In a dramatic strategic pivot, Braze co-founder and CTO Jon Hyman revealed today how the customer engagement platform transformed its entire engineering organization into an AI-first team in just a few months, after nearly 15 years of traditional growth. The move signals a fundamental rethinking of software development for the agentic era.
“We realized that incremental improvements wouldn't cut it anymore—we had to completely rethink how we build, test, and deploy software,” Hyman said in an exclusive interview. “The shift from a human-centric engineering culture to one where AI agents are first-class contributors required us to change everything from our team structure to our code review processes.”
Braze, which went public in 2021, now embeds AI agents across its development pipeline, from automated code generation to real-time system monitoring. Hyman noted the transformation was accelerated by the company's existing data infrastructure, but still required “unprecedented speed of decision-making.”
Background
Founded in 2011, Braze has long prided itself on its engineering-driven culture. Under Hyman's leadership, the company scaled from a small startup to a publicly traded firm with over 1,200 employees. The traditional engineering model emphasized human review cycles and waterfall-style planning.
However, with the rapid rise of large language models and agentic systems in early 2024, Hyman decided to overhaul the entire approach. The company created dedicated AI squads, retrained senior engineers, and implemented new tools for AI-assisted development within a tight three-month window.
What This Means
Industry analysts say Braze's move could set a precedent for other tech companies facing similar pressures. “This is one of the fastest full-scale engineering transformations I've seen at a public company,” said Dr. Elena Martinez, software engineering researcher at MIT. “If Braze succeeds, it will provide a blueprint for other mid-to-large firms.”
The agentic model shifts responsibility from individual engineers to AI “agents” that can autonomously propose, test, and deploy code under human supervision. Hyman emphasizes that the change does not replace engineers but elevates their role. “Our engineers now act more like conductors of an AI orchestra, which is a profound change we're still learning to manage,” he added.
Braze reported no major service disruptions during the transition, and early metrics show a 20% increase in development velocity. Hyman cautions that other firms should expect cultural friction. “The hardest part isn't technology; it's convincing senior engineers to trust AI agents with critical code paths,” he said.
For the broader industry, this signals that the era of purely human-led engineering may be ending. As more companies adopt agentic models, we can expect faster iteration cycles but also new challenges around code ownership and accountability.