Landr

The assistant that's already at the airport when you land.

Series A €25M Led by General Catalyst

01 — Problem

Travel inventory is a list. It was never an orchestration layer.

Travellers trust that when they land in a foreign city at 7am after a 12-hour flight, the world will have been arranged for them. It hasn't. The SIM doesn't work. The cab isn't booked. The expenses are in three inboxes. Everything existed to fix this — it just wasn't connected.

"Travellers give travel platforms an 89-point lower trust score than they give a good hotel concierge."

— 2025 Global Business Traveller Trust Survey · n = 14,200

02 — Market size

$1.61 trillion. Most of it controlled by platforms built for SERP, not agents.

Total travel market (2026)

$1.61T

Global gross bookings across all segments

Business travel (serviceable)

$420B

Frequent business travellers, 4+ trips/yr

Landr initial target (SOM)

$4.2B

Long-haul biz travellers · 47 markets · €47/trip × 90M trips

Bottom-up: 90M long-haul business trips per year in our 47 live markets. At €47 per completed trip and 89% completion rate, that's a €3.7B revenue opportunity at current pricing — before corporate account pricing is layered in. Corporate accounts (€300k/yr average) add a further channel that doesn't scale with trip count.

03 — Why now

Three things converged in 2025 that weren't true in 2023.

AI assistants became distribution channels

Claude, Gemini, and Copilot now handle millions of travel queries per day. They don't link to a SERP — they execute. Any inventory that isn't structured for agent querying is invisible to this distribution layer. Booking.com's inventory is a web page. Landr's is an API the assistants can call and act on.

MCP made real-time service orchestration viable

The Model Context Protocol standardised how agents interact with external services. An agent can now provision an eSIM, book a cab, and push an expense entry in a single workflow — reliably, at scale, without custom integrations. This didn't exist at commercial quality before late 2024.

Business travellers crossed an expectation threshold

The 2025 Business Traveller Trust Survey found 67% of frequent business travellers now expect AI to handle arrival logistics — up from 31% in 2023. Expectation has outrun the product. Landr is the product.

04 — The product

Six agents. One arrival experience.

Agent execution · CX234 HKG → AMS · T-90min

Calendar agent Trip detected · 72hr ahead

Memory agent Preferences loaded

Arrival agent eSIM + cab deployed

Reroute agent Monitoring · standby

Expense agent Concur sync queued

Discovery agent Local recs · post-arrival

Trip status Delivered ✓ · €47

The product isn't a booking app. It's an orchestration layer that wraps the arrival — connecting the flight, the SIM, the ground transport, the hotel confirmation, and the expense system into a single agent-managed workflow.

Inventory is structured as machine-readable, agent-queryable data from ingestion. That means Claude, Gemini, and Copilot can call Landr's API directly when a traveller asks their assistant to sort a trip — no SERP, no click, no separate app open.

89% of trips close without a human

The 11% that require human involvement are genuinely ambiguous. They go to a human in under 4 minutes.

05 — Traction

Numbers we can say in a room without flinching.

Paying travellers

100k

+34% MoM last quarter

ARR (run rate)

€15M

€47 avg revenue / trip

6-month retention

78%

vs 31% industry average

Corporate accounts

50

€300k avg contract value

The unusual metric: 89% of trips completed end-to-end without a human touching them. That number is the product. Everything else — retention, NPS (94), GMV — is a downstream consequence of getting that number right.

06 — Moat

Why Booking.com can't copy this.

Three structural reasons. Not feature-level. Not things that can be copied with a sprint.

1. The inventory architecture is incompatible with SERP

Booking.com's entire inventory layer — pricing, ranking, content — is optimised for human browsing and search engine visibility. Making it agent-queryable isn't a feature addition; it's a fundamental rearchitecture. They would have to rebuild the product for a distribution model that currently generates zero of their revenue. Incumbents don't do that until the alternative distribution is already cannibalising them. By then, Landr has 3 years of preference data per traveller that cannot be recreated.

2. The preference memory compounds across every trip

Every trip Landr handles makes the Memory agent's preference profile more accurate. After 5 trips, Landr knows the traveller's seat preference, hotel requirements, dietary rules, and tolerance for layovers better than the traveller could articulate. This data is not publicly available. It cannot be licensed. It accrues to Landr exclusively — and it is the reason the 89% completion rate will continue improving while a new entrant starts at zero.

3. Booking.com's organisational structure generates the problem Landr solves

A business with 27,000 employees, a SEO team, a content ops team, a category management team, and an affiliate distribution network is structurally incapable of moving fast enough to become an agentic arrival layer. Every team that would need to shrink for Booking.com to compete has a budget, a director, and a headcount review cycle. The company's operating model is the moat — it prevents them from becoming Landr.

07 — Team

Built by people who've seen the problem from the inside.

Alex Mertens

CEO & Co-founder

Previously: VP Product at Travix (acquired by China International Travel Service). Ran the corporate booking platform for 8 years. Left when he realised the product was a list, not an orchestrator.

Yara Osei

CTO & Co-founder

Previously: Staff Engineer at Airbnb (infrastructure + agentic tooling). Built the internal agent orchestration layer that handles Airbnb's real-time pricing. Knows what it takes to run agents reliably at scale.

Jin Park

CPO & Co-founder

Previously: Director of Product at Navan (formerly TripActions). Owned the expense and itinerary management products for the company's largest enterprise accounts. The Concur integration was his idea.

Sara Dubois

CCO & Co-founder

Previously: Head of Corporate Partnerships at Amex GBT. Closed €200M+ in enterprise travel contracts. Built the playbook for landing Fortune 500 corporate travel accounts — now running it for Landr.

08 — The ask

€25M Series A. 18 months to four milestones.

General Catalyst leads. We've run the enterprise + AI thesis together for two years. They know this market. We want a lead who can open doors with corporate travel buyers at scale — General Catalyst's enterprise portfolio is that door.

Capital allocation

Engineering (agent + infra) €12M

Corporate GTM (Field Ops scale) €8M

Distribution partnerships (AI assistants) €5M

Milestone 1 · Month 6

100k paying travellers · €15M ARR · 78% 6-month retention

Milestone 2 · Month 9

50 corporate accounts live · €300k avg contract value

Milestone 3 · Month 12

Landr inventory queryable natively by Claude, Gemini, and Copilot

Milestone 4 · Month 18

89% of trips end-to-end without human touch at full traffic volume

Appendix — reference facts

The numbers behind the narrative.

Global travel market

$1.61T

Total global travel gross bookings, 2026. Source: WTTC Annual Travel & Tourism Report 2026.

Trust gap

89 points

Gap in trust score between a good hotel concierge and the leading online travel platform. Source: 2025 Global Business Traveller Trust Survey, n=14,200.

AI-native VC share

45%

Share of global VC dollars in 2025 that went to AI-native startups — companies where AI is the product, not a feature. Source: Atomico State of European Tech 2025.

Incumbent reference

~$15B

Booking Holdings current market cap. 27,000+ employees. Revenue per employee approximately €14k. Landr runs at 25× that ratio on 42 people. Source: public filings.