developers and agents
getty
Generative AI and agent AI take over the lower-level tasks of software development, such as code generation, testing, and documentation, enabling software to be delivered quickly on demand. But is performance at scale possible for large enterprises? One large insurance company discovered the following in a large-scale pilot employing autonomous software development tools.
GNP Segros, Mexico’s largest insurance company, has seen tangible results from its AI-assisted development efforts, increasing the productivity of its development team by 5-10x.
These are the words of Enrique Ibarra, CIO and Head of Business Transformation at GNP. He moved to autonomous software development at scale, enabling 1,000 developers to improve the speed and quality of their output.
“We want to change the role of humans in the software development process,” he told CXOTalk host Michael Krigsman in a recent interview. “We wanted to do this to achieve agility. We wanted to test a new generation of tools for generating software in an autonomous way.”
According to Ibarra, 80% to 95% of GNP’s development work is now completed autonomously. “Development speed has increased by a factor of 5 to 10. Humans aren’t writing the code; they’re telling the platform how to write the code. This is a big shift in paradigm.”
The transition to autonomous development is not just about AI and the machines that underpin it. Success depends on the people driving these projects. “You can’t just flip a switch and go fully autonomous,” Ibarra said. “We need to build trust through a step-by-step, human-involved approach.”
Cultural change is a key and necessary part of transformation. “We need to change the way we think about engineering,” Ibarra said. “We need to train engineers to move from being creators to being editors and orchestrators. Those are different roles.”
The human leader’s job became “to define the prompts, review the architecture, and validate the AI’s execution,” he added.
Ybarra refuted the idea that AI could significantly reduce opportunities for developers and IT professionals. Instead, roles are being redefined around direction rather than coding, he said. The company says it “moves engineers from line-by-line coding to rapidly creating autonomous output, architecture review, and validation, with the co-pilot handling the rest.”
He stressed that developers will always be in great demand. “If you develop a system, you need to provide technical guidelines. You need to provide guidelines related to the platform the software will run on, and the end user can’t do that.”
GNP deploys its systems on Google Cloud, which requires human knowledge of “technical features and technical configurations,” he explained. “In order for the system to run, it has to be enabled on a cloud platform. End users aren’t going to do that.”
The role will change completely, he added. “Humans aren’t writing the code; they’re telling the platform how to write the code. So this is a big shift in paradigm.”
