Travel & Tourism Data Analytics

NDC (New Distribution Capability) is the IATA XML standard, launched as Resolution 787 in 2012, that lets airlines distribute rich, dynamically priced offers (branded fares, ancillaries, bundles) through any channel. Traditional GDS distribution relies on decades-old EDIFACT messaging and filed fares, which cannot express modern retailing concepts. Most airlines run both in parallel today, which is exactly why a harmonization layer matters.

ONE Order is the IATA standard that collapses the PNR, e-ticket, and EMD into a single order record: one reference carrying the customer, the services purchased, payment, and fulfilment status through delivery and accounting. NDC covers how an offer is shopped and ordered; ONE Order is the living record of what was bought. They are two halves of the same retailing transformation.

AIDM (Airline Industry Data Model) is IATA’s canonical data model, from which both the NDC XML schemas and the Open Air JSON/OpenAPI specifications are generated. It matters because it demonstrates the pattern every travel IT estate needs internally: define meaning once in a governed model, then generate per-format, per-partner projections, instead of maintaining a web of bespoke mappings.

Shopping works: with a connected API or MCP interface, an agent can search and price real offers today. Buying is where it stops. Payment needs separate agentic-commerce protocols and mandates, a non-refundable purchase is precisely the kind of high-consequence action that should stay human-confirmed, and offers expire quickly, so the commitment step must always be re-priced live. Designing that governed flow (cached exploration, live human-gated commit) is a core part of what we do.

No. We work with airlines, OTAs, TMCs, hotels, car rental companies and marketplaces, airports, and travel-tech vendors. The airline standards stack is the deepest specialization, but the underlying disciplines (canonical modeling, caching, governed AI interfaces, revenue analytics) apply across the industry.

No, and we would not suggest it. We do not sell a retailing platform, an aggregator, or an order management system, and we have no reseller arrangements to steer you toward one. Our work is the layer between the systems you already chose: canonical data semantics, integration validation and testing, analytics instrumentation, and AI-agent interfaces. Every new platform in your estate increases the need for that independent layer, not the reverse.

Not yet, and we will not pretend otherwise: that space belongs to established retailing platform vendors, and several do it well. Our track record is in the disciplines around those platforms: canonical data modeling at very large scale, AI agents running in production for clients, machine learning and analytics delivery, and a working airline document-automation proof of concept for a European flag carrier. We have studied NDC, ONE Order, and AIDM deeply enough to publish detailed technical analysis of them, and we bring that standards literacy to the data, testing, and AI work that platform projects leave on the table.

Yes. An MCP interface for AI agents is typically a governed layer over your existing services: it exposes deterministic, documented capabilities with business definitions baked in, rather than raw endpoints. We design the contracts, the caching in front of expensive operations, and the guardrails that keep agents from committing to things they should not.

Every airline offer is computed by a real-time pricing engine, and every shopping request costs compute. AI shopping agents multiply request volumes by orders of magnitude while bookings stay flat, so serving agents economically requires caching normalized offer content with explicit freshness rules, and re-pricing live only at the moment of booking. Get the cached unit wrong and you either mis-price bookings or melt your offer engine.

An inventory of your booking and distribution systems (PSS/OBE, channel managers, aggregator APIs)

Sample payloads from the partner schemas and NDC versions you support

Booking, revenue, and channel-mix history for analytics work

Your current API/interface documentation, if any

The business definitions your teams already agree on (and the ones they don’t)

We offer two engagement models with transparent pricing.

We also offer project-based pricing for well-defined engagements.

Contact us to discuss the best fit for your needs.