Engineering for the Agentic Era: Inside Braze's AI-First Transformation
Introduction
In just a few months, Braze—a customer engagement platform that has grown over nearly 15 years—transformed its engineering organization into an AI-first team. Jon Hyman, co-founder and CTO, spearheaded this shift, rethinking everything from development workflows to team culture to prepare for the agentic era where AI agents work alongside humans. This article explores how Hyman led the transition and what it means for engineering leadership in a rapidly evolving technological landscape.
The Growth Journey: From Startup to Scale-Up
Braze started as a small team building a customer engagement platform. Over nearly a decade and a half, its engineering organization scaled exponentially. Hyman, as CTO, faced classic scaling challenges: maintaining velocity, ensuring code quality, and fostering innovation while hiring hundreds of engineers. However, the arrival of generative AI and autonomous agents forced a strategic pivot.
The Pivot to AI-First: Why and How
In early 2024, Hyman recognized that AI agents were becoming integral to customer experiences—not just as features but as core infrastructure. He decided that Braze’s engineering team must become AI-first, meaning every product decision and development process would incorporate AI capabilities. The transformation, achieved in just a few months, involved:
- Upskilling engineers in machine learning and prompt engineering.
- Redesigning architecture to support agentic workflows.
- Creating internal tooling for rapid AI prototyping.
Hyman says the key was iterative change rather than one big overhaul. Teams started by integrating AI into existing features—like smart notifications—and gradually expanded.
Overcoming Resistance
Engineers initially worried about job displacement or added complexity. Hyman addressed this by framing AI as an amplifier of human creativity. He also set up cross-functional squads where engineers collaborated with data scientists and product managers to build agentic solutions.
Key Strategies for the Agentic Era
Based on Braze’s experience, Hyman advocates three pillars for engineering leaders:
- Culture over Tools: No amount of AI tooling will succeed without a team that embraces experimentation. Braze instituted “AI hackathons” and rewarded risk-taking.
- Data Foundations: Agentic AI requires clean, real-time data pipelines. Hyman invested in data infrastructure first.
- Ethical Guardrails: Build responsible AI from day one—transparency, bias checks, and human oversight.
For more on fostering an AI culture, see Leadership Lessons below.
Leadership Lessons from Jon Hyman
Hyman’s approach offers insights for CTOs, VPs of Engineering, and team leads. Key takeaways:
- Communicate the “why”: Without vision, teams cannot commit to rapid change.
- Lead by example: Hyman learned to code AI integrations himself to understand challenges.
- Balance speed and stability: The transformation happened in months, not years, but with careful testing.
He also emphasizes that engineering leaders must become AI students—continuously learning about new models, agents, and infrastructure.
The Future: Engineering Teams as AI Orchestrators
Braze now treats its engineering org as an orchestrator of agents. Instead of manually writing every feature, engineers define goals and let AI agents propose solutions. This doesn’t eliminate jobs; it elevates them. Hyman predicts that by 2026, most SaaS companies will have similarly transformed—or risk being left behind.
Conclusion
Jon Hyman’s near-15-year journey at Braze culminated in a swift, decisive shift to an AI-first engineering culture. By focusing on culture, data, and ethical design, the team successfully embraced the agentic era. For any engineering leader, the lesson is clear: prepare now to reimagine your organization—because the transformation can happen faster than you think.