capabilities Model"**A central chapter introducing a new ability model for success with AI. Seven technical and cultural practices are listed here. (systemic factors)which statistically enhance the positive impact of AI on team performance - "Directing AI’s potential": a section on how to steer AI efforts in the right direction. The role of Value Stream Management (VSM) As a mechanism to convert local improvements from AI into the end result for business Recommendations and conclusions for leadersThe final part of the report translates the results into practical. Here is a roadmap for implementing AI – which areas to focus on first. The report explicitly states that executives should view AI adoption as an organization transformation, not just an IT project.
P.S. In the comments I will add a PDF with the report.
#AI #Software #Engineering #Management #Whitepaper