How Braze’s CTO is rethinking engineering for the agentic area

2026-05-13 · Source: Stack Overflow Blog · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering, Emerging Technologies & Innovation · Depth: Intermediate, extended

Summary

Jon Hyman, co-founder and CTO of Braze, discusses the company's rapid transformation into an AI-first engineering organization over just a few months. Braze, a customer engagement platform, achieved this by enabling engineers with tools like GitHub Copilot and Claude Code, conducting AI lunch-and-learns, and demonstrating tangible success, such as building an MCP server six weeks ahead of schedule using AI. This led to over 60% of committed code being AI-generated. Hyman emphasizes that model quality, not mandates, drove adoption among their 300-person engineering team. He also addresses the significant challenge of measuring AI's business value and the surprisingly high cost of inference at scale, noting that an engineer can spend $150 daily on inference, totaling $4,500 monthly. The discussion also covers integrating an acquired company, OfferFit, and the evolving relationship between engineering and product management due to AI-driven rapid prototyping and UX debt reduction.

Key takeaway

For MLOps Engineers managing AI integration, prioritize demonstrating clear, measurable gains in model quality and project velocity to overcome skepticism and drive adoption. Your focus should shift to optimizing inference costs, which can be substantial, by standardizing agentic workflows and selecting appropriate models for specific tasks. This approach ensures AI investment translates into tangible business value and sustainable growth, rather than just increased operational expenses.

Key insights

Model quality and tangible results are key to driving rapid AI adoption within large engineering teams.

Principles

• AI enables stepwise increases in team efficacy.
• High inference costs necessitate efficient LLM usage.
• AI induces demand for more software development.

Method

Enable teams with diverse AI coding tools, provide training, demonstrate success with low-stakes projects, and integrate AI into workflows to raise output expectations.

In practice

• Trial Graphite for pull request management.
• Use Vercel and Cursor for rapid interactive mock-ups.
• Codify internal coding standards for AI agents.

Topics

• AI-First Transformation
• Engineering Leadership
• AI-Generated Code
• Inference Costs
• Rapid Prototyping

Best for: MLOps Engineer, AI Engineer, Machine Learning Engineer, CTO, Director of AI/ML, VP of Engineering/Data

Related on AIssential

• The Time Savings Era of AI is Over — AI benefits are shifting from time savings to increased output and new capabilities, driven by agentic and multimodel usage.
• AI is the Concorde of our time (Ep. 309) — Frontier AI's technical brilliance is overshadowed by its economically unsustainable, industrial-scale operational costs.
• You Shipped It Fast. But Did You Ship It Right? — AI-generated code creates an "illusion of correctness" by hiding critical unstated assumptions that lead to production incidents.
• Agentic AI costs more than you budgeted. Here’s why. — Agentic AI operational costs often dwarf development budgets due to compounding expenses in inference, governance, and infrastructure.
• AI economics reshape FinOps as enterprises seek greater visibility and control — AI's rapid adoption necessitates FinOps evolution, integrating cost accountability into code-level decisions and development workflows for optimal value.
• Can tech companies learn to love cheaper AI models? — Rising costs are shifting AI model strategy from "bigger is better" to "efficient is optimal," prioritizing smaller models for most workloads.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Stack Overflow Blog.