Beyond AI tools: Evolving software engineering organizations for the agentic era

Engineering Enablement by DX

0:00

-29:46

Beyond AI tools: Evolving software engineering organizations for the agentic era

Dell’s Jennifer St Pierre explains why the hardest part of AI adoption is leading people through change, not deploying the technology.

Jun 08, 2026

Listen and watch now on YouTube, Apple, and Spotify.

Jennifer St Pierre is Senior Vice President of Developer Experience and Transformation at Dell Technologies, where she leads the strategy for how Dell’s Infrastructure Solutions Group builds, operates, and evolves software.

In this session from DX Annual, Jen argues that the biggest challenge in adopting agentic AI is not the technology itself, but the people transition behind it. Drawing on lessons from earlier shifts like Agile, DevOps, and cloud adoption, she explains why organizations that treat AI as a simple tooling rollout may get compliance, but not commitment.

Jen outlines five leadership imperatives for navigating the transition: building a shared understanding of why change is happening, defining a clear future state, clarifying how roles will evolve, creating psychological safety for experimentation, and aligning metrics and organizational structures with new ways of working. Throughout the talk, she emphasizes that while AI may generate code, humans remain responsible for direction, judgment, and meaning.

Some takeaways:

Treat AI adoption as a people transformation

Technology transitions are really people transitions. The hardest part of adopting agentic AI is not the tooling itself, but helping engineers understand how their roles, workflows, and career paths will evolve.
Compliance is not the same as commitment. Organizations that treat AI as a tooling rollout may achieve adoption metrics, but they will struggle to build the trust and engagement needed for lasting change.
Every major platform shift follows a familiar pattern. New technologies create excitement, skepticism, fear, and eventually productivity gains that become the new normal.

Build a shared understanding of why AI adoption matters

Start with an honest explanation of why the change is happening. If developers do not understand the business rationale, they are likely to assume the initiative is primarily about cost reduction.
Shared understanding does not require universal agreement. It means everyone is working from the same candid view of market pressures, strategic goals, and organizational intent.
Framing shapes emotional response. Positioning AI as a way to help engineers focus on more strategic work creates a very different reaction than simply saying it will increase productivity.

Define a clear future state

A vague vision creates fear. When people cannot picture what their work will look like in 12 to 18 months, they tend to imagine replacement, stagnation, or obsolescence.
Role clarity is essential. Teams need to understand what skills will matter, how performance will be measured, and which responsibilities will increase or diminish.
Specificity beats slogans. Concrete expectations about how AI will be used help people see where they fit in the new model.

Create psychological safety for experimentation

Teams need permission to make mistakes. AI adoption requires experimentation, and experimentation inevitably involves missteps and imperfect results.
Psychological safety helps teams surface problems earlier. When engineers feel safe speaking up, leaders get better information and can address issues before they escalate.
Silence is expensive. Leaders who discourage candor risk making decisions based on filtered or incomplete information.

Align metrics and organizational structures

Old metrics can reinforce old behaviors. Measuring lines of code or heroic firefighting may encourage exactly the habits AI should help organizations move beyond.
Metrics and structure must evolve together. Governance, incentives, funding, and performance systems need to support the behaviors leaders want to see.
Transformation should survive without constant reminders. If the desired behaviors disappear as soon as leaders stop talking about them, the change has not yet become embedded.

Lead the transformation intentionally

Leaders must model the change themselves. Using AI tools, sharing lessons learned, and being transparent about failures builds credibility.
Career paths must be made explicit. Engineers want to know how they can continue to grow and whether deep technical expertise will remain valuable.
AI may generate code, but humans generate direction. Judgment, context, and meaning remain the most valuable contributions people bring to software development.