The Standardization Problem That Shipping Said Was Impossible Got Solved in Two Years

Maritime went from file cabinets to asking for AI in under a decade

Dylan Mace, VP of Sales at Marcura, has spent the better part of ten years in energy and maritime. He started at Kinder Morgan, a midstream oil and gas company, at a time when the industry ran on in-house built systems, Excel spreadsheets, file cabinets and handwritten notes. He describes it as very old school, with every company attached to its own way of working and plenty of resistance to change.

That resistance shaped how he sold. As late as 2017 or 2018 he was pitching cloud SaaS to large shipping, trading and energy companies still running on-premise systems, and had to argue why moving off servers maintained by internal IT was worth it. COVID changed that overnight, with customers suddenly knocking on the door to move to the cloud. Mace says the bigger shift has come in the past two years, with a clear uptick in appetite for new solutions, and for AI in particular.

Knowing the customer's workflow is the sales advantage

Mace did not come up through the usual BDR to AE path. He landed in internal audit at Kinder Morgan through LSU's formal resume drop, moved into commercial analytics at a terminals company where he helped salespeople price storage deals, and then joined Veson Nautical in Houston as a consultant. He spent years sitting next to charterers and operators, drawing their workflows on whiteboards and mapping how their business had to work in the system.

When he moved to London and into a sales role, that background let him build rapport and empathy, and judge honestly whether a solution would solve the problem in front of him. Kucker, who calls himself a jack of all trades across 20 years in shipping, notes the other benefit of staying in one industry: a deep network. Mace agrees. Hiring managers constantly weigh a strong SaaS seller who doesn't know shipping against a strong shipping person who may not adapt to a technology company. Staying put means conferences full of familiar faces, and the operator he once trained may now be a global head of operations. His advice is to treat the CEO the same way you would the janitor, because you never know where people will end up.

AI dissolved the standardization problem shipping called unsolvable

The core of the conversation is a problem Mace heard at conference after conference two years ago. Bulk shipping, meaning non-containerized cargo, is not standard. A dry bulk statement of facts, the time log of everything that happens in port, looks very different from a tanker's. Even a single event like notice of readiness tendered gets written as "N-O-R tendered", "N.O.R tendered" or spelled out in full, depending on which port agent produced the document. The industry consensus was that nothing would improve until everyone agreed to call things the same thing.

A big problem we had in shipping was data standardization. So now with AI, that's kind of been dissolved.

Mace explains that AI can now hold a library of a million variations and map each one to a standard term. Even so, the best statement of facts parser on the market might correctly capture 90 of 100 activities. That is why Marcura builds human in the loop into its products: AI extracts the data, then a person validates it. By the time a demurrage claim reaches a claims analyst or operator, two sets of eyes have already been on it.

Kucker frames this as the difference between older machine learning, which needed something like 100,000 labeled examples, and LLMs, which can make the judgment without that training set. He adds that how much the remaining 10 percent matters depends on the stakes.

If AI is 10 % wrong and coaching a rep on how to close a deal, that's perfectly fine.

In financial and contractual work, where the numbers can get big, the human check is what makes 90 percent accuracy usable.

The AI pitch changes with the persona

Nobody in maritime tells Mace they don't want AI, but there are clear leaders and followers, so he sells by scenario. For a ship owner whose operators handle demurrage claims on top of everything else, the pitch is that Marcura's tool and services take that work off their desk so they can focus on getting the ship in on time and keeping a tight P&L.

For a dedicated claims analyst, the message is different. Mace walks through the hours or days spent chasing port agents for documents, then retyping a PDF into Excel on a second screen.

So I've spent hours or days of work before I can even start doing the strategic part of my job, which is actually negotiating and settling with my counterpart.

Positioning AI as removing low-value admin rather than replacing the analyst matters, because fear of job loss is real in this industry. Mace also sees system fatigue after years of heavy SaaS adoption. Internally, that means pressing product on the questions customers will ask before anything goes to market: how does this make their life better?

Freed-up time turns into a new data set

Kucker points out that "this will free you up for strategic work" is a pitch buyers have heard for years and often don't believe. Mace says skepticism is lower now, but only when the value is shown. His example is the operational data sitting in statements of facts and claims, which he calls a treasure chest people have never really had access to. Combine it with AI's ability to read complex charter parties, and an operator can go back to the charterer with evidence: every time we fix with this counterparty, we lose this much time in port. That lets the charterer negotiate better charter parties that save or make money.

Kucker says he had not thought of AI as creating a new data set, and that this may be where the replacement work comes from: not less work, but a better category of it. He adds a shipbuilding example, where flagging nearby maintenance issues saves inspectors 15-minute walks from one end of a vessel to the other. When Kucker likens the admin to brushing your teeth, Mace jokes that he has wanted a hands-free toothbrush for years so he could spend those minutes on something else.

Salespeople stay, and relationships carry more weight

Inside Marcura, dedicated AI squads run R&D, while Mace's sales team starts with something simple. An AI note taker lets him sit back, listen and ask better questions instead of racing to write everything down. When launching products, he runs call transcripts from the Singapore, US and EMEA teams through AI to find common pain points and see which pitch resonates in which region. Kucker draws the parallel: conversations are the activity revenue teams generate, the way moving goods is the activity shipping generates, and they should be captured permanently.

Looking three years out, Mace expects the SDR and BDR function to change because emails can be automated, but he is firm that salespeople will still exist. C-suite buyers will still want a person to explain why one solution beats another.

I think that human-human to relationship is actually becoming more important as we advanced in technology.

He notes that some people now add deliberate typos to emails so they read as human-written. Kucker argues AI makes relationships more authentic, not less. If he speaks to someone twice a year, a record of past conversations means he remembers their kids and can bring a useful idea next time. In sales, knowing what is in a prospect's 10-K becomes table stakes, so the call can go straight to what is actually happening on the ground.