York Lu is betting that the next customer for hotel inventory is not a person at all - it is an AI agent.

In this episode of the Everything AI in Travel podcast, I speak with York Lu, Head of AI Partnership at Dida and co-founder of RollingGo. York was one of the first three people in Trip.com’s US office, spent three years as a VC investment manager in Shanghai, and is now building what he calls B2A - business to AI agents - on top of one of the biggest hotel supply businesses most people in the West have never heard of.

Connect with York on LinkedIn

Got a travel problem AI might solve? Find the AI you need: tell me what’s frustrating you and I’ll personally introduce you to someone who can help.

Dida is a B2B travel platform that York says is already the third-largest in the world - bigger than the listed player TBO on GMV, by his numbers - with over 2,000 staff, 19 offices across 40-plus countries and around 100,000 directly contracted hotels. RollingGo is the new Dida-invested entity taking that supply and packaging it for machines: net rates through an API, product interfaces designed for agents rather than humans.

We dig into:

• Why Dida’s own AI journey started with building agents for hotels - and why they walked away from it (efficiency gains, but no new bookings)

• The moment the board sent York and one colleague to Hangzhou, home of DeepSeek and Kimi, with no plan and no office: “find your way”

• What B2A actually means: selling hotels and flights to personal AI agents, not to people

• Why RollingGo stayed deliberately quiet while signing up most of China’s top AI companies - and why 1 November changes that

• York’s story of a Google closed-loop booking announcement that knocked 10-20% off Booking.com, Expedia and Trip.com shares in a day - and the reversal a month later

• The Hyatt contract that landed because York’s professor happened to be the VP of Revenue - and what that says about travel as a relationships business

• How net rates through the API leave room for the agent platforms to take their margin - the “toll booth” model

• Why he thinks OTAs and travel agencies gradually get replaced, and what that means for anyone selling travel

One of the most interesting ideas from this conversation is that the AI platforms and the supply owners need each other more than either admits. The AI companies are trying to close the loop inside their own apps and hardware, while the big B2C players fight to keep the customer. Meanwhile Dida quietly contracted 100,000 hotels directly, the hard way, on the ground.

If York is right, the winners on the supply side are the ones whose inventory an agent can actually read, price and book - and that is a plumbing question, not a marketing one.

The battle for who owns the booking is starting now, and it is being fought in the APIs.

Enjoyed it? Subscribe so you don’t miss the next one, and share this with someone in travel who’d get something from it.

Want the weekly version? The Everything AI in Travel newsletter: https://everythingaiintravel.beehiiv.com/

#AIinTravel #B2A #HotelDistribution

Hear the episode here.

In 2026 the podcast is now found on Spotify, Apple Podcasts and now also YouTube (and if you hit subscribe there, you get it before everyone else). You can also catch up on the whole of Series 1 now also fully available in these channels.

This could be your brand’s podcast (and newsletter).

We are currently seeking submissions from leading AI companies in travel who believe that all businesses should be media businesses.

Following the example of dojo.live and Encora we offer the opportunities to communicate your brand’s USP’s to a B2B travel audience from the leading Travel + AI resource on a bi-weekly, done-for-you basis for all of 2027 (and beyond).

Contact Tony if interested to learn more and see if we are a good fit for one another.