Behind Every Seamless Hotel Operation: In Conversation with Sirma’s Antonia Teofilova

A reservation that fails to synchronise. A rate that does not map correctly. An established interface disrupted by a platform migration. In hospitality, gaps between systems can quickly affect hotel teams and the guests they serve.

In an interview, Antonia Teofilova, Head of Support & Managed Services at Sirma Travel & Hospitality, shares how her team keeps hospitality technology connected, why operational experience matters, and where AI is helping support engineers work more effectively.

From hotel operations to technology support

Who does your team support, and why is its background in hotel operations so valuable?

Our customers are technology providers building revenue systems, guest engagement tools, payments and operational software. We support the integrations between their products and the property management systems used by their hotel customers, carrying reservations, rates, profiles and charges between them.

Many of our engineers began in Front Desk, Reservations or Revenue Management. They know what it feels like when a PMS integration fails at 11pm during a full house, because they have lived it from the other side of the desk.

We train hoteliers as application support engineers, building technical expertise on that operational instinct. They also help our Development and QA teams translate real hotel scenarios into something that can be investigated, built and tested.

Our work spans monitoring, troubleshooting, onboarding and rollouts. Increasingly, we also bring customers’ existing integrations under a consistent support model.

The value of the incident that never happens

Support is often judged by how quickly a ticket is resolved. What does that miss?

Our service desk is built around the idea that the best ticket is the one that is never raised.

When prevention works, nobody notices the outage that did not happen. Yet every incident carries information: a configuration pattern, a system boundary or a process that hotel staff were never properly trained on.

A closed ticket and a solved problem are not the same thing. We look at what the incident cost the whole process, not just our share of it - which often means chasing another vendor’s fix long after our own part is done. Generally speaking, that is outside our scope. We have never found that a useful place to stop, however.

Furthermore, what we learn feeds back into monitoring, configuration, integration design and our technology partners’ future releases.

Getting onboarding right before go-live

What separates a successful property rollout from one that creates problems later?

Much of the work that determines success happens before installation.

We need to understand the property’s configuration, its dependencies and what each third party needs to deliver. An API may behave differently in production. Interface provisioning may take longer than expected. A seemingly familiar property may have an undocumented legacy configuration.

Those are familiar situations. The important thing is to account for them in the plan.

We would rather have that conversation at kick-off than agree to a schedule we already know will slip.

Continuity matters too. The people who understand the property during setup remain involved once it is live, preserving knowledge that could otherwise be lost during a handover.

Taking ownership across connected systems

What happens when an issue involves several systems and each provider believes the cause sits elsewhere?

The hotel still needs a resolution, regardless of where the fault sits. The technical diagnosis is often only half the job. The other half is keeping several parties working on the same problem.

We establish the operational impact, trace the issue across systems and check whether other properties could be affected. We document the investigation as it happens - not just for the audit trail, but to prevent the analysis from restarting every time it changes hands.

From there, we keep the relevant teams engaged and customers informed, following up with whoever is responsible for the fix.

Good service goes beyond closing a ticket or establishing that the fault lies elsewhere. It means helping the investigation move towards a resolution, even when the next step depends on another party.

Maintaining continuity through change

How are cloud migrations and AI changing what customers need from support?

Customers are introducing new technology while relying on connections that have worked for years.

During a cloud migration, historical data may move into fields that do not quite match. It can be technically present but difficult to use for reporting, forecasting or personalization.

Established integrations may also need rebuilding while their existing versions remain operational. Essential interfaces cannot simply stop working during that transition.

AI adds another dependency. If an AI capability relies on incomplete or delayed information, the underlying integration may be where the problem starts.

What all of this takes is a team that has watched these integrations under real operational load long enough to know which data mappings will quietly degrade and which interfaces cannot afford a gap. That knowledge is what keeps things running while platforms, interfaces and teams change.

Where AI helps and experience matters

Where is AI making a practical difference within your own team?

It helps us correlate symptoms across systems, find patterns in tickets and draft an initial root-cause analysis.

It also helps us build support tooling, including scripts and monitoring improvements, reducing the distance between noticing a problem and doing something about it.

But judgment remains essential. Deciding which record to trust when two systems disagree requires context. An unusual property configuration may reflect a decision made years ago for a reason nobody documented.

AI helps us get to the right question faster. Reaching the right answer in a live hotel environment still depends on people who understand the operation.

Choosing a support partner

What should a hospitality technology provider ask when evaluating support options?

Response times, coverage hours and escalation processes matter, but they do not tell the whole story.

I would ask three things. What happens when an issue spans several systems and the cause is unclear? What does the team do to prevent a resolved problem from recurring? And does it understand your product well enough to investigate properly before escalating -including telling you when the problem is your design rather than a defect? That last one is uncomfortable, and it is the one worth most.

Then ask for a real example: walk us through the last time something serious went wrong.

The answer will tell you whether you are talking to a vendor closing a case or a team that is as invested in the outcome as you are.

__________________________________

As hospitality businesses connect more systems and introduce AI, the experience of the people supporting that technology becomes increasingly valuable. Antonia’s perspective brings that back to everyday practice: understanding the hotel, preserving knowledge through change and using each resolved issue to prevent the next.