At IMPA London 2026, Closelink founder and CEO Philippe asked the room to picture an ordinary working day in 2030. His answer involved less chasing, more thinking, and a new job title.

It's a Wednesday morning in September 2030. You no longer have "buyer" or "purchaser" on your business card. You're a marine supply manager, and your job is tactical, not operational. You work for a mid-sized owner with a mixed fleet of 30 vessels, and you manage lubricant supply for all of them on your own.

That was the scenario Philippe used to open his keynote, Beyond Today: The Future of Artificial Intelligence at IMPA London 2026. Here's how he described the day, with the details adapted for the people who buy lubricants.

08:00: The report that needs nothing from you

You start with the weekly operations report. It covers every vessel's remaining stock on board, its consumption trend, upcoming port calls, and every open enquiry and order. Quotes have been collected and compared, and orders placed within the rules your company has approved.

The report says no action is needed from you this week.

"You actually just spent two minutes on all the operational orders and enquiries that are happening that week," Philippe said.

08:15: A forecast, not a surprise

Next is your rolling 12-month live forecast. It flags two vessels where lubricant costs are heading up: one by 0.4% and one by 1.1%.

You're the expert, so you check the data before acting. The 0.4% turns out to be a pattern of small top-ups at expensive ports, which you can fix yourself by moving the next order to a cheaper port in the rotation. The technical and finance teams are updated automatically.

The 1.1% needs more thought. The system has already set out options to reduce it. One is a supplier you haven't used before, already checked against your company's compliance rules and the approvals your engines need. Half an hour later you've booked an introductory meeting to understand what they could save you, and whether the quality holds up.

10:00: The cheapest oil isn't always the cheapest

It's nearly month-end, so the performance briefing arrives. Spend on one lubricant category is under target. Normally that's good news.

But the system doesn't only look at procurement data. It also sees the technical side: oil analysis results, maintenance records and the running costs of the equipment that oil protects. It has found a correlation. As spend on that category went down, the cost of keeping the equipment running went up.

This link between departments is where AI earns its keep. The saving in one budget line was showing up as an overspend somewhere else in the business. You pull together comparable cases and supporting records, check that the pattern is real, and set up a meeting with the technical superintendent. You arrive with evidence gathered in minutes, not weeks.

11:30: Seeing risk before it arrives

The system has picked up a strong chance that some of your vessels will change trading pattern in the next two to four months. It isn't in the schedule yet, and chartering may not know. The signal comes from changes in where cargo is moving.

The problem is that one of your key grades has thin availability in the region those vessels are likely to head for. You review the options in front of you: stem more before the change, qualify an approved alternative grade, or arrange supply ahead of time with a regional supplier. You compare the numbers, choose the option with the best cost-to-risk ratio, send the requirement out, and go to lunch.

14:00: A supplier who has done the homework

One of your lubricant suppliers is in town and has asked to meet. They want to show you a new AI model for production and regional stocking. It uses the demand data your company has agreed to share. Seeing your needs early lets them plan production better, position stock in the right ports, and cut the emergency deliveries that both sides would rather avoid.

None of this is new to you when you walk in. Your system has already tested the proposal against your fleet's expected requirements, so you know it's likely to work. The meeting isn't "what do you have and how could it help me?" You go straight to negotiating terms.

17:30: What changed, and what didn't

At the end of the day you notice a few things. You didn't spend a single minute managing enquiries, requisitions or orders. You met a supplier and negotiated. And yes, a captain did call to complain that the coffee on board tastes like bilge water.

"There are things that will drastically change," Philippe said, "and some might remain the same."

This is not science fiction

Philippe was clear that his 2030 isn't speculation. "Everything I said is not 'can that happen?' It will happen, I assure you. Probably some companies will be beyond that."

He also had a message for the buyers in the room: "Don't be afraid that you will lose your job." An AI agent that can handle a simple task quickly is a long way from replacing a team. In his story the person still does the most valuable work. They sense-check the data, decide what matters, build supplier relationships and negotiate. The admin is what goes.

What he did warn about is timing. AI has sped up the pace of change so much that "the later you come to the party, the higher the compounding cost of adapting to it" becomes.

How do you get there?

There are two two things that decide how fast a company reaches that Wednesday in 2030. One is how mature the technology is. The other is how ready your organisation is to use it. We look at each in the next two articles:

• Assist, Advise, Act, Coordinate: the four stages of AI in lubricant procurement, and how to tell where you are.
• Implementation is not adoption: why adding AI to old processes isn't enough.