Episode Overview
In this episode, Jennifer Arnold (Head of Public Sector Network's Industry Research and Advisory Services) is joined by Steve Lewis (EY Global Infrastructure Technology Leader and Head of Infrastructure Advisory, Hong Kong) to unpack a hard truth: despite major investment in digital, infrastructure organisations are still struggling to close the productivity gap. The conversation explores why agentic AI offers a different way to think about productivity, risk, and control—by giving teams better insight earlier, so they can act with confidence in an increasingly complex delivery environment.
Rather than defaulting to “digitise everything,” Steve challenges leaders to focus on outcomes, ROI, and targeted transformation—so AI and digital initiatives actually make organisations fit for purpose for the milestones ahead.
What You’ll Learn
1) Why Digital Investment Hasn’t Closed the Productivity Gap
Why tools alone don’t automatically improve delivery, even with significant spend.
2) What Agentic AI Changes (Productivity, Risk, Control)
How agentic AI can support better decisions earlier in the lifecycle in complex infrastructure environments.
3) Don’t Digitise Everything: Be Targeted
Why some workflows don’t need an overhaul—and some are simply obsolete and should be removed, not digitised.
4) Start with ROI and Outcomes
How to define what you’re trying to achieve, avoid “doing AI somewhere,” and set clear measures of success.
5) Identify the Real Pain Points (and the Real Reasons)
Why leaders need to dig deeper—some processes exist for reasons that aren’t written down, but are critical to the business.
6) How Transformation Programmes Succeed or Fail on People
Why consultation and communication matter, and how different people need different reasons to buy into change.
7) Avoiding Transformation Fatigue
How to prevent agentic AI from becoming “just another initiative” by anchoring every effort to value and outcomes.
Key Takeaways
- Agentic AI offers a new lens on productivity by improving insight and confidence earlier in delivery
- “Digitise everything” creates fatigue—target the right workflows based on outcomes and ROI
- Some processes should be transformed; some should be removed, not digitised
- Define ambition and success measures up front—don’t do AI for the sake of it
- Transformation requires detailed understanding of the work, including undocumented “tribal knowledge”
- People are the deciding factor: consult the doers and tailor communication to what motivates them
Why You Should Listen
This episode is for public sector infrastructure leaders, digital and transformation teams, project and portfolio leaders, and advisory professionals who want a practical perspective on why digital programmes stall—and how agentic AI can support real productivity gains when it’s applied deliberately, measured properly, and implemented with the workforce in mind.
Help your peers
Share what you've learned with fellow public servants