Channels and PMS
Messages, availability, rates and stay records
Independent small and midsize hospitality groups
Guest service depends on back office coordination
Staff rebuild context across the inbox, reservation system, task tools and finance. Work can stall at each handover.
Proposed product
36 proposed workflows connect guest service with hotel operations.
Anonymous operating case informs the design. Wider demand needs validation. Major global brand portfolios excluded.
What we would build and connect
Connect hotel systems through shared services
POS and finance
Sales, purchase orders, receipts and ledger
CRM, HR and tasks
Consent, people, work and building data
Shared services
Property knowledge; context and owners; approvals and readback; audit and measures
One work surface
Inbox, mobile tasks, department approvals and leadership views
Existing systems retain records and action rights. People own service and approvals. Adapters are proposed integrations.
Guest facing inventory
Guest conversations
| Use case | Who uses it | AI work | Intended benefit | Simple example |
|---|---|---|---|---|
| Property questions and conversations | Reservations agent | Summarizes linked requests; drafts a reply for staff review | Fewer unanswered parts of a guest request | An email follows a guest’s earlier chat |
| Multilingual voice notes | Guest services agent | Transcribes and translates voice notes for staff review | Guest intent survives language differences | A voice note asks about late arrival |
| Call handling and handover | Caller and front office | Answers approved public FAQs and summarizes a staff handover | An owned transfer or callback | An evening caller asks for a person |
| Reservation and availability enquiries | Reservations agent | Uses permitted current facts to draft a reviewed reply | A complete sourced answer | Ask whether a room is available tomorrow |
Benefits are intended. Examples are illustrative. Staff decisions and action gates remain.
Reviewed text enquiries start first. Live voice needs its own release gate. No enquiry permits booking, change or payment writes.
Guest facing inventory
Guest service and experiences
| Use case | Who uses it | AI work | Intended benefit | Simple example |
|---|---|---|---|---|
| Service requests and updates | Guest services dispatcher | Extracts details and proposes a staff owned service task | Owned service with visible progress | Towels are requested in an occupied room |
| Lost item returns | Property custodian | Suggests item matches for staff ownership verification | A traceable return with clear custody | A guest reports a missing charger |
| Local itineraries | Guest and concierge | Proposes itineraries from checked local sources | Useful options with practical constraints | An itinerary must finish before dinner |
| Personalized stay preparation | Guest services manager | Proposes preference based offers and preparation | Feasible experiences with a fulfilment owner | A guest opts into a preferred room setup |
Benefits are intended. Examples are illustrative. Staff decisions and action gates remain.
Staff accept service ownership, verify item custody and approve offers. People perform physical work; proposed plans are not bookings.
Guest facing inventory
Guest bookings and payment requests
| Use case | Who uses it | AI work | Intended benefit | Simple example |
|---|---|---|---|---|
| Dining and activity bookings | Concierge and service team | Gathers requirements and drafts current service options | Confirmed references rather than vague promises | A guest requests a table for dinner |
| Eligible changes and cancellations | Reservation approver | Drafts eligible zero financial impact changes for approval | An exact reviewed change and source confirmation | An allowed change keeps price and fees unchanged |
| Approved room bookings | Reservation approver | Gathers details and prepares a booking proposal | Confirmed source status and clear terms | A guest chooses an offer and asks to book |
| Hosted payment requests | Finance or reservation approver | Drafts a hosted payment request explanation | A tracked provider payment handoff | An approved charge needs a time limited link |
Benefits are intended. Examples are illustrative. Staff decisions and action gates remain.
Later gates: exact staff approval, source readback and separate action permissions. No card details through AI; no general payment authority.
Back office inventory
Rooms, food and service quality
| Use case | Who uses it | AI work | Intended benefit | Simple example |
|---|---|---|---|---|
| Housekeeping and maintenance plans | Housekeeping and engineering supervisors | Forecasts cleaning work and proposes maintenance plans | Room readiness and maintenance ownership | A late checkout changes the cleaning plan |
| Room cleaning quality checks | Attendant and housekeeping supervisor | Flags visible omissions for human inspection | Clearer defects before supervisor release | A room photo appears to lack towels |
| Food demand and stock | Food and beverage manager | Forecasts covers and suggests replenishment | Clearer stock and staffing choices | High occupancy changes breakfast demand |
| Guest feedback and recurring issues | Quality manager | Classifies guest themes and sentiment with uncertainty | Recurring service issues reach human owners | Several reviews mention slow maintenance |
Benefits are intended. Examples are illustrative. Staff decisions and action gates remain.
Supervisors release rooms and verify work. Vision is advisory; forecasts need baselines. Complaints stay with people without an automatic reply.
Back office inventory
Commercial and event operations
| Use case | Who uses it | AI work | Intended benefit | Simple example |
|---|---|---|---|---|
| Campaign preparation | Marketing manager | Drafts consent filtered segments and campaign copy | Reviewable offers for permitted audiences | Invite opted in past guests to an offer |
| Market demand insights | Revenue manager | Explains licensed demand signals and proposes scenarios | Evidence beside pricing options | A local event changes the demand outlook |
| Rate and parity checks | Revenue analyst | Explains like for like rate discrepancies | Actionable parity exceptions | Breakfast terms differ between channels |
| Event layouts and setup | Event sales and banqueting | Proposes venue layouts and summarizes revisions | An agreed setup for operational review | A meeting changes from theatre to tables |
Benefits are intended. Examples are illustrative. Staff decisions and action gates remain.
Consent, comparable rate terms and approved venue constraints matter. Commercial changes and commitments remain staff approved in native tools.
Back office inventory
Finance and resource use
| Use case | Who uses it | AI work | Intended benefit | Simple example |
|---|---|---|---|---|
| Payment settlement reconciliation | Finance analyst | Ranks ambiguous settlement matches and explains differences | An owned queue of cash exceptions | An OTA remittance differs because of fees |
| Supplier invoice approval | Accounts payable clerk | Extracts invoice fields and suggests coding | Reviewable matches and approval evidence | An invoice exceeds its purchase order |
| Guest ledger reconciliation | Night auditor and controller | Explains folio and settlement exception clusters | A traceable daily close | A card authorization has not become cash |
| Resource use and intervention proposals | Facilities and sustainability manager | Explains meter anomalies and proposes interventions | Reviewable resource and comfort tradeoffs | Overnight water use rises without more guests |
Benefits are intended. Examples are illustrative. Staff decisions and action gates remain.
Matching does not permit posting or payment. Resource control needs a separate engineering envelope, local override and comfort checks.
Back office inventory
People, identity and shared knowledge
| Use case | Who uses it | AI work | Intended benefit | Simple example |
|---|---|---|---|---|
| Voluntary staff feedback | HR lead | Summarizes redacted voluntary feedback above group limits | Working condition themes for HR action | A team reports recurring rota concerns |
| Staff commitments and reminders | Employee and team manager | Drafts actions and reminders for owner acceptance | Accepted commitments stay visible | A meeting action lacks a named owner |
| Guest profile linking | CRM data steward | Suggests profile matches and explains uncertainty | Evidence for identity decisions | Two guest profiles share a family email |
| Approved knowledge and corrections | Staff member and content owner | Finds approved knowledge and proposes corrections for review | Current property specific answers | Two properties have different checkout policies |
Benefits are intended. Examples are illustrative. Staff decisions and action gates remain.
Feedback is voluntary and aggregated. People accept tasks and uncertain identity matches; content owners approve knowledge publication.
Back office inventory
Governed data and cross system answers
| Use case | Who uses it | AI work | Intended benefit | Simple example |
|---|---|---|---|---|
| Governed data connections | Domain data steward | Suggests mappings and explains data anomalies | Traceable definitions and trusted inputs | A finance extract no longer ties to totals |
| Cross system questions | Operations or finance analyst | Interprets permitted questions and explains computed results | Answers with definitions and source dates | Compare pickup with the approved budget |
Benefits are intended. Examples are illustrative. Staff decisions and action gates remain.
Certified definitions and cutoffs precede reporting. Owners approve release; simulations are modeled choices, not actual results or certified predictions.
Back office inventory
Leadership reporting and scenarios
| Use case | Who uses it | AI work | Intended benefit | Simple example |
|---|---|---|---|---|
| Leadership reporting packs | Finance lead and report owner | Drafts evidence linked commentary and charts | A reproducible pack for owner approval | Prepare a monthly property performance review |
| Operating decision scenarios | Operations leader and domain expert | Helps specify scenarios and explain modeled results | Visible assumptions before an operating choice | Explore later checkout versus cleaning capacity |
| Evidenced objective updates | Department head and executive support | Drafts evidenced objective updates and flags gaps | Comparable progress with visible missing inputs | A monthly target changed since the last report |
Benefits are intended. Examples are illustrative. Staff decisions and action gates remain.
Certified definitions and cutoffs precede reporting. Owners approve release; simulations are modeled choices, not actual results or certified predictions.
Back office inventory
Acquisition and development
| Use case | Who uses it | AI work | Intended benefit | Simple example |
|---|---|---|---|---|
| Development site screening | Acquisition analyst | Structures permitted listings and proposes rankings | Source backed leads for due diligence | A planning update changes a site’s screening |
| Development oversight | Development project director | Summarizes approved baselines, risks and missing updates | Owned exceptions in project reviews | A cost update conflicts with an older budget |
| Site photo evidence | Site manager | Tags dated photos and flags missing coverage | Comparable visual evidence for review | This week’s upload repeats an old image |
Benefits are intended. Examples are illustrative. Staff decisions and action gates remain.
Analysts verify sites and project facts. Photos and AI summaries do not certify construction, safety, payment milestones or investment decisions.
Proposed workflow
An extra night changes reservations and room work
Ask and identify
Guest asks for another night. Staff verifies stay identity and every request.
Check current options
PMS availability, rates and terms. AI proposes options and room work implications.
Confirm and change
Guest accepts terms. Reservations staff handles the priced extension in the native PMS.
Coordinate room work
After PMS readback, notify housekeeping. Supervisor replans the task and confirms ownership.
Verify and inform
Confirmed departure drives the reviewed reply. Housekeeping owns revised work; completion checked separately.
Proposed journey. A priced extension goes to staff in the PMS; automatic reservation change eligibility does not apply.
Proposed workflow
From supplier invoice to an approved accounting draft
Receive evidence
AP inbox holds the invoice. Procurement owns the purchase order and goods receipt.
Match and flag
AI extracts fields; rules check arithmetic, duplicates, PO and receipt. Exceptions return to AP.
Decide expenditure
Budget holder approves. Finance reviews coding, variance and separation of duties.
Create permitted draft
An enabled accounting adapter creates only the approved AP draft and posting proposal.
Read back the result
AP verifies draft reference and status. Finance retains separate posting and payment authority.
Proposed supplier invoice workflow. No automatic supplier creation, bank detail changes, ledger posting or payment release.
Intended value and commercial logic
Measure completed work and total operating cost
| Work | Primary measure | Include the exceptions |
|---|---|---|
| Guest requests | Effort per completed request | Unresolved handovers and reopens |
| Room work | Rooms ready by required time | Overdue work and repeat faults |
| Finance | Validated invoice handling time | Aged reconciliation exceptions |
| Whole platform | Total operating cost | Review, content, integration and support |
Recurring software by group, property and room. Implementation scoped separately. Pricing and margin need validation.
Include staff review, content stewardship, integration, infrastructure and support. No savings, revenue uplift or willingness to pay is established.
Future full platform
Base annual software scenario: USD 262m
Eligibility and pricing unvalidated. No verified global market or buyer census.
158 of 244 areas; 86 unquantified. Mixed stock and dates. Excludes China incl. Hong Kong/Macao, Russia and North Korea.
Substantive alternatives already exist
Compete against complete hospitality workflows
| Category | Examples | Relevant overlap |
|---|---|---|
| Hotel platforms | Mews, Cloudbeds | Reservations, guest journeys and integrations |
| Guest tools | HelloShift, Duve | Guest communication, services and task links |
| Operating tools | hotelkit, Sweeply | Teamwork, housekeeping and operating tasks |
| AI orchestration | Aphy, Inntelo | Back office and cross department overlap |
Lower total coordination effort with clear ownership, workable integrations and verified outcomes. Advantage remains unproven.
Vendor claims describe scope; they do not establish complete execution or independently verified outcomes.
Initial release and expansion
Start with a bounded guest service release
Reviewed guest work
Approved property knowledge, reviewed replies, read only enquiries and staff initiated service tasks.
Own the operation
Content and data owners, dispatch coverage, trained reviewers and protected business testers.
Expand on evidence
Production identity and rates, whole requests, source failures, owned handovers and agreed economics.
Proposed first release. Developer Demo tests are not operator acceptance. No booking or payment scope; support coverage requires agreement.
Operators and domain partners
Agree the first operating validation
Priority workflows, volumes and current effort.
Operator and department ownersSource rights, approved knowledge and accountable handovers.
Integration and domain partnersProtected business UAT and an accepted commercial offer.
Customer and delivery teamEnd of main pitch. Full requirements follow in the separate appendix.
Participation is proposed. No committed pilots, proven economics or funding amount is implied.