Advising

AI’s effect on the software development lifecycle is a developer productivity problem. Faster code generation moves the places where work gets stuck and changes what teams need from their tools and processes. I’ve spent the last decade on exactly these questions, from both sides: leading the teams responsible for developer productivity, and advising the leaders who make the calls. I help engineering leaders understand what AI is changing, decide how to respond, and measure whether their investments improve software delivery.

I work with VPs of Engineering, CTOs, and leaders of platform and developer experience groups on questions like these:

• Where should we invest? Start from the bottlenecks in software delivery and use them to prioritize AI and internal platform work, so the investment goes where work is stuck.
• How will we know it’s helping? Measure what AI changes about software delivery and developer productivity, which usage numbers alone can’t tell you.
• How should we make these decisions? Build AI strategy and governance with input from individual contributors and leaders, and treat people’s concerns and feelings as part of the work.
• How do we explain the choices? Give senior leadership a clear account of the problems, tradeoffs, and evidence behind an AI investment, including what the numbers can’t tell us yet.

Experience #

I’m a Principal Engineer at Honeycomb, focused on AI Operations within R&D. I own the work of maximizing AI leverage across R&D while keeping the culture healthy and inclusive, and I work closely with engineering leadership on AI strategy and governance, especially internal AI usage.

I helped Honeycomb’s engineering enablement team apply a developer productivity framework to prioritize internal AI platform work and measure its impact, which is how the team came to focus on the code review bottleneck. Today, more than 25% of pull requests in key Honeycomb repositories, including the main monorepo, auto-merge without human review.

Automating review also means understanding why people value human code review in the first place, and I’m using an RFC to explore that question before pushing automation further. I bring individual contributors and leaders into decisions like this one, because what people want to preserve matters as much as what they want to change.

As Field CTO at Swarmia from January 2023 to May 2026, I talked with hundreds of engineering leaders about developer productivity, measurement, and what AI is and isn’t changing about software delivery. I co-wrote Build: Elements of an Effective Software Organization with Otto Hilska, contributed to the 2025 DORA report, and hosted Swarmia’s podcast. Before that, I led developer productivity and platform teams at Stripe and Indeed.

Get in touch #

Tell me what your organization is working through at [email protected]. After an initial conversation, we’ll figure out how I can help.