Source inventory and data map
Every system, feed and spreadsheet that touches a trip, with its owner, refresh rate and the fields it holds.
Services · Travel Data Engineering & Integration
One honest record of what's actually bookable
Personalization, routing and prediction are only as good as the data underneath. We unify reservation, PMS, CRM, supplier and content data into one travel-shaped model, deduplicated, mapped and governed by contracts, so every system works from operational truth.
Three records, one lodge, no single truth
The same canyon lodge lives in your systems three times: once from a bedbank feed, once in the contracts spreadsheet, and once typed in by an advisor years ago with the old phone number. Its room types don't match the channel manager's. Your CRM counts a couple who share an email as one traveler who took two trips. Each problem is small. Together they mean nobody can say with confidence what's bookable on a given Tuesday, and an AI feature built on top will answer confidently anyway.
- Supplier content duplicated across feeds, contracts and manual entry
- Room, product and rate codes that don't map between systems
- Opening hours and seasonal closures with no owner and no date
- Traveler records split apart, merged wrongly, or holding data nobody agreed to keep
What we deliver
Content mapping
Supplier, product, room and rate codes mapped across reservation systems, channel managers and GDS and bedbank feeds, with a review queue for uncertain matches.
Deduplication
Entity resolution for suppliers, venues and travelers: scored matches, human review below a threshold, and an audit trail for every merge.
Supplier truth registry
A rule for each field: which source wins, how fresh it must be, and what happens when sources disagree.
Travel-shaped warehouse
Trips, legs, services, suppliers, travelers, consents and events modeled in your warehouse, ready for analytics, forecasting and AI grounding.
Data contracts and pipelines
Schemas, freshness targets and change alerts agreed with each source, plus pipelines over REST APIs, webhooks, CSV/SFTP or MCP.
Supplier truth, field by field
“Which source is right?” has no single answer. It has a different answer for each field, and writing those answers down is most of the work. Here is a simplified version of the rules we might draft for a small collection of lodges:
| Field | Authoritative source | Fallback | Freshness target |
|---|---|---|---|
| Room inventory and rates | Property management system | Channel manager | Minutes |
| Seasonal opening dates | Supplier contract record | Supplier website, flagged for review | Monthly, and before each season |
| Accessibility details | Site-visit notes from your team | Supplier self-report, labeled as such | Annually |
| Cancellation terms | Contract system | None: missing terms block publication | On every change |
| Guest contact preferences | CRM consent record | None | Real time |
The fallback column matters as much as the first one. When an authoritative source goes quiet, the system has to know whether to use the next-best source, label the fact as unverified, or refuse to publish it. That last option is what keeps an itinerary generator from recommending a café that closed in March, a failure we pick apart in Phantom trails and closed cafés. For the fuller method, read Supplier truth: content mapping, deduplication and the data behind personalization.
How it works
Map, model, match, move, monitor
- 01
Map
1–2 weeksInventory the sources, interview the people who maintain them and sample the data. Most of the surprises surface here.
- 02
Model
2–3 weeksAgree the target model and the supplier-truth rules, field by field, with your operations team in the room.
- 03
Match
2–4 weeksRun content mapping and deduplication, clear the review queue with your staff, and document every rule we automate.
- 04
Move
3–6 weeksBuild the pipelines and contracts, load the warehouse, and switch downstream systems to the unified records one at a time.
- 05
Monitor
OngoingFreshness, volume and schema checks on every source, with alerts to the owner before bad data reaches a traveler.
Side by side
What changes underneath
AspectWithout integration workWith integration work
Supplier recordsDuplicated across feeds and spreadsheetsOne record per supplier, with a source and date for each field
When sources disagreeWhoever edited last winsWritten rules decide, and real conflicts go to a person
Upstream schema changesDiscovered when a report breaksCaught by a data contract and sent to the owner
AI groundingModels fill gaps with plausible guessesModels read verified fields or say they don't know
Checklist
Outcomes to expect
The operational changes usually show up within weeks. The analytical ones depend on what you build on top.
0 of 6 ready
We'd called it a software problem for years. It was a which-phone-number-is-right problem, multiplied by every supplier we work with. Once that was settled, the software questions got easy.
Where unified data goes next
- PlatformIntegrationsThe system categories and integration patterns we support.
- PlatformSignalsForecasting that depends on exactly the clean history this work produces.
- PlatformBooking FlowSupplier requests and confirmations that rely on current contacts and terms.
- ServiceResponsible AI & Traveler Data GovernanceConsent and minimization designed into the data model from the start.
Questions
Frequently asked questions
Which systems do you integrate with?
By category: property management systems, channel managers, reservation systems, GDS and bedbank feeds, payment providers, CRMs, messaging channels, mapping and transit data, weather data and data warehouses. Our integrations page explains the patterns.
Do you replace our PMS or reservation system?
No. Your operational systems stay the systems of record for what they own. We sit between them, map them to each other, and give everything downstream a consistent view.
Whose warehouse does the data live in?
Yours. If you have one, we build into it. If you don't, we set one up in your own cloud account, so you own it from day one.
What is a data contract?
A written, testable agreement with a data source: the schema, what each field means, how fresh it will be and who to call when it changes. When a source breaks its contract, the pipeline stops and the owner hears about it before a traveler does.
Is this only for large operators?
No. Hotels and lodges with one PMS and a channel manager need a lighter version. Tour operators and DMCs with hundreds of suppliers need the full one. Travel advisors mostly need CRM and supplier mapping.
Related
Keep exploring
- PlatformIntegrationsHow Nova connects to PMS, reservation systems, channel managers, payments, CRMs and messaging: REST, webhooks, CSV/SFTP and MCP, with written data contracts.
- WorkDesert lodge collectionHow a small collection of desert lodges flagged 7 in 10 late cancellations two weeks early and filled quiet shoulder-season midweeks with packages.
- Field Notes · Data & AnalyticsSupplier Truth: Content Mapping, Deduplication and the Data Behind PersonalizationPersonalization fails on a messy catalog. A practical guide to content mapping, match scoring, field-level authority and freshness for travel suppliers.
- Field Notes · Data & AnalyticsPredicting Cancellations Before They HappenA practical walk-through of cancellation prediction for lodges and tour operators: defining the label, choosing features, calibrating scores and acting on them.
- ServicesResponsible AI & Traveler Data GovernanceConsent, minimization, careful handling of mobility, health and dietary details, bias reviews, red-teaming and clear policy for the AI in your travel business.
Find out what your data can support
Send us a list of the systems and spreadsheets that touch a booking. On a discovery call, we'll tell you where we'd start mapping and what it would unblock.