RU
The State of Software Delivery 2025+ Report

The State of Software Delivery 2025+ Report

Tags: LLM, AI, job, SDLC, Software Delivery, Software Development, CI/CD, DevOps

My comment

The main thing that comes to mind is that AI has not yet learned to "cook". And those who learn to "cook" will be more efficient.
As a result, there will be lower development costs and greater financial efficiency.

Introduction

The report is based on a survey of 500 technical leaders and specialists. The main goal of the study is to understand how AI-based tools affect modern development processes, what challenges they create, and how the role of a developer is changing in the face of growing market demands.

1. The Developer Crisis and "Toil"

The role of a developer has changed dramatically. In addition to writing code, developers are now expected to be experts in security, operational processes (Ops), system performance, and UX. This leads to cognitive overload.

  • Routine takes time: 78% of developers spend at least 30% of their time on manual, repetitive tasks (writing policies, manual testing, and bug fixing).

  • Financial losses: For an organization with 250+ developers, this is equivalent to an annual loss of $8 million in productivity.

  • Burnout: 88% of developers work more than 40 hours per week.

    • 50% report an unhealthy work-life balance.

    • 48% experience increased stress and anxiety.

2. The Paradox of AI Implementation

AI CodeGen tools are seen as a solution, but they create new challenges in practice.

Expectations vs Reality

  • Optimism: 98% of developers believe that AI tools are a great way to reduce burnout.

  • New challenges: Implementing AI shifts the burden from writing code to testing and debugging it.

    • 67% of developers spend more time debugging AI-generated code.

    • 68% spend more time fixing security vulnerabilities.

    • 92% say AI increases the "blast radius" of failed deployments.

    • 59% face issues deploying AI-generated code at least half the time.

3. The Shadow AI Problem

The lack of clear policies and user-friendly tools pushes developers to use unauthorized software.

  • 52% of developers do not use AI tools provided by the company's IT department.

  • This creates risks of intellectual property leakage and security issues, as code is sent to external services without control.

  • 60% of respondents currently do not evaluate the effectiveness of the AI tools they use.

4. AI Beyond Code Generation

The main conclusion of the report is that the benefits of AI will not be fully realized if it is limited to code writing.

  • 95-96% of respondents agree that the benefits of AI will only be realized when it is integrated into the entire software development lifecycle (SDLC).

  • Plans for the next 12 months: Companies plan to implement AI in the following areas:

    1. CI/CD (continuous integration and delivery) - 50%.

    2. Performance optimization - 48%.

    3. Security and compliance - 42%.

5. Fear of Job Loss and Role Transformation

There is significant anxiety about the future of the profession.

  • 90% of respondents are concerned that AI will replace developers.

  • Reality: The report states that instead of replacement, there will be a transformation of the role. AI will take over the routine, while developers will focus on architecture, business logic, and validating AI results. New features such as prompt engineering and AI integration into workflows will emerge.