AI, Demand & the Expertise Crisis

By Allison Buenemann, Senior Product Marketing Manager at UptimeAI 

With CERAWeek 2026 in the rearview, the UptimeAI team shares our reflections, the challenges the energy industry faces, and how AI Reasoning Agents are making operational excellence not just aspirational, but attainable.

“In the next 4-5 years, I expect to lose 30-40% of my engineering workforce.”  

This isn’t the first time we’ve heard something like this, but it is the first time the CEO of a major North American power generator has brought it up in 1:1 conversation with a tone of genuine concern. In a single sentence, they summed up the criticality of the conversations that happened in Houston, and the urgency that’s being felt far beyond the conference walls.

The UptimeAI team spent the last week meeting with energy industry executives, joining keynotes and panel discussions, and launching a new line of AI Reasoning Agents to address the highest value operational challenges these leaders face. Every conversation, whether on the plenary stage or at the infamous Hilton Americas lobby bar, eventually reached the workforce, the growing expertise shortage, and the pitfalls of the last decade of digital transformation. Here’s what we heard, and how it affects the future of the industry.  

We’ve shifted from “Does AI work?” to “Can AI scale?” 

“We’re past proof of concepts. The concepts have been proven. Now it’s proof of competency.”  
-Julien Debard, Databricks in conversation with Microsoft and Plains All American Pipeline 

For Plains All American, the secret to driving value from AI is simple: You have to say no to hundreds of really good ideas so that you can focus on ten excellent ones. For ExxonMobil, it’s about setting yourself up for success at scale. Rob Crane, Technology Scouting & Venture Manager, shared that they’ve driven over $500M in value through scaled AI deployments. The question isn’t whether AI delivers; it’s whether your organization is set up to capture it.

With last week’s launch of our inaugural AI Reasoning Agents for Operational Excellence, we’ve focused on creating scaleable agents designed to mimic the skillsets and reasoning capabilities of your plant experts to make the critical operational decisions that deliver bottom line impact.  

In the Innovation Agora, Pablo Peris of Microsoft said, “The AI you’re using today is the worst AI you’ll ever use.” This technology revolution is moving fast and pairing the right technology partners and the right business problems is a recipe to maintain momentum for your company.

Growing demand for energy and natural resources = High stakes  

Global supply uncertainty stemming from the war in the middle east has players in all regions, and all segments of the petroleum value chain, pushing rates and deferring planned outages. Data center energy demand is creating unprecedented demand in power generation. In this perfect storm of energy and resource demand, companies can’t afford any reliability hiccups.

What struck us most wasn’t the size of the challenge. It was the clarity amongst leaders about what’s required to rise to it. In a panel session on AI Potential: Hype v Reality, our CEO Jag Gattu postulated that the companies that will capture AI’s value in this environment aren’t necessarily the ones with the biggest AI budgets. Panelists from Energy Impact Partners and UC San Diego agreed. The winners will be the ones who focus on the business problem, then prove them at scale.  

In the energy industry, the reality for what AI can deliver in operational excellence has finally caught the hype. The unlock came from analyzing where decades of past digital innovation fell short. It was never in the ability to detect an issue was present. It was in the expert ability to interpret detected alerts, diagnose root causes, and recommend optimal solutions. With the launch of Root Cause Agent, we’ve bottled up that expert ability into a production-ready agent that scales without reliance on human domain expertise. 

The expertise crisis isn’t coming. It’s here. 

This thread was woven into every other topic at CERAWeek. With eighty percent fewer petroleum engineers entering the workforce and experienced operators retiring, leaders across the industry forecasted that they’ll lose up to half of their skilled workforce within the decade.

What’s walking out the door isn’t just knowledge and headcount. As Accenture explained in Monday’s AI Hub session, “It’s hunch.” It’s decades of hard-won operational judgement that can’t be easily retained or replaced, that keeps the ‘human’ at the front end of ‘human-AI decision making.’

UptimeAI Reasoning Agents are our answer to replicating the hunch. When we launched the agents on Wednesday in a standing-room-only Agora Pod session, the reaction of the room—the questions, the hallway conversations that followed—told us the industry is ready. ‘Human-AI’ decision making will inherently be limited by human expert decision capacity. UptimeAI Reasoning Agents embody the skills, logic, and cross-functional reasoning of plant experts to effectively scale the human element of it.  

Where we go from here 

The answer to the expertise crisis underpinning broader operational excellence won’t come from yesterday’s technology. It will come from thinking boldly, pushing boundaries, and keeping the metrics that matter top of mind.

As our own Senior Solution Consultant Cody Berra joked when asked what we’d be talking about at CERAWeek next year on the panel Where Agentic AI is Now & What Comes Next …

“We won’t be talking about anything; it’ll be the agents doing the talking.”  

Laughs? Yes. Healthy skepticism? Yes. But the stakes are higher than ever before. The urgency is higher than ever before. And the appetite for pushing the limits in one of the most historically conservative industries is higher than ever before.  

As with the technological advances of previous decades and centuries, AI reality will follow the hype. For energy, this presents tremendous opportunity. The window to act — before expertise walks out the door and poor operational excellence impacts demand — is narrowing. And UptimeAI Reasoning Agents couldn’t be coming at a more perfect time. We don’t know what next year will bring, but we’re excited to be there to find out!