An engineer working in office. Photo courtesy of Groove Technology

However, it is not replacing engineering but reshaping how organizations build and scale software systems, according to executives at Groove Technology.

Rather than committing to large, upfront projects, companies are increasingly testing ideas through smaller, faster development cycles. At the same time, requirements around control, scalability and reliability remain central, prompting a reassessment of how software can be developed in a more practical and business-aligned manner.

Traditionally, enterprise software development followed a structured approach, involving extensive planning, high upfront investment and long implementation timelines. Once projects were underway, making changes was often costly and complex.

Research by McKinsey & Company shows that large IT projects run, on average, 45% over budget and deliver 56% less value than expected. These outcomes reflect structural constraints, including long development cycles, limited flexibility and high initial costs.

According to Matt Long, CEO of Groove Technology, one of the main limitations of traditional development has been the difficulty of testing ideas in a cost-effective way. Companies often needed to validate concepts extensively before building systems, as rework at later stages could be expensive and time-consuming.

Matt Long (2nd, R), CEO of Groove Technology, said companies often needed to validate concepts extensively before building systems. Photo courtesy of Groove Technology

AI-assisted development is beginning to alter this model by accelerating the transition from concept to working software. Gartner estimates that by 2028 more than 75% of enterprise developers will use AI coding assistants, compared with less than 10% in 2023.

Despite growing adoption, many organizations remain cautious. Common concerns include data security, the reliability of AI-generated code and the long-term scalability of rapidly developed systems.

Mai Nguyen, general director at Groove Technology, said businesses are exploring ways to experiment with AI while managing risks, particularly among small and medium-sized enterprises.

As AI tools lower technical barriers, a hybrid development model is emerging. Instead of relying solely on software-as-a-service platforms or large custom systems, organizations are building smaller applications tailored to specific operational needs.

These applications often focus on internal processes such as workflow automation, data tracking and operational reporting. Hung Do, business development manager at Groove Technology, said companies are increasingly identifying opportunities to replace manual processes with targeted digital solutions.

Hung Do, Business Development Manager at Groove Technology, said organizations are building smaller applications tailored to specific operational needs. Photo courtesy of Groove Technology

This approach allows businesses to modernize incrementally while maintaining control over costs, security and scalability. Rather than investing in large, complex systems, organizations can expand applications step by step based on actual needs.

To support this shift, some providers, like Groove Technology has introduced an AI Custom Software Pilot Package designed to help organizations build and validate internal applications safely. These pilots typically involve building a functional application within a short timeframe, enabling evaluation under real operating conditions before broader deployment.

Industry experts note that while AI is accelerating development, it does not eliminate the need for human expertise. Decisions related to system architecture, scalability and long-term design remain dependent on experienced engineers.

As development costs decline, attention is shifting from speed alone to the quality of decision-making in software design. Organizations are placing greater emphasis on selecting the right problems to solve, structuring systems effectively and balancing rapid development with long-term stability.

In this context, AI is increasingly viewed not as a replacement for traditional development, but as a tool enabling a gradual transition toward more flexible, iterative and business-oriented approaches to software engineering.