Site map

Find the Tamaga page that answers your question.

Tamaga works on the points where hotel meaning, guest decisions, booking systems, human responsibility and public representations meet. Start with the problem you are trying to solve, or explore the complete knowledge map below.

01
Tamaga Hospitality
The proposition and product architecture behind Tamaga's first hospitality system.
- Tamaga Hospitality: direct booking and hotel knowledge
  The overview of the proposition for independent hotels: an owned direct-booking path or a governed layer around an existing stack.
- Product architecture: House Record, Guest View, Host Desk and Channel Lens
  See how the current product model turns that proposition into four product surfaces and optional continuity layers.
02
Room and choice
Will the room actually work for this guest? Suitability is more than capacity or availability.
- Tamaga Hotel: a fictional working demonstration
  Inspect Room Fit, conditional requests, Host Attention, and representation checking without mistaking a proof environment for a live property.
  - Open the working Hotel DemoExternal
    Open the public proof environment itself.
- How guests book the wrong room even when the website is accurate
  Building Room Fit exposed a deeper problem: hotel inventory compresses physical rooms into sellable categories. The guest needs enough of that physical reality restored before availability can become a useful answer.
03
Promise and action
What remains a request, and who must act before it becomes a promise?
- How Tamaga works: from hotel promise to host action
  Follow a hotel condition from approved meaning through guest understanding, a pending request, Host Attention, and responsible resolution.
04
Systems and authority
Which system is allowed to know, decide, or write?
- Hotel PMS and booking integrations
  Keep transactional authority explicit while adding only the governed context and handoffs a property actually needs.
- Owned at the source. Interoperable at the edges.
  Travel knowledge should be able to move without losing the source that can maintain, qualify and correct it.
- Direct booking needs more than a booking button
  A traveler may begin on an OTA and book on the hotel website. The harder question is whether room meaning, conditions, and unresolved requests survive the handoff.
- Keep your PMS
  PMS adoption does not remove the need for a stronger owned hotel layer. It changes the problem: keep the systems that transact the stay, then protect the room meaning, guest context, conditions, and human confirmations that can disappear between them.
- What an independent-hotel chatbot should never answer from memory
  Hallucination is only one failure mode. A hotel answer can be accurate in wording and still be wrong because it is stale, applied to the wrong room or rate, or promoted from a possibility into a promise.
05
Representations and evidence
Does approved hotel meaning survive when it becomes a website, feed, booking surface, or synthesized answer?
- One Hotel, Seven Versions: the open research method
  Read the public-source case for tracing decision-critical hotel facts across seven representations.
  - Download the fixed PDF editionPDF
    Inspect the byte-stable downloadable research edition.
- Seven-Version Snapshot: trace one real hotel promise
  Request the current manual, human-reviewed diagnostic for one consequential hotel fact.
- Tamaga releases a hotel that does not exist
  Tamaga Hotel is a fictional reference environment with public guest-facing surfaces and a documented bounded release scenario showing what happens when a room does not fit, a request remains pending, a person still has to decide, or a representation loses an important condition.
- What can the travel internet safely forget?
  Travel platforms have to compress places. The difficult question is which distinctions can disappear safely, which ones change a traveler’s decision, and when a representation has reached the limit of what it can establish.
- The website no longer owns the final sentence
  Machine-readable does not mean authoritative. A hotel sentence still needs an owner, a condition, evidence, and a responsible next action.
- Schema.org is not the source of truth. It is the public edge of travel knowledge infrastructure
  Search, AI, OTAs, booking engines, email, APIs, and staff systems do not read one website; they assemble fragments. The durable asset is governed travel knowledge, with Schema.org as its public semantic spine.
- Tamaga releases an open method for tracing hotel promises across seven versions
  The open method behind One Hotel, Seven Versions traces decision-critical hotel facts across owned, transactional, structured, distributed and synthesized representations. Its first public-source case shows how a fact can remain visible while the condition that makes it trustworthy disappears.
06
Experiments and synthetic evidence
Controlled material designed to test retrieval behaviour, not ordinary authoritative hotel truth.
- AI Citation Lab: a controlled late-arrival retrieval experiment
  Inspect deliberately synthetic hotel material that tests whether a system preserves conditions instead of flattening them.
07
Newsroom
Research, field notes, and releases in chronological order. The semantic map above gives each public article a primary home.
- Newsroom: the chronological public record
  Browse the complete public record by publication date and editorial stream.
  - Follow the Newsroom through RSSRSS
    Subscribe to the public publication feed.
08
About the work
Travel knowledge infrastructure and the people behind it.
- Travel knowledge infrastructure: the category definition
  The governed system underneath the travel website: meaning, representation, authority and action.
- About Tamaga: why we are building it
  Understand the company and the hospitality system through which the wider idea currently takes shape.
- Dan Metille: the questions and practice behind the work
  Read the field-notebook account of the questions, work, and method behind Tamaga.
- Contact Tamaga
  Choose the right conversation for a property, platform, research question, or correction.
09
Legal
The site's privacy boundary.
- Privacy
  Read how the site handles analytics, newsletter, and contact data.