Frequently Asked Questions About Una

If you’re researching a hotel AI chatbot, virtual concierge, AI host, or digital front desk, these are the questions most hospitality teams ask first.

The difference is that Una by Polydom is not just another chatbot for hotels or vacation rentals. Una is an AI digital employee for hospitality that can communicate with guests, execute workflows across your operational systems, and escalate to staff when human judgment is needed.

The FAQs below explain what Una is, how she differs from an ordinary hotel chatbot, how she works with PMS and other systems, how she handles multilingual communication and why Polydom positions Una as an operator — not just an interface.

FAQ

1. What is Una?

Una is Polydom’s AI digital employee for hotels and vacation rentals. She operates like a teammate: working across guest communication channels and operational systems such as PMS, CRM, and task tools to execute workflows and escalate when needed.

If you are searching for a hotel AI chatbot, vacation rental chatbot, virtual concierge, or AI host, Una fits that need — but goes further. She is designed to do real operational work, not just answer questions.

2. Is Una a hotel AI chatbot?

No. A normal hotel chatbot usually focuses on answering FAQs or sending simple replies.

Una by Polydom is an AI digital employee for hospitality. That means she can help handle guest communication, support booking workflows, dispatch tasks, and operate inside your stack instead of sitting on top of it as a chat-only layer.

In simple terms: a standard chatbot talks. Una helps operate.

3. What does Una do day to day?

Una supports three core areas of hospitality work:

Guest communication She helps manage omnichannel communication across phone or voice and text-based channels such as web chat, email, and messaging apps.

Booking workflows She helps handle the booking lifecycle through PMS-connected workflows, including actions such as create, modify, or cancel, based on verified system data and configured permissions.

Task dispatch and escalation She turns guest requests into structured tasks for teams such as housekeeping, maintenance, or front desk, and escalates when human judgment is needed.

4. How is Una different from a typical hotel chatbot or virtual concierge?

Most hotel chatbots are built to answer common questions, route inquiries, or provide lightweight guest messaging support.

Una is different because she is designed to work across your systems and workflows. Instead of only replying to guests, she can help execute real tasks, coordinate requests, support booking-related actions, and involve staff only when needed.

That is why Polydom uses the category AI digital employee for hotels and vacation rentals instead of “just chatbot.”

5. Which systems does Una integrate with?

Una is designed to work with your existing hospitality stack rather than replace it.

That can include:

• PMS platforms
• CRM systems
• task management tools
• messaging channels
• calendars
• internal operational systems

The important point is not only that Una connects to systems. It is that she is designed to do work inside those systems, using verified operational data and predefined workflows.

6. Does Una work across channels like phone, WhatsApp, email, web chat, and kiosk?

Yes.

Una is built for omnichannel hospitality communication and can support interactions across channels such as:

• phone and voice,
• web chat,
• WhatsApp,
• email,
• social messaging,
• kiosk or guest-facing interfaces (QR code),
• and other digital touchpoints used by hotels and vacation rental teams.

This gives teams more consistent guest communication without requiring staff to manually manage every interaction.

7. Can Una support multiple languages?

Yes.

Una is built for 24/7 multilingual guest communication, helping hospitality teams respond more consistently across languages and reduce the pressure of repetitive guest questions across different markets.

For hotels and vacation rentals serving international guests, multilingual capability is a core part of making automation genuinely useful.

8. What happens when Una needs a human staff member to step in?

Una is designed to handle repetitive and rules-based work automatically while involving staff when human judgment is needed.

If a case requires approval, nuance, exception handling, or a sensitive guest interaction, Una can hand the conversation or task over to the right team member.

This is a core part of the operating model: automate the repetitive work, escalate when judgment matters.

Reliability, Hallucinations, and Control

9. Does Una hallucinate?

Polydom’s answer is that the LLM is not treated as the source of truth.

Una is not designed as an autonomous free-form LLM making unchecked decisions. She is built so the model cannot invent factual operational data such as room availability, pricing, reservation details, or policies.

In other words, Una does not rely on language generation alone for critical operational output.

10. Is it true that all large language models hallucinate by default?

Yes.

Large language models are probabilistic text-generation systems. Under uncertainty, they can produce plausible but incorrect information. Polydom assumes this risk by design and builds Una around control mechanisms that prevent unsupported AI output from becoming the final operational result.

11. How does Polydom prevent hallucinations in practice?

Una operates with practical guardrails such as:

• the LLM is not the source of truth,
• factual data such as rates, availability, policies, reservations, and task status must come from connected systems or configured business rules,
• when confirmed data is unavailable, Una asks for clarification or escalates to a human,
• actions are constrained by workflow logic, permissions, and validation steps.

This is how Una is positioned as a reliable operational system, not an unconstrained chatbot.

12. Can Una invent a price, room availability, or property rules?

No.

Una cannot generate or confirm prices, availability, reservation details, or operational conditions without a valid PMS response or a pre-configured ruleset.

If verified data is not available, she does not guess.

13. What happens if Una doesn’t have instructions for a specific scenario?

If Una encounters a situation that is not covered by the property’s rules, workflows, or approved instructions, she will do one of three things:

• state that the information or action is not available,
• ask for clarification,
• or escalate the case to a human team member.

That is an important difference between an AI operator and an unconstrained LLM chat experience: Una follows operational boundaries and hands off exceptions when human input is required.

14. Does Una make independent decisions?

Una executes predefined scenarios and actions within permitted authority.

Her actions are constrained by:

• business rules,
• connected system data,
• permissions,
• and escalation logic.

She is designed to operate within controlled boundaries, not make unchecked decisions outside policy.

15. Does Una learn autonomously from live guest conversations?

No.

There is no autonomous self-learning in production. Changes to prompts, policies, workflows, or models are versioned, tested, and deployed through controlled releases.

That means Una does not “pick up bad habits,” drift unpredictably, or change behavior without intentional updates.

16. How are incorrect responses prevented or detected?

Polydom uses multiple control layers to improve reliability across guest communication and operations, including:

• validation before actions are executed,
• controlled workflow logic,
• system-based verification,
• review of conversation flows to identify where corrections or escalation are needed,
• and audit logs of responses and completed operations.

This helps both reduce errors before they happen and detect cases where a response, workflow, or action needs adjustment.

17. Is Una suitable for mission-critical hospitality operations?

Yes.

Polydom positions Una as suitable for operational hospitality work because she is designed as an executor of constrained tasks, not as a free-form reasoning system. She operates within policies, system permissions, and escalation paths.

That makes her much better suited to real operational environments than a generic chatbot alone.

Security, Privacy, and Stability

18. How does Una handle guest privacy on kiosks or shared devices?

With Una Kiosk Mode, chat history can be automatically cleared after each session or inactivity period so the next guest does not see private information from a previous interaction.

This is important for privacy, operational hygiene, and reducing legal or reputational risk on shared devices.

19. What does “security” mean in Una’s architecture, in plain language?

In practical terms, security and reliability mean that Una is designed to operate in a controlled and auditable way.

That includes:

• working from verified system data rather than guesses,
• using constraints and escalation when data is missing,
• and maintaining logs of actions and responses for traceability.

For operators, that means more confidence in what the system says and does.

20. Why are updates needed at all?

Updates are needed because the environment changes.

For example:

• PMS APIs may change,
• property rules or rates may change,
• workflows may change,
• operational logic may need adjustment.

In those cases, Una needs to be synchronized with the new reality — just like any employees.

21. Can the quality of Una’s responses deteriorate over time?

No.

Because Una does not autonomously learn from production conversations, she does not gradually drift or become worse over time. If left unchanged, she will not automatically become smarter — but she will also not become less reliable on her own.

22. Can Una get used to mistakes or become unpredictable?

No.

Because there is no autonomous self-learning in production, Una does not absorb mistakes, copy bad patterns, or change behavior unexpectedly.

Her behavior changes only through controlled updates.

Outcomes and Business Impact

23. What outcomes can hotels expect from Una?

The business case for Una is centered on practical hospitality outcomes such as:

• 24/7 guest coverage,
• faster response times,
• less repetitive work for staff,
• more consistent service,
• and more captured booking demand.

One example referenced in your materials is Lee Gardens Extended Stay in Orlando, Florida, where Una reportedly delivered:

• 24/7 coverage
• 100% call answer rate
• 232 direct bookings
• $81,200 in direct booking revenue over 10 months
• with no added staffing

This type of result helps explain why Polydom positions Una not as another software tool, but as an operational AI teammate for hotels and vacation rentals.