A guest arrives at the front desk and asks whether the hotel can send the invoice from their last stay to the email address on their profile.
The property management system (PMS) shows one guest profile. The booking engine shows another. The point of sale has a room charge under a slightly different name, while the spa system has the guest’s current email address. Each system contains a record called “guest”. None is describing the guest in quite the same way.
Now give an agent access to those systems and ask it to help.
The agent may understand that a guest can have reservations, stays, folios and preferences. It may know exactly what each field means. But it still has to answer a more basic question: which of these records belongs to the same person?
This is where two different data problems are often collapsed into one. A semantic layer gives systems and agents a shared language. Guest identity resolution establishes the person that language is describing.
What a hotel semantic layer is
A semantic layer is a shared business representation of data. It translates the different terminology, structures and rules used by individual systems into definitions that people, reports, products and agents can understand consistently.
In business intelligence, the familiar version is a shared definition of a metric. If finance and revenue both ask for occupancy or RevPAR, the semantic layer ensures that the same calculation, inclusions and exclusions are applied each time. Two dashboards should not produce two answers simply because they were built by different teams.
The same principle can be applied to the entities underneath those metrics. A hotel semantic layer can define what a guest, reservation, stay, room, rate or folio is, and how those entities relate to one another. It can establish, for example, whether the person who made a reservation is always considered the staying guest, or how a reservation becomes a completed stay.
This matters because each hotel system represents the operation from its own perspective. A property management system (PMS) may organise data around reservations and guest profiles. A point-of-sale system sees transactions and room charges. A spa system sees appointments and customers. A semantic layer maps those different structures into a shared model without pretending that the source systems were designed around the same definitions.
That shared model gives anything connected to it a common language. A report, a front-desk assistant or an agent that has not been introduced yet can work from the same meaning of guest, reservation and stay. Without it, each tool may be correct according to its own source system while the hotel remains collectively wrong.
What it is not
A semantic layer does not create one guest from records scattered across different systems.
It can define what a guest is, map the different fields used to describe one and establish how a guest relates to a reservation, stay or folio. But it does not determine that an OTA booking made with a proxy email address, a point-of-sale room charge and a spa appointment all belong to the same human being.
That is guest identity resolution: matching records, merging them when there is enough evidence and deciding which values should survive when the source systems disagree.
The distinction matters because a shared definition of “guest” can still point to several disconnected records. Teaching an agent what a guest profile means does not tell it which guest profile is the right one. The agent may understand the structure of the data perfectly and still act on an incomplete or duplicate view of the person.
A useful way to separate the two jobs is:
- A semantic layer answers: what does this record mean?
- Guest identity resolution answers: which records describe the same person?
A semantic layer can speak consistently about a person once that person has been resolved. It cannot produce the resolved person on its own.
Why the order matters
Most hotels are not going to replace their entire technology stack with one system or move every part of the operation onto a single data model. They will keep the property management system (PMS), point of sale, booking engine, spa system and other tools that already run the hotel.
As a result, the same guest can exist several times across the stack, under different identifiers and with different contact details. Each source system may contain a valid part of the picture without any one of them containing the whole person.
A semantic layer can teach an agent how to interpret the records returned by those systems. It can explain what a guest profile, reservation, stay or room charge represents. But the agent is still left asking which records belong to the guest in front of it.
This is why the order of work matters. Semantic modelling and identity resolution can be developed in parallel, but before a hotel can reliably act on a guest across systems, it must resolve the person those systems describe.
Once one Master Profile exists, a shared vocabulary has something stable to attach to. Reports, applications and agents can use the same language about the same person. Without identity resolution, a semantic layer provides a consistent way to describe fragments.
What Casa Layer is
Casa Layer is the guest identity layer for the hotel that keeps its existing technology stack. It connects guest records from the property management system (PMS), point of sale, booking engine and other source systems, determines which records belong to the same person and creates one Master Profile per real guest.
Casa Layer then builds a semantic layer onto those Master Profiles. It gives guest data, attributes and relationships consistent meanings so reports, applications and agents can work from the same understanding of the same person.
It is not a PMS, CRM or messaging product. Those systems continue to run the operation, while Casa Layer sits above them as the persistent identity and semantic foundation for guest data.
Identity resolution establishes who the guest is. Casa Layer’s semantic layer makes that resolved guest consistently understandable and usable.
Frequently asked questions
What is a hotel semantic layer?
A hotel semantic layer is a shared business representation of data. It defines what entities such as guests, reservations, stays, rooms and folios mean, how they relate and which business rules should be applied when people, reports, products or agents use them.
Does a semantic layer decide which records belong to the same person?
No. A semantic layer can define what a guest profile means, but it does not determine that records from a property management system (PMS), point of sale, booking engine and spa system belong to the same human being. Matching and merging those records is guest identity resolution.
Do you need a semantic layer if the stack uses a single data model?
Not always. If every part of the operation genuinely uses the same underlying records, definitions and identifiers, there may be little need for a translation layer. The distinction becomes important when a hotel retains multiple systems that represent guests, reservations and stays differently. Even with a shared data model, the hotel must still establish whether records entering from outside that model describe an existing guest. A common vocabulary and a resolved identity remain different jobs.
What does Casa Layer do?
Casa Layer resolves guest identity across the hotel's existing systems. It matches records that belong to the same person and creates one persistent Master Profile per real guest. Casa Layer then applies a semantic layer to those Master Profiles so reports, applications and agents can work from shared definitions attached to the same resolved person.