The evolving role of DevProd teams in the AI era
AI is creating a new mandate for DevProd teams. Here’s what they should focus on to remain relevant and impactful.
Welcome to the latest issue of Engineering Enablement, a weekly newsletter sharing research and perspectives on developer productivity.
Developer productivity teams have always had a broad, sometimes ambiguous mandate. With AI now transforming software development, these teams are being asked to take on even more—evaluating and rolling out AI tools, setting guardrails, measuring impact, and finding ways to apply AI to challenges like migrations and technical debt—all while balancing their traditional responsibilities.
We recently held a live discussion on this shift, exploring the “new mandate” for DevProd teams—what they should focus on now to remain relevant and impactful. The conversation touched on topics such as vendor selection, measurement practices, paved paths versus tool sprawl, and the one-to-many opportunities for platform teams to leverage AI at scale.
This week’s newsletter highlights key questions from that discussion. You can also rewatch the full conversation here.
Q: How is AI changing the mandate for DevProd teams?
Abi: DevProd teams have always been the governors of the developer toolchain, so it’s natural that organizations are now turning to them for leadership in selecting and guiding AI adoption. This is a major opportunity. AI is top of mind for executives, and platform leaders are well-positioned to be the stewards of this change.
We’re seeing DevProd teams asked to evaluate the AI tooling landscape, recommend which vendors to pilot, decide on build-versus-buy strategies, manage procurement, roll tools out to developers, and then demonstrate impact. In other words, the full lifecycle of AI adoption is quickly becoming a core responsibility and should be treated as a central pillar of the platform roadmap.
Q: Why is measurement so critical in this new era?
Laura: For DevProd leaders rolling out AI tools, having strong measurement in place is essential. These are big decisions with big budgets and consequences. You need to be able to answer questions like: Are we picking the right tools to invest in? Are we making them available to developers in the right way?
Some measurement principles stay the same, but AI also requires new approaches. Because the way software is developed is shifting, leaders need to adapt their data collection and metrics to capture AI’s unique impact.
Abi: Last month, Laura and I did a live session on the AI Measurement Framework, which lays out how organizations are approaching this. Historically, the industry focused heavily on delivery metrics like DORA. More recently, attention has shifted toward broader productivity signals and developer experience.
AI is the next evolution. DevProd teams now need to ask: How do we measure AI? How do we measure agents? What signals matter most as the way we build software changes? This is the new page-turn in how measurement programs need to evolve.
Q: What about tool sprawl—how should teams handle that with AI?
Abi: The results from AI tools vary widely depending on the context—your codebase, language, and technology stack. Forward-thinking DevProd leaders see this as similar to past challenges with tool sprawl. Standardization has always been a source of leverage: narrowing down to a smaller set of tools and ways of working lets you invest in making them excellent and seamless for developers.
Right now, many organizations are piloting lots of tools and throwing them at developers, which creates uneven results. Many developers also don’t feel fully equipped to maximize the benefits. That’s where DevProd teams come in—not just with training and enablement, but by creating paved paths, narrowing down tools for specific use cases, and building reusable workflows and templates. This is the same kind of work DevProd teams have always done, but AI makes it an even bigger opportunity.
Laura: This is really the same problem we’ve faced with tool sprawl before. The fact that it’s AI doesn’t make it different. It’s still the DevProd team’s job to solve it. That said, there are two unique aspects here. First, experimentation is good, but tool sprawl is bad. When teams are hacking together new AI assistants and agents, we want to encourage that experimentation because it helps uncover new use cases. But when it spreads unchecked, it creates friction.
The key is to provide the right channels for experimentation and to get curious about why teams are building their own tools. What problem are they solving? Can we unify and standardize? This is where the “platform as a product” mindset is critical. Think of it like gardening: encourage organic growth, but prune and trim where needed so the system stays healthy and the best solutions can spread across the organization.
Q: The use of AI in engineering drives faster iteration and higher throughput, resulting in more code shipped overall. How can developer productivity teams ensure the platform scales to meet this increased demand?
Abi: The unbelievable pace at which code can be generated with AI tools needs to be counterbalanced with safeguards, guardrails, and quality checks. SRE and Platform teams have always emphasized the importance of systematic quality checks, feedback loops, and tests. It's more important than ever because we're moving faster than ever with AI. DevProd teams must ensure that the code we ship to production is good quality, secure, and has checks in place to prevent issues.
Laura: The volume of code that's going to start coming through these platforms and build systems will test the boundaries and quickly show you what's broken. DevProd engineering teams can stay ahead of this by ensuring that the fundamentals are in place. Ask yourself the questions:
How will we know that code shipped, where 50% of it is AI-authored, is secure and compliant according to our standards?
How will we know when our systems can't check that sufficiently?
Q: Should DevProd teams also use AI themselves?
Abi: We’ve noticed that while DevProd teams are rolling AI tools out to developers, they often aren’t using these tools themselves. There’s no clear role yet for DevProd to fully apply and leverage AI directly, and that’s a missed opportunity.
Developers don’t always have the time, expertise, or motivation to apply AI to problems it’s particularly well-suited to solve, like refactoring, patching, or other KTLO work. These tasks rarely make it onto product roadmaps, which means they get overlooked. DevProd teams are in a unique position to step in here. They can think about one-to-many opportunities, owning a kind of “AI headcount” for the organization and applying it horizontally across teams, rather than just handing tools to individuals and hoping for acceleration.
Laura: Whenever DevProd teams can apply AI to workflow problems or organizational issues, the leverage is significantly greater, and adoption will be stickier because you're solving actual, real problems at the team level, not just for individuals trying to speed up their individual tasks.
Final thoughts
One of the most important things that platform and developer productivity teams can do right now is to stay focused on the bigger picture. Everyone is focused on AI tools, but when we look at the data, we see two interesting things:
1. Across the board, the acceleration gains from AI are still outweighed by some of the inefficiencies that exist across the SDLC. All of the traditional issues, such as excessive meetings, interruptions, and code quality, still hinder us today and remain problems we need to solve.
2. There's a refocus on developer experience with AI because many organizations are discovering that the same factors that enable human developers are also what makes AI effective. Things like well-documented code, feedback loops, and documentation all have a significant impact on the efficacy of LLMs’ ability to navigate the system and make changes confidently.
AI should be seen as part of a broader strategy. DevProd teams that pair AI adoption with a renewed focus on developer experience and productivity fundamentals will create lasting, organization-wide impact.
Who’s hiring right now
This week’s featured DevProd job openings. See more open roles here.
Deliveroo is hiring a Staff Platform Product Manager | London, UK
Rippling is hiring a Director, Platform Engineering | San Francisco, CA
Atlassian Williams Racing is hiring a Software Engineer - Engineering Acceleration | Grove, UK
RH is hiring a Director, Platform Engineering | Pleasanton, CA
ScalePad is hiring a Head of AI Engineering & Enablement | Canada (Remote or in-office)
That’s it for this week. Thanks for reading.
-Abi