The increasing use of AI in software development is reshaping how engineering work is approached, particularly in large-scale development tasks. This shift is already changing how development teams design and deliver software.
In this expert opinion, Oleksii Zarembovskyi, our Lead Data Engineer, compares traditional and AI-supported workflows in a real development context. He focuses on how different stages of the development process are affected, the role human expertise continues to play, and the broader implications of AI adoption in the IT industry.
Can you describe your experiment and its results?
The experiment was designed to evaluate the AI's effectiveness compared to a traditional manual development approach. It is based on a hypothetical project to migrate data from a legacy system to a new application. The process is divided into four key stages. Time and effort are compared across each stage.
1. Codebase analysis stage: Manual analysis of a medium-sized codebase typically takes around 12 hours. AI can do this in a fraction of the time — but only if given the right context. That means feeding it technical documentation, business requirements, and task details upfront. The quality of the output depends directly on the quality of the input.
2. Architecture and development stage: This is where developer expertise really matters. An experienced engineer knows how to define the architecture, migration principles, and tech stack clearly within a prompt. Using precise technical language — and prompting in English, since most AI agents are trained primarily on English data — leads to significantly better results.
3. Testing and debugging stage: Execution and debugging are traditionally one of the most time-consuming stages. AI provides substantial support by automatically generating test cases and validating different scenarios. This does not replace a quality assurance specialist. It reduces routine workload and allows QA engineers to focus on higher-value quality concerns. Human validation remains necessary, as AI lacks business context and practical experience.
4. Change requests stage: Handling change requests—an integral part of any project—is also significantly accelerated. Tasks such as adding a new table to a migration process can be completed in under an hour using AI.
Overall, the productivity increase is significant. However, this level of efficiency is largely attributed to prior domain knowledge. For less experienced specialists, a significant portion of time would be spent in the discovery phase exploring options and evaluating trade-offs, thereby reducing the net productivity gain.
How do you assess the role and limitations of AI in software development?
AI should be viewed as a powerful tool rather than a replacement for developers. Expertise, domain knowledge, and an understanding of business context are becoming increasingly important.
AI models are trained on publicly available data and do not have access to proprietary corporate knowledge. A widespread shift toward AI-generated code could lead to stagnation and reduced innovation, as new knowledge may not be incorporated into training datasets. The expert compares this to an ouroboros—a system consuming its own output.
Rather than decreasing workload, AI introduces new complexities. Developers must perform detailed reviews, optimise outputs, and debug AI-generated code. This shift increases cognitive load and the risk of fatigue-related errors. In critical systems, the combination of AI limitations and human error presents significant risks.
What are your views on AI’s influence on the software engineering job market?
There is a growing misconception that a single senior developer equipped with AI can replace multiple junior developers. This may lead to reduced hiring at the entry level, creating long-term risks of talent shortages, as junior developers are essential for developing future senior professionals.
Public narratives around layoffs in major tech companies such as Meta and Oracle often overlook the fact that many specialists are later rehired. This indicates that demand for skilled developers remains strong.
How do you sum up AI's role in software development?
The goal of software development is to deliver a product that meets user and business needs. Artificial intelligence serves as a tool to support this goal, and its effectiveness depends entirely on how well professionals who understand both technical and business contexts use it.
Related insights
The breadth of knowledge and understanding that ELEKS has within its walls allows us to leverage that expertise to make superior deliverables for our customers. When you work with ELEKS, you are working with the top 1% of the aptitude and engineering excellence of the whole country.
Right from the start, we really liked ELEKS’ commitment and engagement. They came to us with their best people to try to understand our context, our business idea, and developed the first prototype with us. They were very professional and very customer oriented. I think, without ELEKS it probably would not have been possible to have such a successful product in such a short period of time.
ELEKS has been involved in the development of a number of our consumer-facing websites and mobile applications that allow our customers to easily track their shipments, get the information they need as well as stay in touch with us. We’ve appreciated the level of ELEKS’ expertise, responsiveness and attention to details.
