Why Una Uses Frontier OpenAI Models

Originally published November 24, 2025. Updated October 15, 2026.

Why Una Runs on Frontier OpenAI Models—and Why That Matters for Your Property

Editor's Note: Artificial intelligence continues to evolve, and so does the technology powering Una. At Polydom, we regularly evaluate advancements in OpenAI models to ensure that Una benefits from improvements in reasoning, language understanding, and operational reliability.

Our approach is not about attaching Una to a specific model generation. It is about continuously improving what Una can accomplish for hospitality operators.

Because ultimately, the question isn't which AI model powers Una. It's what Una can do for your property.

Hospitality Needs AI That Keeps Getting Better

Imagine hiring an employee who never learns anything new.

Their knowledge remains fixed. Their ability to solve problems never improves. Every new situation requires additional instructions or someone else stepping in.

No hospitality operator would deliberately build a team that way.

Yet technology can create a similar limitation when an AI product relies on capabilities that are never meaningfully updated.

Hospitality is a dynamic industry. Guest expectations change, portfolios expand, operational procedures evolve, and new challenges appear every day.

That's why we believe the intelligence behind Una should continue evolving alongside the businesses it supports.

By building Una on advanced OpenAI models, we can incorporate improvements in AI capabilities as they become suitable for real hospitality operations.

But technological progress is only valuable when it produces practical results.

For a guest, that might mean receiving an accurate answer without repeating a question. For a property manager, it might mean fewer interruptions, smoother reservation workflows, and more consistent execution of company procedures.

Why We Build Una on OpenAI's Frontier Models

OpenAI develops advanced models capable of processing natural language, interpreting complex instructions, reasoning through unfamiliar situations, and maintaining conversational context.

These capabilities are particularly important in hospitality, where a seemingly simple request often involves multiple systems, rules, and decisions.

At Polydom, we continuously evaluate developments in AI technology and incorporate improvements into Una when they meet our operational requirements.

This approach allows our product to evolve without requiring hospitality operators to manage model selection or technical upgrades themselves.

However, we don't assume that every newly released model is automatically the right choice for every workflow.

Different models can offer different advantages in reasoning, responsiveness, reliability, and efficiency.

A guest calling reception expects an immediate response. A complex reservation change requires careful interpretation and validation. A policy exception may require additional reasoning or human approval.

Our responsibility is to select and configure the technology that best supports these situations.

The objective is not to have the most impressive model name. It is to deliver an AI Digital Employee that performs reliably in real hospitality environments.

Better Reasoning for Real Hospitality Situations

Hospitality operations rarely follow perfectly predictable scripts.

Consider a guest asking whether they can bring a pet.

For a single property, the answer may be straightforward. For an operator managing 100 vacation rentals, the situation becomes more complicated.

Some properties may allow pets without restrictions. Others may charge additional fees, limit animal sizes, or prohibit pets entirely.

The correct answer depends on the specific property, reservation, and applicable policies.

Una needs to understand the question, identify the relevant information, and respond according to the operator's instructions.

The same principle applies to more complex requests.

A guest might ask to extend a reservation, reserve several rooms under a corporate agreement, or arrange an early check-in subject to availability.

Each situation requires more than producing a convincing conversational response.

It requires understanding context, recognizing operational boundaries, and determining which actions are permitted.

Advanced reasoning capabilities help Una interpret these situations. Connected property systems, configured workflows, and clear permissions make reliable execution possible.

Both are essential.

From Understanding Requests to Completing Work

Many conversational AI tools are designed primarily to answer questions.

A guest asks about a reservation, and the chatbot explains where to find the information.

Someone requests a booking modification, and the chatbot provides an email address.

A traveler reports a maintenance problem, and the chatbot recommends contacting reception.

The conversation may be successful, but the underlying work remains unfinished.

Una is designed around a different principle:

Tools need operators. Una is the operator.

Rather than functioning solely as a conversational interface, Una operates within supported hospitality systems and configured workflows.

Depending on the property's integrations and permissions, Una can retrieve reservation details, assist with booking actions, coordinate guest requests, route operational tasks, and involve human employees when necessary.

Consider a guest who wants to extend their stay.

Una may need to identify the existing reservation, verify availability, check applicable policies, explain available options, and support an authorized modification.

The value isn't simply that Una understands the request.

It's that understanding can lead to meaningful operational action.

This is where advanced AI technology becomes useful for hospitality businesses: not as a standalone feature, but as the intelligence supporting an integrated digital employee.

Guests Don't Speak in Perfectly Structured Instructions

Real conversations are messy.

Guests interrupt themselves, change their minds, provide incomplete information, and ask several questions at once.

A traveler might begin a phone call in English and switch languages. Another might send a WhatsApp message referring to a reservation without providing the booking number.

Someone may ask about availability for two people, then mention that a child will be joining them.

These details change the request.

Una needs to recognize those changes, maintain the available conversational context, and respond using the property's actual information.

Advanced language understanding helps make these interactions more natural and productive.

However, conversational fluency should never replace operational accuracy.

An AI Digital Employee must use verified information from the relevant systems and follow the property's configured rules.

The objective is not simply to make an interaction sound more human.

It's to minimize misunderstandings, unnecessary repetition, and avoidable work for both guests and employees.

Supporting Revenue Opportunities Around the Clock

Guest inquiries don't stop when the front desk closes.

A traveler researching accommodation late at night may want to compare room categories, confirm cancellation conditions, or discuss a direct booking.

If no one responds until the following morning, that opportunity may disappear.

Una provides continuous guest communication coverage and can support eligible reservation workflows outside traditional staffing hours.

More sophisticated AI reasoning is particularly useful when booking conversations become complicated.

For example, a guest may mention finding a lower rate through an online travel agency.

A reliable AI Digital Employee shouldn't automatically promise a discount or disregard the hotel's revenue policies.

Instead, Una should interpret the request, retrieve relevant information, and help the guest understand the options available under the property's commercial rules.

For supported Apaleo integrations, Una can assist with presenting appropriate room and rate alternatives and guiding guests toward eligible booking options.

Similar principles apply to extended stays, group bookings, and inquiries involving multiple conditions.

The commercial value comes from connecting guest intent with the appropriate operational workflow.

AI improvements can strengthen that process, but actual revenue impact should always be evaluated through measurable business results.

Built for Growing Hospitality Portfolios

Managing AI for one property is different from managing AI across dozens or hundreds of locations.

Policies vary. Different buildings have different access instructions, maintenance procedures, parking arrangements, and guest services.

Operators may also work with several property management systems, communication channels, and operational tools.

As portfolios expand, maintaining consistency becomes increasingly challenging.

Una is designed to operate within this complexity.

Its intelligence helps interpret requests, while its integrations and configured workflows connect those requests to the relevant business processes.

As advanced AI models improve, Una can benefit from stronger underlying capabilities.

However, model improvements do not automatically update operational information.

A change to a property's cancellation policy, rates, availability, or maintenance instructions still needs to be reflected through the appropriate connected systems or knowledge sources.

This distinction matters.

The quality of an AI Digital Employee depends on both its underlying intelligence and the accuracy of the operational environment in which it works.

We Evaluate Improvements Based on Operational Value

New AI models and capabilities appear regularly.

But adopting technology simply because it is new is not a meaningful strategy for hospitality operations.

A misunderstanding about check-in instructions can leave a guest unable to access a property.

An incorrect interpretation of a cancellation policy can create a dispute.

An unauthorized reservation modification can affect availability, payments, and customer satisfaction.

For these reasons, Polydom evaluates AI improvements according to practical operational requirements.

The questions that guide our approach include:

Can the technology interpret complex hospitality requests more accurately?
Does it reliably follow property-specific instructions?
Can it distinguish between verified information and missing information?
Does it recognize when a request requires human involvement?
Is its response time suitable for real-time guest communication?
Does it work reliably with Una's existing integrations and workflows?

These considerations matter more than a model's release date or version number.

Our objective is continuous improvement that supports consistent, dependable execution.

A technical advancement is valuable when it translates into a better guest experience or more effective operations.

Human Judgment Remains Essential

More advanced AI doesn't eliminate the importance of hospitality professionals.

Some situations require discretion, empathy, local expertise, or decisions that fall outside established procedures.

A guest may dispute an unexpected charge.

A family might experience an urgent issue during their stay.

A reservation may require an exception to company policy.

In these situations, the appropriate response may be to involve a human employee rather than attempt an automated resolution.

Una is designed to support routine communication and authorized operational tasks while recognizing the boundaries of its responsibilities.

When a situation requires additional approval or human judgment, the workflow can include an appropriate handoff.

This allows hospitality teams to concentrate their attention where it matters most.

The purpose of advanced AI in hospitality isn't to remove people from service.

It's to reduce repetitive digital work so that employees can focus on decisions, relationships, and experiences that require their expertise.

Una's Intelligence Will Keep Evolving. Your Operations Should Benefit.

Artificial intelligence will continue to advance.

New capabilities will emerge, existing technologies will improve, and expectations for AI-powered products will evolve.

At Polydom, we view this progress as an ongoing opportunity to improve Una.

Rather than defining our product around a particular generation of AI technology, we focus on the long-term capabilities required to support hospitality operations.

Better reasoning helps Una interpret complex requests.

Reliable integrations connect those requests to real operational actions.

Clear permissions establish what Una can do independently and when human involvement is required.

Together, these elements make Una more than a chatbot that answers questions.

They create an AI Digital Employee designed to support the daily work of hospitality businesses.

Our approach is built on a simple principle:

The technology behind Una will continue evolving. Its purpose will remain the same: helping hospitality operators communicate with guests, manage operational workflows, and deliver reliable service around the clock.