[1/2] AI research in SDLC from IT One and Skolkovo (AI column)
Recently I was at a presentation. study from IT One and Skolkovo, where participated panel discussion. Then I took this research and studied everything. 59 pages with approximately the following structure
- World practice: the current and upcoming transformation of SDLC under the influence of AI - here the authors conducted a meta-study, analyzing the reports of other respected guys (I will give links below when describing the results.)
- AI in SDLC in Russia: market landscape, services, practices and expectations for development - here the authors analyzed public services available in the Russian market + analyzed large announcements from players, even if the solutions themselves are not yet available to a wide range of users
- Technology Leaders’ View on AI in SDLC: Results of Interviews with STO/CIO and Major Company Development Executives – Here the Authors Interviewed with 50+ respected people in the industry (I did not find a list of these people in the report, but we will take the word for it.)
- Forecasts, conclusions and recommendations about the opportunities and risks associated with the use of AI in SDLC - the authors' opinion on how AI will further develop and how to integrate it into the development of AI
Let’s start with global practice in this post. Market AI tools for development : $6,9 billion29,6 billion 2032 (x4). The greatest effect at the stages of development and testing. Source Spherical Insights
- 62% of developers are already using AI 13,8Percentage of plans (data 2024 year). Managers rate penetration lower, but the trend is accelerating. Source StackOverflow 2024 (me handler report 2024 years and handler new report 2025 year) Value shifts from personal effectiveness (faster) to the command: now +10% to the coding speed, in the horizon of several years +25–30% to the productivity of teams in the work of "AI-level-teams" (whatever it means). Source Mia Platform
- SDLC acceleration: already 15–20Right now. (source: Forrester)potential 30–50% on the medium-term horizon (source: medium from Satish Rama, Director Gen AI @ Paypal)
- Horizon of process transformation - 1–3 year; y 32% of tech leaders have already exceeded expectations (Source: MIT, but there is no specific link to the article)
Separately, the authors mention the METR study, where on complex tasks / repositories AI can slow down the work of experienced engineers. Detailed analysis is in my posts. 1 and 2 even podcastwhere it is shown that the study itself is designed for a specific result + represents a very narrow case.
Interestingly, the study authors found a maturity model from the UK government’s DEFRA ministry.The AI-Powered SDLC: A Comprehensive Technical and Cultural Maturity Assessment Framework" At first, I didn’t understand how the agency responsible for environmental protection, food policy and rural development related to digitalization, but then I studied the question and realized:) From the middle 2010DEFRA has become one of the leaders of digital transformation in the British public sector, largely thanks to active cooperation with the Government Digital Service. (GDS) Central Digital and Data Office (CDDO). DEFRA has thousands of data sources (From satellites to farm reports) and a huge variety of users. This is where the British government is actively testing its AI/ML themes and agency systems.
If we talk about the risks of AI, which are noted by Western reports, then they are: Code/data leaks and compliance problems Degradation of engineers’ competencies and blind trust in tips; Growth of tech debt and vendor lock-in on providers
As a result, it is clear that the topic of AI-fication of software development processes is hotter than ever. In the West, only the lazy do not publish their reports/forecasts/products. And for the most part, they're now in a positive tone. But many note that the leap in efficiency will be in the transition from personal copilots to integration into the processes of companies. Next. Let's talk about Russia.
#Software #Engineering #Productivity #DevEx #AI #Management #RnD #Leadership #Economy