About Mat

Mat Orrego is CEO and co-founder of Cornerstone, a travel data platform he has built over 30-plus years after moving from financial services into one of the world's most transactional industries. Cornerstone automates the most fragmented post-booking workflows for travel management companies, corporate buyers, and OTAs like Expedia — processing hundreds of thousands of reservations daily through deterministic business rules at scale.

What we talked about

Post-booking as the hard problem — Why Cornerstone lives in the messy middle after purchase — schedule changes, policy compliance, reconciliation, and real-time messaging across the reservation lifecycle.

Business rules at a million transactions a day — How parameterized rule engines enforce policy, trigger Slack alerts, and replace month-end reports with instant reactions.

Legacy infrastructure meets agentic search — Why combinatorial AI shopping queries are crushing old distribution systems built for human look-to-book ratios of one to two percent.

Making travel data agent-ready — Schema descriptions, master data management, and intent tagging so LLMs can answer forecasting questions humans used to need analysts for.

Intermediaries under pressure — AI threatens labor-heavy agency roles, but customer relationships, bundled experiences, and post-booking support still defend the middle layer.

Booking windows flattening — Shorter advance purchase cycles on business and leisure travel — and what Cornerstone's Travel Index data shows about impromptu buying.

Sharpest moments

On what Cornerstone does:

take the most complex, most fragmented areas of the travel industry and kind of figure out how to automate them

On the old way of working:

it was always an after action analysis.

On daily scale:

we're processing anywhere from 800,000 to over a million transactions on a daily basis

On who sits behind every reservation:

Every one of those reservations is not a bot. It's an actual person.

On legacy distribution strain:

look to book ratios... one to 2 % actually convert to an actual booking.

On AI's threat to intermediaries:

AI is competing for human labor costs

On shifting booking behavior:

you don't have those big year-long booking windows like you used to.

Key takeaways

• Cornerstone's core is real-time reservation lifecycle automation — not front-end search — serving TMCs, corporate buyers, and major OTAs.
• Business rules engines already enforce policy at transaction speed; agentic layers are emerging on top, not replacing the foundation overnight.
• AI agent shopping floods legacy airline and hotel systems with queries never designed for combinatorial load — distribution costs are rising.
• Travel data for LLMs needs explicit schema, intent tags, and master data — dashboards built for humans assume context agents do not have.
• Intermediaries lose defensibility on pure search, but bundled itineraries, customer trust, and post-booking support still matter.
• Travel agencies still employ large front-line and support teams — consumers research with AI, then call a human to finalize details.
• Booking windows are compressing: less year-ahead planning, more last-minute and bleisure-blended corporate travel.