Why We Built Una: The Origin of an AI Digital Employee

In the fall of 2023, Alexander Udalov, founder of Polydom, was thinking about a problem that nearly every hospitality operator recognizes.

Hotels had become increasingly dependent on software. Property management systems, reservation platforms, CRMs, guest communication tools, and operational applications had become essential to running a property.

Yet every new system introduced another demand on the people expected to operate it.

Employees needed to learn multiple interfaces, understand how information moved between them, and remember the procedures required to complete everyday tasks.

Training took time. Experienced employees carried valuable operational knowledge. And when someone left, much of that knowledge had to be rebuilt with the next hire.

For a hotel managing 50, 100, or 200 rooms, the problem wasn't a shortage of technology.

It was that every piece of technology still needed someone to operate it.

That observation became the starting point for Una by Polydom.

Not another hospitality chatbot. Not another dashboard. But an AI Digital Employee designed to carry out work inside the systems hotels already use.

The Problem That Started It All

Hotel technology has traditionally been designed around a straightforward assumption: a person will use the software to complete a task.

A property management system stores reservations, but an employee still needs to create or modify them.

A guest messaging platform organizes conversations, but someone must interpret the request and decide what happens next.

A task management system tracks maintenance or housekeeping work, but a person still needs to create the task and pass along the relevant information.

These systems perform important functions. The difficulty emerges in the spaces between them.

Consider a guest calling to extend their reservation.

An employee may need to identify the guest, find the reservation, check availability, confirm the rate, update the booking, and communicate the result.

The workflow is familiar. Its steps are often repeatable.

But it still requires someone who knows the property, understands the rules, and can navigate the systems involved.

Now multiply that interaction across hundreds of guest requests, several shifts, and multiple properties.

The operational bottleneck isn't always the software. It's the amount of human coordination required to make the software work together.

For Alexander, this raised a different question.

What if hospitality technology didn't simply help employees complete tasks?

What if a digital employee could take responsibility for some of those tasks itself?

A Different Idea: Build the Employee, Not Another Tool

When the concept behind Una emerged in 2023, advances in generative AI were opening new possibilities for how machines could interpret and respond to natural language.

But conversational intelligence alone wasn't the goal.

The idea was to create something that could understand a request, determine the necessary steps, interact with connected systems, and complete an authorized workflow.

An AI Digital Employee.

The distinction was fundamental.

A conventional chatbot might tell a guest how to change a reservation.

A digital employee, given the appropriate integrations and permissions, could make the change in the property management system.

A chatbot might acknowledge a maintenance problem.

A digital employee could create a structured task and route it to the appropriate team.

A chatbot might answer a question about room availability.

A digital employee could retrieve relevant information and support the booking process.

The purpose wasn't to make conversations more impressive. It was to make them operationally useful.

This became the central philosophy behind Una: tools need operators, and Una is designed to be the operator.

The goal was never to eliminate the role of human hospitality professionals.

It was to separate the repeatable work that technology could execute from the situations where human judgment, experience, and relationships matter most.

Why Building an AI Digital Employee Was More Complicated Than Building a Chatbot

The concept was simple enough to explain.

Building it was another matter.

A hotel employee doesn't operate through isolated questions and answers. They work with incomplete information, changing priorities, property-specific rules, and requests that frequently involve more than one department.

For Una to become useful in real operations, the development team needed to address several interconnected challenges.

1. Understanding What the Guest Actually Wants

Guests rarely communicate in perfectly structured requests.

Someone might call to ask about extending their stay and then mention that they also need parking.

Another guest might ask to change arrival dates while checking whether the new rate includes breakfast.

A message may contain several requests, missing details, or an explanation that doesn't match the language used in the hotel's systems.

An AI Digital Employee needs to identify the underlying intent before taking action.

That requires more than recognizing keywords.

The system needs to determine which information is relevant, what additional details are necessary, and which actions are appropriate.

2. Connecting Conversations to Hotel Systems

Understanding the request is only the beginning.

The next challenge is execution.

A reservation change involves the PMS. A housekeeping request involves an operational task. A booking inquiry may require checking availability, rates, and property conditions.

Una therefore needed to work with the technology hotels already relied on.

Instead of creating another isolated interface, the development approach focused on connecting communication with existing operational systems.

That is what allows Una to move beyond providing information toward performing supported actions.

Learn how Una connects to PMS platforms and hospitality workflows.

3. Knowing When to Involve a Human

Not every guest request should be automated.

A routine reservation modification may follow established rules. A complicated complaint, unusual exception, or sensitive guest situation may require human judgment.

The challenge is knowing the difference.

Una was designed to support a model in which routine workflows can be automated while exceptions and decisions requiring discretion are escalated to employees.

The human team remains responsible for the property's service standards and operational decisions.

Automation should reduce repetitive work, not remove accountability.

4. Maintaining Context Across Interactions

Hotel requests rarely exist in isolation.

A guest may book through an online travel agency, contact the property through WhatsApp, and later call the reception desk.

The information needed to respond may be distributed across different systems and conversations.

One of Una's development challenges was therefore understanding how guest identity, reservation information, communication history, and operational context fit together.

When the appropriate data is available through connected systems, Una can use that context to support more relevant responses and workflows.

This remains dependent on the integrations, access permissions, and configuration of each deployment.

The broader principle is straightforward: a useful digital employee needs to understand not only the latest message, but also the operational situation surrounding it.

The Questions That Shaped Una's Development

Some of the most revealing development questions came from ordinary hospitality scenarios.

What happens when a reservation originates on Booking.com, but the guest contacts the hotel through WhatsApp?

How should the system recognize that a returning guest has an existing relationship with the property?

What should happen when a guest asks about parking, but the request requires more than simply explaining where the parking lot is?

These questions exposed the limitations of treating guest communication as a standalone function.

The issue wasn't merely how to generate a good answer.

It was how to connect that answer with the information, people, and systems needed to deliver the requested service.

Each scenario helped define the operational responsibilities an AI Digital Employee would need to support.

The resulting product had to be adaptable to different property rules, booking procedures, service standards, and technology environments.

A boutique hotel, an extended-stay property, and a professional vacation rental operator may share similar guest communication challenges. But their workflows are not identical.

That is why property-specific configuration and integration became essential parts of Una's operating model.

From an Idea to Real Hotel Operations

The strongest test of an AI Digital Employee isn't whether it can demonstrate an impressive conversation.

It's whether the technology can perform useful work in a live hospitality environment.

Una moved from its original concept into deployments involving hotels and professional accommodation operators.

Those deployments tested the idea against real guest inquiries, reservation processes, communication channels, and operational requests.

At Lee Gardens Extended Stay in Orlando, Florida, the results illustrated how an AI Digital Employee could contribute to both reservations and daily operational activity.

Over a ten-month reporting period, the property reported:

4,104 guest conversations handled by Una.
2,475 phone calls handled.
232 direct reservations made by Una, generating $80,968 in reported booking revenue.
172 structured operational tasks generated for hotel employees.
80 conversations requiring human involvement, approximately 1.95% of all recorded conversations.

These figures reflect one specific deployment, not a universal performance benchmark or guaranteed return on investment.

They also demonstrate an important distinction between conversational automation and operational execution.

Una wasn't only responding to questions. She was participating in booking workflows and helping connect guest requests to hotel staff.

Read the Lee Gardens Extended Stay case study: $81K in Direct Bookings.

What Una Has Become

Today, Una by Polydom is a voice-first AI Digital Employee designed for hospitality operators, including independent hotels, hotel groups, serviced apartments, and professional vacation rental managers.

Her role extends beyond answering guest questions.

Depending on the property's integrations and configured workflows, Una can:

Communicate with guests through supported phone and digital channels, including after-hours.
Retrieve relevant reservation information and check availability.
Create, modify, or cancel reservations through supported PMS integrations.
Help guests complete booking-related workflows.
Capture and route housekeeping, maintenance, and other operational requests.
Escalate exceptions and situations requiring human judgment.

These capabilities are designed to work within the operator's existing technology environment rather than require a replacement of its core systems.

For hotel teams, the practical objective is to reduce repetitive execution and the administrative work involved in coordinating tools, information, and people.

For guests, the benefit can be much simpler: their request moves forward without unnecessary waiting or repeated explanations.

That is the operating model Una was created to support.

For a detailed overview of the product and its current capabilities, see What Is Una? The AI Digital Employee for Hotels & Vacation Rentals.

Why the Human Role Still Matters

When people hear the term digital employee, an obvious question follows: what happens to the human team?

For us, the answer has been part of Una's philosophy from the beginning.

Hospitality involves far more than processing requests.

It requires recognizing when a guest needs reassurance, handling unusual situations, making exceptions, and creating the feeling of being genuinely welcome.

Those responsibilities aren't simply collections of repetitive tasks.

They depend on experience, empathy, accountability, and judgment.

At the same time, valuable human attention is frequently consumed by predictable administrative work.

Employees answer the same questions, transfer information between systems, look up reservation details, and coordinate routine requests throughout the day.

The purpose of Una is to take on supported, repeatable execution so employees can devote more of their attention to work that benefits from their expertise.

A digital employee can help maintain operating capacity across shifts and communication channels.

A human employee remains essential for the decisions and relationships that define hospitality.

The ambition isn't hospitality without people. It's hospitality where people spend less time operating software and more time serving guests.

What We Learned About the Future of Hospitality Work

Looking back at the original idea from 2023, one observation remains central.

Hospitality doesn't necessarily need more software interfaces.

It needs a better way to turn information into completed work.

For years, technology investment has largely focused on helping employees manage more information: reservations, messages, reports, tasks, and guest records.

AI introduces the possibility of changing that relationship.

Instead of software waiting for a person to initiate every action, a properly configured digital employee can execute defined workflows within connected systems.

That doesn't mean every hotel process should become autonomous.

It means operators can begin identifying which parts of their daily workload are repeatable, rules-based, and suitable for reliable automation.

For an independent hotel, that may begin with after-hours reservations and routine guest inquiries.

For a serviced apartment operator, it may involve coordinating requests across multiple properties.

For a hotel group, it may mean creating more consistent operational coverage without requiring every property to manage the same repetitive work independently.

The business problem may look different, but the underlying question remains the same:

How much of the work surrounding hospitality can technology handle so people have more capacity for hospitality itself?

Why We Created Una

Una began with a frustration familiar to many operators: too much of the daily workload depended on people manually connecting systems that were supposed to make their jobs easier.

Alexander Udalov's original idea was to change that relationship.

Not by introducing another tool that employees would need to learn and supervise constantly.

But by developing an AI Digital Employee capable of working inside the existing hospitality technology stack.

The product has evolved since that first concept. Its capabilities, integrations, and real-world applications continue to develop.

Yet the original purpose remains.

We didn't create Una because hotels needed another way to talk to guests. We created Una because the work behind those conversations needed a better way to get done.

Meet Una by Polydom

Curious how the idea works in real hotel operations?

Talk to our human sales team, explore real customer deployments, and see how Una can support guest communication, reservations, and routine operational workflows.

Book a quick call with our team to discuss what an AI Digital Employee could do within your existing hospitality technology stack.