THE AI TRAVEL DECISIONING ENGINE

AI understands the request. Our engine makes the decision.

A proprietary travel decisioning layer between an open-ended enquiry and a commercially useful shortlist. Our intelligence meets your catalogue and business rules — without rebuilding the engine for every customer.

See it on your catalogue

Explore Discovery API

THE DECISION BOUNDARY

“An active holiday in October, about 10 nights, around £3,000 each.”

MyEscapePlan engine

AI intent + constraints + proprietary signals + ranking

Your catalogue and policies apply inside the decision

Ranked, explainable options

Conceptual workflow · not a live inventory search

Built, not just prompted

One proprietary engine. Configured for your business.

A general-purpose model can understand a traveller. Making a consistent travel decision requires specialist data, reliable constraints and a controlled ranking layer around it.

INPUT

Traveller request, written naturally

Dates, budget, party, interests and what can flex.

PROPRIETARY PLATFORM

MyEscapePlan Decisioning Engine

AI understanding

Interpret the request

Turn open-ended language into structured intent: dates, budget, party, interests and flexibility. The model understands the brief; it does not invent the final ranking.

Constraint integrity

Protect requirements

Separate firm requirements from preferences. Apply typed date, budget, duration and suitability constraints, with any permitted relaxation made explicit.

Proprietary travel intelligence

Travel-specific signals

Bring together synthetic tourism scores, derived destination intelligence, seasonal fit and learned signals from observed travel and verification outcomes.

Decisioning and ranking

Explainable results

Evaluate eligible destination and date options against the same versioned data, score multiple signals and return reasons for each recommendation.

The data construction, signal weights and ranking methods remain our IP.

YOUR CONFIGURATION

Your catalogue and commercial rules

Supply eligible trips or products, approved months, suitability rules and supplier policies. We agree the mapping and validate the rule behaviour during evaluation.

Catalogue scope

Eligibility and season

Commercial preferences

These controls govern the decision; they are not an after-the-fact filter.

OUTPUT

Recommendations your business can act on

Ranked, explainable options and structured enquiry context — delivered through a widget, an API or an advisor workflow.

An evidence loop, not a black box

Observed route, availability and verification outcomes can inform learned signals in future published data versions. Rankings do not retrain themselves after every click.

The data advantage

Proprietary knowledge that isn't in a prompt.

Some inputs are public. The value comes from the travel-specific signals we derive, the synthetic scores we maintain and the observations accumulated as the engine verifies real-world options.

SYNTHETIC INTELLIGENCE

Tourism scores

Derived destination and tourism-fit indicators help compare options at scale. We describe what the scores support — not the recipe used to construct them.

LEARNED SIGNALS

Evidence from real searches

Observed route coverage, affordability and verification outcomes inform destination and month signals when sufficient evidence exists.

REPEATABLE DECISIONS

Versioned travel intelligence

The engine evaluates options against published data versions, helping teams reproduce and inspect decisions without exposing internal scoring formulas.

A worked decision

An appealing option isn't always an eligible one.

A fictional specialist operator and an illustrative October request. Select a decision to see how the engine and catalogue policy combine.

THE ENQUIRY

“An active holiday in October, around ten nights, with about £3,000 per person.”

The real decision depends on the approved catalogue, constraints and any evidence that still needs checking.

WHY THIS DECISION?

A suitable October trip

The engine can recommend an operator-approved trip and explain the match. The specialist still confirms departure, total price and availability.

Illustrative only. No connected operator inventory, live availability or confirmed prices.

Built once. Configured for many workflows.

You shouldn't have to build the decisioning infrastructure.

A model alone is not a travel product. The engine brings interpretation, travel-specific intelligence and repeatable decisioning. Your business contributes what may be sold, to whom and under which rules.

Trip Finder

Help visitors explore operator-approved trips before an enquiry reaches the specialist.

Explore Trip Finder

Website widgets

Add guided travel discovery to an existing member or consumer journey.

Explore widgets

Discovery API

Bring pre-shopping travel decisioning into a product through a structured integration.

Explore API

Provider-backed live verification is a separate, optional step. Catalogue setup, delivery and integrations are scoped and tested before launch; this page does not promise instant self-service onboarding.

START WITH WHAT YOU SELL