It takes a lot more than a GPU for AI to recognize a hot dog.

At Megaport Connect 2026, David Boland, Wasabi’s Vice President of Cloud Strategy, walked us through what really happens behind an AI app like this one.

This clip only shows a small part of it.

In the full session, David follows a single photo all the way through the AI pipeline:

• Collecting it alongside billions of other images
• Cleaning it up so the model only sees the food
• Training and testing the model
• Moving it to a system that can answer questions
• Telling you whether it’s a hot dog (or not)

And every one of those steps runs across the network.

(Plus, the data keeps growing the whole way through.)

If you want to see the whole journey, the full session is on our channel.