StayMesh holds properties, units, bookings, tenants, obligations, tasks, evidence and communication in one model. Agents prepare the work. Your team approves what matters.
Channels, imports and property records are the foundations, not the product. Each pillar names the operational problem, the records involved, what AI prepares and what stays with your team.
Revenue and bookings
Booking requests, channel evidence, availability, amendments, cancellations and canonical booking records.
In development
The problem: The same booking arrives twice, changes once and cancels somewhere else. Nobody is certain which version is true.
Connected records
Booking requests and channel evidence, kept immutable
One canonical booking per case, with a revision number
Canonical availability per unit, separate from channel claims
What AI prepares
A match between an incoming signal and the right unit and case
A conflict warning when dates overlap an existing booking or block
A proposed amendment or cancellation with its source event
What stays human
Confirming a booking
Applying a date change
Applying a cancellation and releasing the calendar
A platform can report a cancellation without your calendar releasing anything. StayMesh shows both truths separately until a person applies the change.
Tenants and trust
Tenant onboarding, identity, solvency, documents, communication and tenant history.
In development
The problem: Tenant facts live in a booking platform, an inbox, a shared drive and somebody's head. Verification status is a feeling, not a record.
Connected records
Candidate details taken from booking evidence
Tenant records linked to bookings, with their own history
Documents and verification outcomes with their source
What AI prepares
A possible match with an existing tenant, never applied on a name alone
The list of information still missing before move in
A document request in the tenant's language
What stays human
Creating or linking a tenant record
Accepting or rejecting verification evidence
Rejecting an applicant
Verification carried out by a booking platform and verification carried out in StayMesh stay separate and traceable. StayMesh never presents someone else's check as its own.
The problem: Expected income and received income are treated as the same number until the bank statement disagrees.
Connected records
Expected rent periods derived from the booking dates
Deposits, fees and channel payouts as separate obligations
Received payments recorded against what was expected
What AI prepares
A full expected schedule from the commercial terms of the unit
An exception when two authoritative sources disagree on an amount
A reconciliation proposal when a receipt looks like a match
What stays human
Approving a financial schedule
Resolving a conflicting amount
Any movement of money
Expected is not received. StayMesh labels every amount as one or the other and never merges them.
Property operations
Check-ins, check-outs, inspections, cleaning, inventory, maintenance and external contractors.
In development
The problem: A booking changes on Friday and the cleaner still turns up on the old date.
Connected records
Arrivals and departures derived from canonical booking dates
Cleaning, inspection and turnover tasks per unit
Maintenance issues, contractor assignments and unit history
What AI prepares
Turnover work as soon as dates are confirmed or amended
A triage summary and troubleshooting steps for a reported issue
A contractor assignment with the access and scope details
What stays human
Dispatching a contractor
Granting property access
Approving spending
A booking prepares the operational work. It never dispatches people, opens doors or spends money on its own.
Compliance and control
Local requirements, required documents, renewal deadlines, permissions, approvals and audit history.
Planned
The problem: Rules differ per city and per contract type, and the deadline is usually noticed after it passes.
Connected records
Required documents per property and per contract type
Renewal and expiry deadlines
An immutable history of who approved what, and when
What AI prepares
A reminder before a document or permit expires
A checklist of what a specific location appears to require
An audit-ready summary of a booking's full life
What stays human
Confirming that a requirement applies
Signing or filing anything legal
Changing permissions and roles
StayMesh organises requirements and deadlines. It does not provide legal advice and it is not a substitute for professional counsel.
Shared operational foundation
Every agent works from the same operational truth.
Property, unit, booking, tenant, financial obligations, tasks, evidence and communication are one connected model. An agent does not need to be told the context, because the context is the product.
Property
Unit
Booking
Tenant
Financial obligations
Tasks
Evidence
Communication
One shared operational record. One history. One set of policies.
One booking amendment, seven prepared consequences
1Recalculate expected rent periods
2Reschedule check-in
3Reschedule cleaning
4Update tenant communication
5Flag a contract change
6Recheck availability
7Preserve the original channel evidence
Nothing is changed silently. Each consequence is prepared, shown with its source and applied only when it is approved. This shared model is the difference between StayMesh and running several disconnected AI tools next to each other.
How agentic operations work
From signal to resolved.
A signal enters StayMesh, an agent prepares an action, a person approves it, and the outcome is recorded next to the evidence that started it.
1
Observe
StayMesh receives an authorized signal from a booking channel, e-mail, tenant or connected service.
2
Understand
The relevant agent extracts the facts, connects them to the correct booking, tenant, unit and policy, and identifies missing information.
3
Prepare
StayMesh creates a proposed action, schedule, message or task with its source and reasoning.
4
Approve
Sensitive decisions remain with your team. You see what will happen before anything is applied.
5
Act
Approved actions can be executed within the rules and limits of the workspace.
6
Record
The outcome becomes part of the shared operational history without changing the original evidence.
StayMesh uses AI models to read and prepare work inside your workspace. It does not train models on your private workspace information.
Human control
Autonomous where it should be. Human where it matters.
Level 1
Observe
AI reads authorized signals, organizes evidence, detects exceptions and recommends the next action.
Level 2
Prepare
AI prepares booking decisions, tenant requests, payment reminders, cleaning tasks, maintenance assignments and communications for approval.
Level 3
Act within policy
Only explicitly permitted, low-risk actions may run automatically within workspace rules and defined limits.
Booking decisions, cancellations, tenant rejection, legal declarations, financial movements, property access and material spending require human approval unless the operator has explicitly configured and authorized a safe policy.
Sources and reasoning remain visible
Role-based permissions
Immutable activity history
Workspace isolation
Human approval controls
Agent limits
Kill switches
Spending limits
Privacy-conscious data handling
Honest product status
Start with today's operation. Grow into the autonomous one.
StayMesh is in early access. This is what is genuinely usable today, what is being built now and what is still on the roadmap.
Available now
Usable today inside a StayMesh workspace.
Multi-property and unit inventoryProperties, buildings, apartments and rooms in one structured inventory.
Property and commercial listing dataAttributes, amenities, pricing periods, fees, deposits and terms per unit.
Booking Inbox and Booking CasesEvery incoming signal becomes a reviewable case with its own evidence trail.
Human reviewed Spotahome e-mail ingestionBooking e-mails are parsed into structured facts that a person confirms.
Gmail label integration in betaRead only access to one chosen label, with no access to the rest of the mailbox.
Confirmation, amendment and cancellation workflowsApprove a booking, apply a date change or a cancellation with a full audit entry.
Availability calendarCanonical occupancy per unit, with conflicts surfaced before they are applied.
iCal import and exportOccupancy signals in and out, kept separate from financial booking records.
Role based workspacesOwner, administrator, operator and viewer roles inside an isolated workspace.
In development
Being built now, with parts already testable.
Tenant handoffTurning a confirmed booking candidate into a tenant record, with match checks.
Expected booking financial schedulesExpected rent periods, deposits, fees and payouts prepared for approval.
Additional channel adaptersMore booking sources normalised into the same canonical booking model.
Tenant onboarding and verificationRequired information, identity and solvency evidence, documents and follow up.
Cleaning and turnover workflowsCheck in, check out, inspection and cleaning work prepared from the calendar.
Maintenance coordinationIssue triage, urgency checks and prepared contractor assignments.
Tenant communicationContext aware multilingual messages prepared for a person to send.
Planned
On the roadmap, not built yet.
Bank and payout reconciliationMatching money actually received against what StayMesh expected.
Local compliance engineLocal requirements, required documents and renewal deadlines per location.
Agent policiesPer workspace rules that define what each agent may prepare or run.