Pallavi Mahajan is the Chief Technology and AI Officer at Nokia.
Somewhere in your organization today, a team of three is doing what used to require thirty. They’re not on your radar and they didn’t ask for a budget increase or run it through the architecture review board. They quietly rebuilt a workflow your company has been executing the same way for years and it works better, costs less and took them six weeks.
You’ll hear about it during a quarterly review. Someone will call it a “great example of AI adoption.” There will be a slide.
And you will have completely missed what it’s telling you.
I’ve spent my career building and scaling technology through multiple waves of disruption, from software‑defined networks to high‑performance supercomputing and enterprise AI. Every cycle taught me the same lesson: Technology is rarely the limiting factor. The real constraint is whether senior executives can update how work gets done once the operating model shifts beneath them. We are in that moment again, and this wave is moving faster.
The Five Internal Conversations That Are Stuck
Every enterprise technology company is grappling with the same questions. The honest answers are costly, and silence erodes trust.
1. How do you manage a two‑speed workforce when some teams move three to five times faster than others, but performance expectations haven’t caught up?
2. What do you do with roles that are no longer backfilled as AI absorbs more work, especially when silence around those decisions erodes trust?
3. Who owns quality when output is increasingly AI‑generated and accountability becomes political?
4. How do you price and contract when AI has already decoupled delivery from effort and your sales model hasn’t caught up?
5. Is AI enabling something genuinely new, or simply optimizing the old?
These questions force choices about power, roles and redesign. They force leaders to update beliefs their credibility was built on. That is the transformation; everything else is tooling.
The Signal Your Metrics Don’t Capture
Most enterprise technology companies measure AI adoption through familiar metrics: tools deployed, licenses consumed, productivity gained. These numbers matter, but they miss the structural shift underneath them: the divergence between the company you’re running and the one emerging inside it.
The gap compounds quietly, then snaps. AI‑native teams surge; the rest stall.
The danger isn’t that AI is moving too fast for your organization to follow. It’s that parts of your organization have already moved and the rest of the structure hasn’t caught up. You’re not behind the technology; you’re behind your own people.
What’s Actually Happening On Your Engineering Floors
The headline: AI writes the code, velocity doubles, teams get smaller. True, but it only captures the output change. The harder shift is the identity change underneath it.
Your senior engineers are no longer writing code in meaningful volume. They are specifying outcomes, reviewing agent-generated work at a systems level and catching architectural drift before it becomes technical debt. The 80/20 split between implementation and architecture has inverted, redefining what expertise looks like.
Engineers who haven’t made this transition are operating in a mode that made sense 18 months ago but now creates a performance gap that is uncomfortable to name and politically difficult to address. Your frameworks, compensation bands and promotion criteria weren’t built for it.
The fix requires rethinking how people are evaluated: rewarding judgment, outcome ownership and the ability to direct AI effectively toward business results.
The Product Function: Being Rewritten
Your product managers built their craft in a world where engineering time was scarce. The skills that defined the role, from customer understanding and feature prioritization to roadmap negotiation, were shaped by constrained capacity. Every framework they’ve learned assumes execution is expensive.
AI is collapsing those assumptions. As execution becomes dramatically cheaper, scarcity-based frameworks distort product judgment. The job shifts from fighting for engineering time to deciding which questions are worth answering. The strongest product managers move upstream to customer truth, market dynamics and strategic framing.
This is an organizational design crisis most companies haven’t acknowledged. The honest move is to redesign the coordination function openly. Don’t wait for it to self-resolve.
The Middle Layer: Where Transformation Stalls
In my 20 years of enterprise transformations, I’ve watched leaders get the top right, teams adapt at the bottom and the organizational middle stall the change. The middle layer existed to manage scarcity: engineering capacity, context gaps and decision bottlenecks.
AI is removing much of that scarcity. In AI-native teams, what once created control now creates drag. The most difficult question is not whether this layer can be upskilled, but whether the coordination function still needs to exist in the same form.
The leaders who answer that question clearly and redesign accordingly will likely build organizations that compound their advantage.
In practice, that means following how work flows, collapsing layers that no longer earn their place and reorganizing into small teams that own outcomes end-to-end.
The same operating model gap eventually shows outside the organization too. Once AI changes how work is delivered internally, it also changes how value should be priced, sold and contracted externally.
Delivery: Quietly Repricing Itself
Delivery economics changed faster than sales methodologies and pricing models could absorb. Sales still leads with effort, team size and timeline. Delivery leads with outcomes. AI-compressed execution has decoupled outcomes from effort.
Companies that realign before their customers close that gap will own the premium end of the market. The rest compete on price. No high-cost organization wins that race.
The Compounding Gap
The organizations that will likely define enterprise technology aren’t the ones with the best AI strategy documents. They are the ones having hard conversations with enough honesty to act.
Real AI leaders are compounding advantage through workflow redesign and agentic execution, while everyone else is still talking about productivity.
The signal already inside your organization? The team of three doing what used to require thirty.
The question is, what will you do with it?
Start with whichever of the five questions you most want to avoid. That’s where the transformation begins.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
