Developers and agents

getty

Generative and agentic AI are taking up the lower-level tasks of software development, such as code generation, testing, and documentation, enabling the rapid delivery of software as needed. But can it perform on a massive scale, for a large enterprise? One major insurance company is finding out in an extensive pilot that employs autonomous software development tools.

GNP Seguros, Mexico’s largest insurer, is seeing tangible results from its AI-assisted development work, with a five to ten-fold productivity boosts for its development teams.

That’s the word from Enrique Ibarra, CIO and head of business transformation for GNP, who has moved into autonomous software development in a big way, enabling its 1,000 developers to ramp up the speed and quality of their deliverables.

“We want to change the role of humans in the software development process,” he told Michael Krigsman, host of CXOTalk in a recent interview. “We want to do it to achieve agility. We wanted to test a new generation of tools for generating software in an autonomous way."

Between 80% and 95% of development work at GNP is now completed autonomously, Ibarra said. “We accelerated development velocity of between 5X and 10X. The human is not writing the code. The human is directing a platform on how to write the code. That’s a huge change in paradigm.”

Moving to autonomous development isn’t just about AI and underlying machines – it’s success depends on the people who will move these projects forward. “You don’t just flip a switch to full autonomy,” said Ibarra. “You have to build trust through a phased human-in-the-loop approach."

Cultural change is a key and necessary part of the transformation. “We need to shift the engineering mindset,” Ibarra said. “We have to train our engineers to transition from being creators to editors and orchestrators. It's a different role.”

The human leader’s job, he added, “now becomes defining the prompt, reviewing the architecture, and validating the AI's execution.”

Ibarra pushed back on the notion that AI may severely reduce opportunities for developers and IT professionals. Instead, roles are being redefined around direction, not coding, he said. The company is “shifting engineers from line-by-line coding into prompt authorship, architecture review, and validation of autonomous output, with co-pilots handling any residual work.”

Developers will always be in great demand, he emphasized. “If you’re going to develop a system, you need to provide technical guidelines. You need to provide guidelines related to the platforms where the software is going to run, and an end user cannot do that."

GNP deploys its systems in Google Cloud, requiring human knowledge about “technical features and technical settings,” he explained. “You need to enable in the cloud platforms in order for the system to run. An end user is not going to do that.”

The role changes completely, he added. "The human is not writing the code. The human is directing a platform on how to write the code. So that’s a huge change in paradigm.”