Andrew Sales, Chief Product Officer, Scaled Agile – Interview Series

Andrew Sales, Chief Product Officer and Chief Methodologist at Scaled Agile, is an experienced Lean, Agile, and DevOps leader who has spent more than two decades helping enterprises transform the way they deliver value. In his current role, he guides product strategy and the evolution of the Scaled Agile Framework (SAFe), working with global organizations to adopt agile practices at enterprise scale. Before joining Scaled Agile, Sales held senior leadership and consulting positions at CA Technologies, Rally Software, and Pearson, where he advised organizations across Europe, the Middle East, and Africa on agile transformation initiatives spanning financial services, telecommunications, retail, software, and education.

Scaled Agile, Inc. is the company behind the Scaled Agile Framework, better known as SAFe, a widely adopted framework for helping enterprises apply Lean, Agile, and DevOps practices at scale. Through its framework, training, certification programs, partner network, and SAFe Studio platform, Scaled Agile supports organizations looking to improve alignment, delivery speed, product quality, employee engagement, and business agility. The company reports that SAFe has been adopted by more than 20,000 enterprises globally and that more than one million professionals have been trained in the framework.

What lessons from large-scale transformations most directly influenced the creation of AI-Native SAFe?

Over the past 15 years, we’ve watched more than 20,000 organizations adopt SAFe to solve one fundamental challenge: coordinating complex product development at enterprise scale. One lesson has remained remarkably consistent: Lean and Agile principles continue to work. What changes is how those principles need to be applied as technology changes.

AI doesn’t make those principles obsolete; it changes where the constraints are. Teams are becoming smaller, more AI-augmented, and increasingly focused on directing work rather than performing every task themselves. Decision cycles are compressing, and the interaction between human judgment and AI execution is becoming central to how value is created.

AI-Native SAFe builds on the proven foundation of SAFe while evolving the operating model for this new reality. It preserves the focus on customer value, cross-functional collaboration, portfolio alignment, and continuous improvement, while introducing new ways to govern AI, manage data as a strategic asset, and establish accountability in AI-enabled organizations.

Why are traditional Agile frameworks no longer sufficient for the AI era?

Traditional Agile frameworks remain incredibly valuable. What has changed is the environment they operate within.

For years, organizations were constrained by how quickly they could design, build, and deliver software. AI dramatically changes that equation. As one customer recently told us, they’re no longer limited by how fast they can build things; they’re limited by how quickly they can determine whether those solutions create meaningful value, are safe to deploy, and align with business objectives.

The bottleneck has shifted from production to judgment. From outputs to outcomes.

Organizations now need operating models that help them continuously evaluate value, govern AI responsibly, and coordinate work between people and AI. That’s an evolution beyond traditional Agile practices, not a replacement for them.

What are the biggest mistakes enterprises make when trying to scale AI initiatives?

The first mistake is treating AI as a tooling problem. Most enterprises already have access to more AI capabilities than they’re effectively using.

The second is layering AI onto existing workflows without questioning whether those workflows should exist in the first place. Automating inefficient processes rarely produces transformation.

A third mistake is treating AI as an isolated innovation initiative instead of an organizational capability. Sustainable AI adoption requires changes to strategy, governance, product development, operating practices, and leadership, not simply deploying another technology platform.

Organizations that see the greatest impact redesign how work happens. AI becomes part of the operating model rather than another tool employees are expected to use.

What gap does the AI Value Architect fill, and how does it change team dynamics?

AI dramatically lowers the cost of creating solutions, but it doesn’t lower the cost of solving the wrong problem.

The AI Value Architect exists to help organizations uncover opportunities for leveraging AI while ensuring they stay focused on business outcomes. This role helps balance customer value, technical feasibility, cost, governance, legal considerations, and ethical responsibility before organizations invest in scaling AI-powered capabilities.

Rather than becoming another approval layer, the AI Value Architect role creates alignment between business leaders, product teams, and technical organizations so decisions about AI are both faster and better informed. It provides a practical mechanism for balancing innovation with responsibility while keeping measurable business value at the center of every AI investment.

How do you balance innovation speed with responsible AI oversight?

Governance shouldn’t be something that happens after innovation. It should be built into the operating model so innovation can happen safely and continuously.

Large enterprises have always needed to balance speed with accountability. AI raises the importance of that balance because AI systems introduce new considerations around data quality, transparency, bias, compliance, and organizational responsibility.

AI-Native SAFe makes governance explicit rather than implicit. It incorporates governance into portfolio decision-making, elevates data management as a first-class concern, keeps the human in the loop, and introduces role clarity around AI accountability. When governance is embedded into how organizations work, it becomes an enabler of innovation rather than a constraint.

How will enterprise team structures change over the next five years?

We see four major shifts emerging across high-performing organizations.

First, people increasingly focus on defining intent and desired outcomes rather than executing every task manually.

Second, learning and experimentation cycles become dramatically faster as AI accelerates exploration and feedback.

Third, innovation becomes more distributed because AI lowers the cost of experimentation across the enterprise.

Finally, teams become increasingly AI-augmented, with AI acting as an active participant in delivery rather than simply another productivity tool.

In the future, AI-Native teams will likely be defined less by headcount and more by how effectively they orchestrate collaboration between human expertise and AI capability.

Why is data readiness still such a major obstacle, and what should organizations prioritize?

Data remains one of the biggest barriers to realizing enterprise-scale AI because AI is only as effective as the information it’s built on. Fragmented data, inconsistent governance, and workflows that weren’t designed for an AI-native environment continue to keep many organizations stuck in pilot mode rather than scaling AI across the business.

Research from the Return on AI Institute found that 55% of global executives cite the lack of AI-ready data as the single biggest inhibitor to realizing value from AI. That reinforces what we’re seeing in practice: before organizations invest heavily in new AI applications, they need to treat data as a managed strategic asset, with clear ownership, governance, quality standards, and trust. AI scales whatever information it’s given, whether it’s good or bad.

That’s why AI-Native SAFe incorporates curated data management as a structural element of the operating model rather than treating it as a prerequisite or someone else’s responsibility. Data readiness isn’t simply an IT challenge; it’s a leadership and operating model challenge that directly impacts an organization’s ability to generate business value from AI.

How does AI-Native SAFe help organizations shift from measuring outputs to measuring outcomes?

AI makes it dramatically easier to generate output. As execution becomes faster and less expensive, output alone becomes a much weaker measure of progress. The real competitive advantage shifts to measuring whether those outputs create meaningful business outcomes.

AI-Native SAFe is built around an outcomes-driven product development cycle. It provides the roles, practices, and governance needed to continuously connect AI investments to measurable business value.

The same research from the Return on AI Institute supports this shift: organizations that regularly measure and communicate the value of AI at the executive and board level report significantly stronger returns on their AI investments. AI-Native SAFe helps make that outcome-based discipline a continuous part of how the enterprise operates.

What factors should organizations consider when deciding whether to transition to AI-Native SAFe?

Core SAFe remains the right operating model for many organizations. AI-Native SAFe is not intended to replace it; it extends it for organizations whose operating models are increasingly shaped by AI.

Organizations should consider where AI is creating meaningful opportunities to redesign work rather than simply automate existing activities. They should evaluate whether their governance, leadership practices, and data capabilities are ready to support AI at enterprise scale.

Most importantly, this is not an all-or-nothing decision. AI-Native SAFe includes adoption guidance so that organizations can evolve incrementally, adopting AI-Native SAFe practices.

Every organization starts from a different place, and AI-Native SAFe is designed to support that evolution.

What does a truly AI-native enterprise look like in 2030?

By 2030, the highest-performing organizations won’t think of AI as a technology problem but as a business opportunity. It will become integral to how they do work, remain competitive, and deliver value to their customers.

Human leaders will increasingly define intent, make strategic decisions, exercise judgment, and remain accountable for outcomes. AI agents will perform more of the analysis, coordination, and execution that previously required significant human effort.

What will separate industry leaders won’t be access to AI, everyone will have access to powerful AI capabilities. The differentiator will be the quality of their operating model: how effectively they orchestrate people, AI, data, governance, and continuous learning to deliver measurable business outcomes at scale.

Thank you for the great interview, readers who wish to learn more should visit Scaled Agile.