AI Otel glossary: Tourism Brain, shared core, decision record and other terms
This glossary defines AI Otel terms with fixed Turkish forms and status words. It clarifies what is built, live, or planned, ensuring accurate communication.
Contents
Key takeaways
- Every AI Otel term carries one of five status words, and only OtelCiro, the ecosystem's hotel growth and consultancy arm, is live.
- The Tourism Brain has three levels, Hotel Brain, Platform Brain and Industry Brain (Sektör Beyni), and none of them is trained: the Hotel Brain is architected and entering training.
- A decision record logs what changed and lets a person undo it, except a message already sent to a guest; an audit record is append-only and tamper-evident.
- The benchmark that will choose AI Otel's models holds 30 cases and 5 traps and has not been run, so no model has been chosen by measurement yet.
- The shared core defines every hotel action once and serves it to hotel teams, partners and the AI assistants people already use, in a public demo environment on a fictional hotel.
This AI Otel glossary fixes one meaning per term and attaches a status word to each. The Tourism Brain is the intelligence layer AI Otel is building for Türkiye's tourism; the shared core defines every hotel action once; a decision record logs each change and can be undone. Nothing in AI Otel's technology is live.
A hotelier reads that a system is built, an investor reads that it uses AI, and a journalist reads that it is live. When the same three words can mean three different things, a fixed vocabulary is the cheapest form of honesty a technology company can offer. This page defines the terms AI Otel uses, each with its Turkish form and its status word, so that other articles can link here instead of redefining. After reading it, you will be able to quote any AI Otel term exactly, and to ask any vendor the same five questions this glossary is built to answer.
How do you read an entry in this glossary?
Each entry gives five things: the English term, its fixed Turkish form, a plain definition, a status word, and the article that goes deeper. Read the status word before the definition.
The status word comes first because it answers the only question that matters commercially: is this in customers' hands today? A definition without a status is a description of an idea, not of a product. The Turkish forms are fixed decisions, not translations of convenience: Turizm Beyni, Otel Beyni, Platform Beyni, Sektör Beyni, ortak çekirdek, misafir uygulaması, dağıtım merkezi, ek satış. Using a synonym in one language breaks the link between the two versions.
Acronyms get a plain gloss at first use: KBS is the guest identity notification, KVKK is the personal data protection law, e-Fatura and e-Arşiv are electronic invoices. Where a term has an approved sentence, quote it rather than paraphrasing it upward. A glossary exists to stop a definition from growing in the retelling.
The terms fall into four families: vision (the brains), software (the operating system and its parts), safety (records and tiers) and model (how the AI itself is chosen). They are grouped by family, not alphabetically, because a reader usually arrives with one question, not with a letter.
ChecklistBefore quoting any AI Otel term, check
- The English term as written
- Its fixed Turkish form
- The plain definition in one sentence
- The status word: live, built, in development, architected or roadmap
- The article or page that carries the longer explanation
What is the Tourism Brain, and what are its three levels?
The Tourism Brain is the technology and intelligence layer AI Otel is building to connect hotels, travelers, tourism businesses and distribution channels across Türkiye. It has three levels, and each level carries its own status word.
The fixed wording is "AI Otel is building the Tourism Brain of Türkiye." Always building, never is: the vision is a direction of work, not a finished object. The three levels are separate terms with separate statuses, and mixing them is the most common reading error.
The Hotel Brain (Otel Beyni) understands one property. It is architected and entering training; today the PMS assistant answers from each hotel's own data. The Platform Brain (Platform Beyni) is designed to learn patterns across the network, with privacy thresholds so that no single hotel can be identified. The Industry Brain (Sektör Beyni, never Endüstri Beyni) is on the roadmap: network signals joined with public tourism data to understand Türkiye's tourism market. Public institutions are never described as its users.
The line that holds the three together is "Three levels of brain. One promise: a hotel's knowledge stays its own." It is a design promise, not a residency claim. Words that are never used for any level: trained, fine-tuned, learns from your hotel. Each of them would turn a design into a false claim about today. For the full picture, read The Tourism Brain of Türkiye: what AI Otel is building.
Watch outThe honest zero, in full
"We have no trained model yet, and the product says so on screen. What exists is the architecture, the decision infrastructure a model would run inside, and a benchmark to choose the model."
What is the hotel operating system, and how does the shared core differ?
The hotel operating system is the software a hotel runs its day on. The shared core is the layer underneath it that defines each hotel action once, so people, partners and AI assistants all work from the same definition.
The hotel operating system covers reservations, front desk, housekeeping, billing, restaurant sales, reports and compliance in one place. It is built and working, ahead of launch, measured at 400+ database tables, 150+ built-in reports and 5 languages (June 2026). The shared core (ortak çekirdek, AI Otel Core) defines every hotel action once and serves it to hotel teams, partner systems and the AI assistants people already use. It is built and runs in a public demo environment, on a fictional hotel. If the phrase "agent-native" appears, it means a core built around actions an assistant can call, not around screens; the plain words are shared core.
| Term | Turkish form | Plain definition | Status |
|---|---|---|---|
| Hotel operating system | otel işletim sistemi | reservations, front desk, housekeeping, billing, restaurant sales, reports, compliance in one place | Built |
| Shared core | ortak çekirdek | every hotel action defined once, served to people, partners and AI assistants | Built, demo |
| PMS assistant | PMS asistanı | plain-language answers about one property; a person confirms anything sensitive | Built |
| Guest app | misafir uygulaması | each hotel's own app from the browser, no app store | Built |
| Distribution hub | dağıtım merkezi | contracted rates to agencies; every quote shows its working | Built, seed data |
| One room inventory | tek oda envanteri | one inventory feeding own site, OTAs, agencies, assistants | Built |
Two notes on the table. The guest app holds the bill, room service, spa bookings, requests and offers; an accepted offer goes straight onto the guest's bill, which is ek satış, not a conversion figure. The distribution hub is built with seed data and no agency is connected yet, and one room inventory is a build statement, not a claim that all channels run on it today. The mechanism is explained in AI Otel Core: how hotel teams, partners and AI share one system.
What is a decision record, and how is it different from an audit record?
A decision record is the running log of what the system changed and who confirmed it. An audit record is the append-only chain that shows whether anyone later altered that log.
The decision record works in three moves: small changes apply automatically and are logged, larger ones wait for a person, and every change can be undone, except a message already sent to a guest. It is built and demonstrated in the demo environment. The audit record is stricter: past records can only be appended, and changing or deleting them is closed in the database. If someone still alters one, the chain shows it. Both are built, demo.
Tamper-evident means a process that makes alterations to the data easily detectable, the definition given by the NIST Computer Security Resource Center glossary (opens in a new tab) (NIST CSRC, 2026-11-01). It does not mean an alteration cannot happen, and the term is never written as tamper-proof or blockchain.
Two related terms sit beside these. Tenant isolation means each hotel's data is separated at the database level; that is the whole sentence, with no table count and no certification attached. Safety tiers mean every action the assistant can take sits in one of three tiers: some run directly, some need one person's confirmation, the most sensitive need two. The threshold between tiers is fixed in code, so the glossary does not say the hotel sets it. A reader who needs the numbers should ask for the registry snapshot rather than trust a figure repeated in a slide. The reasoning behind this layer is in Why AI Otel built the safety layer before choosing a model.
Key pointOne line, repeated wherever the term appears
Tamper-evident is not tamper-proof. It means an alteration cannot be hidden, not that it cannot happen.
What does model-agnostic mean, and what is the benchmark for?
Model-agnostic means the architecture is not tied to one AI model: the model is chosen per deployment by measured evaluation. The benchmark is the test set that will do the measuring.
Because the choice is made by measurement, AI Otel material names no AI vendor and no model. The phrase used instead is "the AI assistants people already use." The benchmark holds 30 cases and 5 traps, and it has not been run. Results will be published when measured; until then no score exists to quote.
The traps matter more than the count. A case where the correct answer is to ask a person, or to refuse, tests a system more honestly than a case with an easy answer. The safety layer came before the model on purpose: the decision record, the safety tiers and the audit record exist so that any model can run inside limits a person set, whichever model is finally chosen.
Words this family excludes: predicts, forecasting, machine learning, AI pricing engine. Each describes a capability that a benchmark result would have to support first. A hotelier can reuse the idea without any AI at all: write down the ten questions your front desk answers every day, and the three where a wrong answer costs money. That is a benchmark, and it is the right shape.
NoteBenchmark status in one line
30 cases, 5 traps, not yet run. Results will be published when measured.
What do the five status words mean?
Each status word answers one question: is this in customers' hands? Live is. Built is not yet. In development, architected and roadmap sit further back.
| Status (EN) | Status (TR) | What it means | Example |
|---|---|---|---|
| Live | Faaliyette | operating today | OtelCiro |
| Built | Geliştirildi | working software, ahead of launch, not in customers' hands | hotel operating system, shared core, PMS assistant, decision record, benchmark |
| In development | Geliştiriliyor | being built now | TRBNB |
| Architected | Tasarlandı | designed | Hotel Brain, Platform Brain, ÖKC connection, door locks |
| Roadmap | Yol haritasında | planned | Industry Brain, flights and connectivity, hosting in Türkiye |
TRBNB is the ecosystem's sales and distribution arm, and it is in development. ÖKC refers to fiscal cash-register devices, designed to connect outlet by outlet. Nothing in AI Otel's technology is live, and the Turkish never uses canlı. A status word is only useful if it is used against interest, which is why the weakest one appears most often on this page. The full reasoning is in Built or live? The five status words AI Otel uses.
What does AI Otel not have yet?
No trained model, no connected agency, OTA or partner, no certifications, and no count of hotels, bookings or users on its technology.
The honest zero, in the words the product uses: "We have no trained model yet, and the product says so on screen. What exists is the architecture, the decision infrastructure a model would run inside, and a benchmark to choose the model." The benchmark is built and awaiting measurement: 30 cases, 5 traps, not yet run.
Integrations are built and fixture-tested; going live needs the authority and integrator credentials. That is not the same as connected. Hosting and AI processing in Türkiye are on the roadmap; until it is done, the glossary does not claim it and does not soften the gap with a residency phrase. There is no TGA integration, and the Ministry or TGA is never described as a partner or an endorser.
On the other side of the ledger, OtelCiro works with paying hotel clients today, and the hotel operating system is built and working, ahead of launch. A glossary that lists absences is doing its job. A term a company cannot define precisely is usually a claim it cannot support.
ChecklistFive questions to ask any vendor term, including ours
- Which status word applies?
- Which exact sentence do you publish?
- What is not built yet?
- Who confirms a sensitive action?
- What happens when the system is wrong, and can it be undone?
What to do next
Pick one term you use about your own hotel's software, and ask which status word applies to it and which exact sentence the vendor publishes. If the answer is vague, the term is doing marketing work rather than describing software. The rest of this glossary then becomes a reading order: the vision terms explain where the work is going, the software terms explain what exists, the safety terms explain what limits it, and the status words tell you which of the three you are reading. To walk through any entry with us, request a briefing or write to info@ai-otel.com.
Frequently asked questions
Is the Tourism Brain trained?
No. There is no trained model yet, and the product says so on screen. The Hotel Brain is architected and entering training; today the PMS assistant answers from each hotel's own data. The Platform Brain is designed, the Industry Brain is on the roadmap, and the benchmark that will choose the model is built and has not been run.
What is the difference between Built and Live?
Built means working software, ahead of launch, not in customers' hands. Live means operating today. Only OtelCiro is live; nothing in AI Otel's technology is. The distinction exists so that a working system is never described as a used one.
Why tamper-evident and not tamper-proof?
Tamper-evident means a process that makes alterations to the data easily detectable, the definition in the NIST Computer Security Resource Center glossary (2026-11-01). It does not claim that alteration is impossible. Records are append-only, changing or deleting is closed in the database, and if someone still alters one, the chain shows it. No blockchain is involved.
What is the shared core in one sentence?
Every hotel action is defined once and served to hotel teams, partner systems and the AI assistants people already use. It is built and runs in a public demo environment on a fictional hotel, so it is a demonstration of the mechanism, not a live hotel.
Does AI Otel have an integration with the Ministry or the TGA?
No. There is no TGA integration, and the Ministry or TGA is never a partner or endorser. There are also no connected agencies, OTAs or partners: the distribution hub is built with seed data, and the booking-site integration is in pre-production testing. Integrations are built and fixture-tested; going live needs the authority and integrator credentials.
Sources
- Glossary: tamper evident (NIST CSRC; definition sourced from NISTIR 8202) (opens in a new tab)NIST Computer Security Resource Centercsrc.nist.gov