Data Stars & Scars

Twenty years ago, Julien Debard was running a rig that cost a million dollars a day. At 23. Offshore in Nigeria. With his best data arriving by USB stick in someone’s pocket on a chopper. Now he’s a 2026 S&P Global Energy Awards finalist and the person a lot of energy leaders call to work out where AI actually earns its keep.

This episode is the throughline of his career: technology only creates value when it changes how the work gets done. Harj Chand and Emily Kucukalic get him on the hard questions; who holds the decision rights when AI is in the loop, why senior leaders still won’t ask their own data a question, and why “industrial AI” is its own beast when a wrong call shuts a safety valve rather than ruins your dinner.

• Julien’s field years running a rig that cost a million dollars a day offshore Nigeria at 23, when getting data to shore meant a USB stick on a helicopter.
• The retirement cliff facing energy, and why the old phone-a-mentor model is running out of mentors.
• Who actually holds the right to make a decision on site once AI is in the loop, and how you know the call is a good one.
• Why AI should augment people rather than replace them, and the case for solving small before shooting for the moon.
• The productivity problem energy hasn’t faced yet, with an ageing workforce and fewer active people funding more retirees.
• Why you should back the experienced person’s gut over the data when the two disagree.
• Why senior leaders still won’t ask their own data a question, and why it’s habit rather than fear of the tech.
• How the Strait of Hormuz closure separated the companies that were ready from the ones that weren’t.
• The two kinds of companies that win with AI, and the ways the rest get stuck.
• The valley of death between a single use case and a scalable platform, and why it bites harder in Australia.
• Why a proof of concept is really a proof of competency, and what that changes about how you start.
• What industrial AI actually means when a wrong call shuts a safety valve, and why AI needs the industries as much as they need it.
• The physics advantage legacy players are sitting on, and why good enough beats perfect.
• Supplier concentration, frontier labs that won’t all survive, and building an architecture you can pivot.
• Julien’s advice to leaders to dive in and not trust the vendors, and to newcomers to learn the domain first.
• What’s next with GREJU, the venture connecting executives who want value with the smaller players who can deliver it.