Chief Technology Officer, ShipboardAI
James Calder spent a decade flying helicopters off the decks of warships before transitioning into defense AI and simulation engineering. He holds a Master's in Human Systems Integration and has led the design, deployment, and testing of AI systems for some of the most demanding operational environments on Earth.
At ShipboardAI, he oversees the technical architecture of every deployment — from GPU cluster design to model optimization for bandwidth-constrained maritime networks. He writes about what he knows: making AI work where the cloud cannot reach.
His operational background — 500+ hours of night-vision flight, combat deployments, and operational test pilot experience — shapes how he thinks about edge computing. When the link drops at 2am in heavy weather, abstract best-practices stop mattering. Only what is physically on the vessel still works.
A practical buyer's guide for on-vessel GPU hardware. When to pick H100, when L40S is the right sweet spot, when A100 makes sense, and the power and cooling realities nobody tells you before you sign the PO.
James Calder·April 11, 2026
Starlink promised 200 Mbps. You're getting 11. Here's what satellite connectivity actually looks like on a vessel — and how to build AI systems that work regardless.
James Calder·April 9, 2026
SpaceX is talking about gigabit-class speeds for Starlink in 2026. Here's why the throughput number is the wrong thing to fixate on if you care about AI that works on your vessel.
James Calder·April 9, 2026
Cloud-based chatbots fail the moment your ship leaves port. Here's what an on-vessel AI assistant actually looks like — and why your guests won't know the difference from shore-side service.
James Calder·April 1, 2026
Managing 2000+ crew across nationalities, certifications, and compliance requirements — without reliable connectivity. Why on-vessel AI is becoming essential for cruise line HR operations.
James Calder·March 25, 2026
When your vessel's systems show early signs of trouble, you need answers immediately — not after the data uploads to shore. Here's why on-vessel AI is changing how maritime maintenance works.
James Calder·March 18, 2026
Man overboard detection, restricted area monitoring, and bridge watchkeeping — running on hardware aboard the ship, responding in under a second, no cloud required.
James Calder·March 11, 2026
Your satellite link can't support cloud-based AI on a vessel. Here's why — and what actually works when you're 200 miles from shore.
James Calder·March 4, 2026