Una by Polydom FAQ: Capabilities, Integrations, Security & Results

Can an AI receptionist actually change a hotel reservation? What happens if it gives a guest the wrong information? Can it access your PMS, handle calls overnight, and know when to involve a human?

These are the questions hospitality operators should ask before giving AI responsibility for guest communication and daily operations.

Una by Polydom is a voice-first AI Digital Employee for hotels, serviced apartments, and professional vacation rental managers. She is designed to communicate with guests, execute supported booking and operational workflows within connected hospitality systems, and escalate situations requiring human judgment.

Unlike an information-only chatbot, Una doesn't stop at answering a question. When the necessary integrations, permissions, and property rules are in place, she can help complete the work behind it.

This FAQ addresses Una's capabilities, PMS integrations, multilingual communication, operational controls, security, reliability, implementation, and documented customer results.

What Is Una and What Can She Do?

1. What Is Una by Polydom?

Una by Polydom is a voice-first AI Digital Employee designed to handle repeatable communication and operational workflows for hotels and vacation rental businesses.

Una works with existing hospitality technology, including compatible property management systems (PMS), guest communication channels, and operational task management tools.

Depending on the integration and configuration, Una can answer guest inquiries, retrieve reservation information, create or modify bookings, coordinate service requests, and escalate situations requiring employee judgment.

Her purpose is to reduce the manual coordination required to move a guest request from conversation to action.

Tools need operators. Una is the operator.

For a complete product introduction, read What Is Una? The AI Digital Employee for Hotels and Vacation Rentals.

2. Is Una a Hotel AI Chatbot?

Una can perform functions commonly associated with a hotel AI chatbot, but her operational scope is broader.

A communication-focused chatbot typically answers questions, shares property information, or forwards requests to employees.

Una is designed to execute supported workflows through connected hospitality systems.

For example, when a guest asks to extend a stay, Una can potentially retrieve reservation information, check availability, and perform an authorized modification through a compatible PMS.

The distinction is not the conversational interface. Some advanced chatbots also support transactional workflows.

The difference lies in which actions the system can actually complete after understanding the guest's request.

3. What Does Una Do During Daily Hotel Operations?

Una supports three primary areas of hospitality work.

Guest communication: Handling incoming calls and supported digital conversations, answering routine questions, and providing relevant property information.

Reservation workflows: Retrieving availability and reservation information, supporting direct bookings, and creating, modifying, or canceling eligible reservations through compatible integrations.

Operational coordination: Capturing guest requests, creating or routing structured tasks, and escalating exceptions to the appropriate employees.

For hotel operators, the objective is to reduce repetitive work while maintaining consistent guest communication and operational follow-through.

4. How Is Una Different From a Virtual Concierge or AI Receptionist?

A virtual concierge generally focuses on guest assistance, such as answering questions about amenities, recommending local experiences, and helping guests access hotel services.

An AI receptionist focuses on front desk communication, including telephone calls and routine inquiries.

Una can support these functions, but she is designed to go further by executing authorized actions inside connected hospitality systems.

A guest asking about availability may need information. A guest requesting a reservation change needs an action.

Una's operating model connects both types of requests to the appropriate workflow.

For a practical vendor comparison, see How to Choose a Hotel AI Chatbot: 10 Criteria That Actually Matter.

PMS Integrations and Guest Communication

5. Which Property Management Systems Does Una Integrate With?

Una supports integrations with hospitality platforms including Apaleo, Guesty, Mews, Hospitable, Zeevou, and other compatible systems.

Depending on the platform and deployment, an integration may provide access to:

• Reservation information.
• Availability and rates.
• Property details and applicable booking conditions.
• Supported reservation creation, modification, or cancellation.
• Other authorized operational actions.

Not every integration supports identical workflows.

A connection that allows Una to retrieve reservation information does not necessarily permit every booking or payment action.

Operators should confirm which capabilities are supported by their specific PMS and configuration.

Learn how Una connects to your PMS and operates your hotel's technology stack.

6. Does Una Work Across Phone, WhatsApp, Email, Web Chat, and Kiosks?

Yes. Una supports guest communication across compatible voice and digital channels.

Depending on the deployment, these can include:

• Incoming telephone calls.
• Website voice and text interactions.
• WhatsApp and supported messaging applications.
• Email.
• Supported OTA messaging channels.
• Social messaging.
• QR-based guest access points.
• Lobby kiosks and shared guest-facing devices.

The available actions may differ between channels.

For example, an operator should verify whether a booking workflow supported over the phone is also available through a particular messaging integration.

Multi-channel availability does not automatically guarantee shared guest context or identical functionality across every platform.

7. Can Una Communicate in Multiple Languages?

Yes. Una supports multilingual guest communication across compatible voice and text channels.

Her text communication supports more than 90 languages, while telephone language detection and voice communication support a smaller, defined set of languages.

The available voice languages include English, Spanish, French, German, Hindi, Russian, Portuguese, Japanese, Italian, and Dutch.

Language availability should be confirmed for the intended channel and deployment.

For hotels serving international guests, multilingual support can help provide more accessible communication without requiring every staff member to speak every guest language.

However, accurate responses still depend on correct property information, supported workflows, and applicable policies.

8. Can Una Make, Modify, or Cancel Hotel Reservations?

Yes. Una can perform supported reservation actions through compatible PMS integrations and configured permissions.

These may include creating bookings, modifying eligible reservation details, and processing permitted cancellations.

For example, a guest requesting a date change may need the system to identify their reservation, check availability, apply the relevant booking conditions, and update the record.

Una can support that sequence when the connected PMS permits the necessary actions.

However, some modifications may be restricted by reservation source, payment status, rate conditions, or property policies.

Una should escalate requests that require additional authorization or fall outside the supported workflow.

9. Can Una Help Hotels Capture Direct Bookings?

Yes. Una can support direct booking workflows by responding to inquiries, retrieving relevant availability and rate information, and completing authorized reservation actions through compatible systems.

For example, a prospective guest calling outside staffed hours may be able to discuss accommodation options and proceed with a reservation instead of waiting for the front desk to reopen.

At Lee Gardens Extended Stay in Orlando, Una made 232 direct reservations over a ten-month reporting period, representing $80,968 in reported booking revenue.

These results describe one hotel deployment, not guaranteed incremental revenue for other properties.

Read the Lee Gardens Extended Stay case study.

10. Can Una Handle Housekeeping and Maintenance Requests?

Una can help capture housekeeping, maintenance, and other operational requests and create or route structured tasks through compatible systems.

For example, when a guest reports that the air conditioning isn't working, Una can collect relevant information and initiate the supported task workflow.

The maintenance team remains responsible for assessing and repairing the problem.

Creating a task is not the same as completing the physical work.

Una's role is to reduce the manual coordination required to transfer the request from the guest to the people responsible for resolving it.

11. What Happens When Una Needs a Human Employee?

Una is designed to escalate situations that require human judgment, additional authorization, or work outside her configured responsibilities.

These may include:

• Unusual refund or cancellation requests.
• Booking exceptions requiring managerial approval.
• Sensitive guest complaints.
• Urgent situations that require a person.
• Requests unsupported by the connected systems.
• Cases where necessary information is missing or contradictory.

Depending on the workflow, Una can pass relevant context or create a structured request for the appropriate team.

Human employees retain responsibility for decisions requiring discretion, physical service delivery, and the property's guest relationships.

Reliability, Hallucinations, and Operational Control

12. Can Una Hallucinate or Give Incorrect Information?

Like other systems using large language models, Una is exposed to the possibility of inaccurate AI-generated responses.

Polydom's architecture is designed to reduce this risk by separating language generation from verified operational information.

The language model is not intended to serve as the authoritative source for room availability, reservation details, prices, or property policies.

Instead, Una uses information from connected systems, configured property rules, and controlled workflows.

These safeguards reduce risk; they do not justify a claim that errors are impossible.

Operators should still use appropriate testing, monitoring, and escalation procedures.

13. Why Do Large Language Models Sometimes Produce Incorrect Answers?

Large language models generate responses based on learned patterns and the information available in their context.

They can produce plausible statements that are factually incorrect, incomplete, or unsupported.

In hospitality, such errors can create operational consequences.

An incorrect room rate, cancellation condition, or access instruction can affect a guest's booking or experience.

This is why an AI system used for reservations and operations needs more than a capable conversational model.

It also needs reliable data sources, defined permissions, validation, and procedures for handling uncertainty.

14. How Does Polydom Reduce the Risk of Hallucinations?

Una is designed around several operational control principles:

• Verified information: Critical booking and property data should come from connected systems or configured rules.
• Defined permissions: Actions are limited by the authority available to the relevant workflow.
• Validation: Supported actions are subject to applicable checks and operational conditions.
• Clarification: Missing or ambiguous information can trigger additional questions.
• Human escalation: Requests requiring judgment or unsupported actions can be handed to employees.
• Monitoring and traceability: Conversation and operational records can support review and improvement.

These controls help make Una more suitable for real hospitality workflows than relying on unconstrained language generation alone.

However, reliability should be assessed through the specific deployment, not assumed from the architecture's design.

15. Can Una Invent a Room Price or Confirm Availability Without PMS Data?

Una is designed not to invent operational facts such as room prices, live availability, or reservation status.

Those details should be retrieved from an appropriate connected system or an approved source of property information.

If the necessary information is unavailable, the expected behavior is to ask for clarification, explain the limitation, or escalate the request.

For example, if the PMS does not return a valid rate for a particular booking, Una should not manufacture a price to complete the conversation.

Operators should test these scenarios during implementation to confirm that the configured workflow behaves as intended.

16. What Happens if Una Doesn't Have Instructions for a Guest Request?

When a request falls outside the property's configured procedures, Una is designed to avoid taking unsupported action.

Depending on the situation, she may:

• Explain that the required information or action is unavailable.
• Ask the guest for additional details.
• Escalate the request to an authorized employee.

For instance, if a guest asks for an exception to a cancellation policy, Una should not independently approve it unless the relevant authority and workflow have been configured.

This helps distinguish routine execution from situations requiring human discretion.

17. Does Una Make Independent Decisions?

Una executes supported actions within configured workflows, permissions, and property rules.

She is not intended to make unrestricted business decisions on behalf of the hotel.

For example, Una may be authorized to process a standard reservation modification.

That does not automatically authorize her to waive a cancellation penalty or approve an unusual refund.

The property's operating procedures determine which actions can be completed automatically and which require employee approval.

The objective is controlled operational execution, not unlimited autonomy.

18. Does Una Learn Automatically From Live Guest Conversations?

Una is not designed to modify her production behavior through uncontrolled self-learning from guest conversations.

Changes to prompts, policies, workflows, and models are handled through controlled updates and testing.

This matters because hospitality operators need predictable behavior.

A guest's unusual request should not automatically become a new rule applied to future guests.

However, the absence of autonomous self-learning does not eliminate the need for monitoring or guarantee that behavior will remain perfect.

Changes in property data, integrations, or external systems can still affect results.

19. How Does Polydom Detect and Address Incorrect Responses?

Polydom describes Una's reliability approach as combining workflow controls, system-based verification, action validation, operational records, and conversation review.

These mechanisms can help identify situations where:

• Information was missing or inaccurate.
• A workflow did not complete as intended.
• A request should have been escalated.
• A property rule requires clarification.
• An integration or configuration needs adjustment.

Errors identified during monitoring can inform controlled improvements.

Operators should establish a review process appropriate to the workflows Una is authorized to perform.

20. Is Una Suitable for Important Hotel Operations?

Una is designed for defined hospitality workflows involving guest communication, reservations, and operational task coordination.

However, suitability depends on the specific operational requirement, connected systems, testing results, and available safeguards.

A routine booking inquiry and an urgent safety-related situation involve different levels of risk.

Operators should establish clear boundaries for automated actions and maintain appropriate human escalation.

Una should be evaluated as part of the hotel's overall operating environment rather than treated as a universally autonomous solution for every situation.

Security, Privacy, and Ongoing Maintenance

21. How Does Una Protect Guest Privacy on Kiosks or Shared Devices?

Una's Kiosk Mode includes functionality designed to clear conversation history after a session or configured inactivity period.

This helps reduce the risk of a subsequent user viewing information from the previous guest's interaction.

For example, a guest using a lobby tablet should not be able to review another guest's reservation discussion simply because the device was left open.

However, session clearing is only one part of a secure shared-device deployment.

Properties also need appropriate device configuration, access controls, data-handling procedures, and privacy safeguards.

Kiosk Mode should not be presented as a guarantee of complete privacy or regulatory compliance.

22. What Does Security Mean in Una's Architecture?

Una's operational security approach includes controlled access to information, defined workflow permissions, validation, escalation procedures, and records that can support traceability.

In practical terms, a hotel should understand:

• Which systems Una can access.
• Which information she can retrieve.
• What actions she is authorized to perform.
• When staff approval is required.
• How activity and exceptions can be reviewed.
• How guest information is handled across supported channels.

Technical security and privacy arrangements should be reviewed against the applicable agreement, product documentation, and the operator's requirements.

For formal privacy information, see Polydom's Privacy Policy and the relevant contractual documentation.

23. Why Does Una Need Updates?

Una operates within a changing technology and business environment.

Updates may be necessary when:

• PMS or communication platform APIs change.
• Property policies or operating procedures are revised.
• New integrations or supported capabilities are introduced.
• A workflow requires improvement.
• Testing identifies an error or operational limitation.
• Underlying AI models or supporting infrastructure change.

These updates help keep the configured system aligned with current requirements.

They should be managed and tested rather than introduced through uncontrolled changes in production behavior.

24. Can Una's Performance Change Over Time?

Yes. Performance can change even when the AI is not autonomously learning from guest conversations.

For example, an external API may change, property information may become outdated, or a previously reliable workflow may encounter a new type of request.

Changes to underlying models or configurations can also affect behavior.

Una's controlled release process is intended to make changes more manageable and testable.

But controlled updates do not eliminate all risks.

Ongoing monitoring remains important for maintaining reliable operations.

25. Can Una Pick Up Bad Habits From Guests?

Una is not designed to autonomously rewrite her operating instructions based on conversations with guests.

This reduces the risk of uncontrolled behavioral changes caused by individual interactions.

For example, a guest requesting an unauthorized discount should not cause Una to permanently change the hotel's pricing policy.

Nevertheless, AI systems still require safeguards against misleading instructions, incorrect information, and attempts to manipulate workflows.

Defined permissions, system validation, escalation rules, and controlled updates are important parts of that protection.

Business Outcomes and Implementation

26. What Business Results Can Hotels Expect From Una?

Una is designed to support several operational outcomes:

• 24/7 operating capacity: Handling supported guest communication outside staffed hours.
• Revenue capture: Supporting direct booking inquiries and authorized reservation execution.
• Labor leverage: Reducing repetitive communication and administrative steps.
• Operational consistency: Applying configured procedures across supported workflows.
• Task coordination: Helping requests reach the relevant operational team.
• Human capacity reallocation: Allowing employees to focus more attention on exceptions and personal service.

The measurable impact depends on the property's existing workload, technology stack, operating procedures, and deployment scope.

Operators should evaluate actual completed bookings, human escalations, task outcomes, and staff workload rather than rely on generic AI productivity claims.

27. Has Una Delivered Measurable Results for Real Hotels?

Yes. Lee Gardens Extended Stay in Orlando, Florida, provides a documented example.

During a ten-month reporting period, the property recorded:

• 4,104 guest conversations handled by Una.
• 2,475 phone calls handled.
• 232 direct reservations made by Una.
• $80,968 in reported direct booking revenue associated with those reservations.
• 172 operational tasks generated for employees.
• 80 conversations requiring human interaction, approximately 1.95% of the total.

The 98.05% conversation-level automation rate does not mean that 98.05% of all hotel operations were automated.

Likewise, the reported booking revenue does not independently establish incremental revenue or net profit.

These figures illustrate what Una accomplished within a specific hotel deployment and reporting period.

Explore the complete Lee Gardens Extended Stay customer case study.

28. Does Una Replace Front Desk or Reservation Employees?

Una is not positioned as a replacement for the hospitality professionals responsible for guest relationships and service quality.

Her role is to handle supported, repeatable workflows that would otherwise require employees to perform routine administrative work.

A human team remains responsible for:

• Complex guest complaints and sensitive situations.
• Exceptions requiring professional judgment.
• Physical service delivery.
• Operational management and quality control.
• Relationship-building and personal hospitality.

The objective is to give employees more capacity for responsibilities where their judgment and experience matter most.

29. How Is Una Implemented at a Hotel or Vacation Rental Business?

A typical implementation includes five stages.

1. Define the scope. Identify guest communication channels, booking workflows, operational tasks, and escalation requirements.

2. Connect systems. Establish the appropriate PMS, communication, and operational integrations.

3. Configure property knowledge. Provide relevant policies, standard operating procedures, guest information, and workflow requirements.

4. Test representative scenarios. Verify booking actions, guest responses, exception handling, and system behavior.

5. Launch and monitor. Introduce Una into the agreed operating environment and review performance, exceptions, and required improvements.

The time and technical effort depend on the integrations, configuration, and scope of deployment.

30. How Much Does Una Cost?

Una's commercial terms depend on the deployment agreement and the property's requirements.

Relevant factors can include portfolio size, communication channels, expected usage, supported workflows, and integration scope.

Operators should request current pricing and confirm what is included in implementation, ongoing service, and support.

Historical promotional offers should not be assumed to represent current pricing.

31. How Can I See Una in Action?

You can explore Una through Polydom's interactive voice demonstration or speak with the team about a real hospitality workflow.

For a meaningful evaluation, bring a scenario your employees handle regularly, such as:

• An after-hours reservation inquiry.
• A booking modification.
• An early check-in request.
• A maintenance issue requiring task creation.
• A situation requiring human escalation.

Ask which steps Una can complete through your specific systems and which remain the responsibility of the hotel team.

That provides a more useful assessment than evaluating conversational quality alone.

The Most Important Question: What Work Can Una Complete?

A hotel AI chatbot can provide fast answers. A virtual concierge can make information and services more accessible. An AI receptionist can support continuous telephone coverage.

All of these capabilities have value.

But for many hospitality operators, the larger workload begins after the guest asks a question.

Someone needs to modify the reservation, create a maintenance request, check availability, apply a property policy, or coordinate the next operational step.

Una by Polydom was designed to address that execution gap.

As a voice-first AI Digital Employee, she works within supported hospitality systems to carry out authorized workflows while involving human employees when judgment or approval is needed.

The goal isn't simply to answer more guest questions. It's to reduce the amount of routine work that still needs to be done after those questions are answered.

Still Have Questions About Una?

Our team can help you understand which integrations, communication channels, and operational workflows are available for your property.

Email us at una@polydom.ai, try the Una voice demo, or book a call with our team to discuss your hotel's specific requirements.