What if a General Manager could know exactly what the hotel needs before a large group arrives?

It is Thursday afternoon.

The GM gets a call.

“A large group is arriving on Saturday. They are going to push occupancy significantly for the next five days.”

That sounds like good news.

And it is.

But every experienced hotelier knows that a large group can also create pressure across almost every department.

More arrivals.

More luggage.

More rooms to clean.

More breakfast covers.

More airport transfers.

More requests at Front Office.

More pressure on Engineering.

More movement through the lobby.

And, inevitably, something unexpected.

The GM doesn’t need another report saying:

“Occupancy will be 92%.”

The GM needs to know:

“Are we actually ready?”

So the GM asks Zoya.

“We have this group arriving Saturday. Prepare me for the next five days. What could go wrong, and what do we need to do now?”

Zoya starts connecting the dots.

She reviews the forecast.

Room inventory.

Arrivals and departures.

The group’s rooming list.

VIP arrivals.

Housekeeping staffing.

Front Office staffing.

F&B requirements.

Airport transfers.

Open maintenance issues.

And the operational information already available across the hotel.

Then she starts looking for bottlenecks.

The occupancy number is not the whole story.

Suppose the hotel has 280 rooms and occupancy is expected to reach 94%.

On a dashboard, that looks straightforward.

But Zoya may discover something different.

The group has a large number of arrivals within a short window.

Several rooms are still marked for maintenance.

Housekeeping has fewer people scheduled on the busiest turnover day.

The breakfast restaurant may not have enough capacity for the group’s expected morning peak.

Airport transfers are already heavily booked.

Several VIP guests are arriving independently on the same day.

And there is a known Engineering issue affecting a block of rooms.

None of these problems necessarily appear in one place.

But together, they create operational risk.

That is where a GM needs help.

Zoya doesn’t just identify the problems.

It turns them into a plan.

For example:

Front Office

Prepare group check-in process, pre-assign rooms where appropriate, confirm rooming list and establish a dedicated escalation point.

Housekeeping

Rebalance staffing for peak turnover days, prioritize group rooms and identify rooms requiring Engineering clearance.

F&B

Review expected breakfast and meal volumes, staffing requirements, table capacity and group-specific requirements.

Engineering

Resolve priority room issues before the group’s arrival and establish an escalation plan for incidents during the stay.

Transportation

Confirm airport transfers, vehicle capacity, arrival schedules and contingency arrangements.

Security

Review expected lobby traffic, group movement and any special security requirements.

And then Zoya brings everything back to the GM.

“Here are the five biggest risks for the next five days, what we know about each one, what remains uncertain, and what I recommend doing today.”

That is much more useful than another spreadsheet.

And the GM can challenge the plan.

Maybe Zoya recommends adding transportation capacity.

The GM knows that the group has arranged its own buses.

So the GM says:

“Transportation is already covered. Remove that risk and focus on Housekeeping and F&B.”

Zoya updates the plan.

Or perhaps the GM says:

“The group has several VIPs. Treat their arrivals separately.”

Again, Zoya incorporates that context.

This is important because AI should not pretend it knows the hotel better than the people running it.

It should bring analysis, preparation and options.

The GM brings judgment, relationships and context.

Together, they can make a better plan.

The five-day horizon is where things get interesting.

A hotel GM is constantly dealing with tasks that don’t fit neatly into a single transaction.

A group arriving Saturday affects Friday.

Friday affects Thursday.

Today’s staffing decision affects tomorrow’s room readiness.

Today’s Engineering issue can become Saturday’s Front Office problem.

A VIP arrival changes how the lobby needs to operate.

A group breakfast changes F&B staffing.

A rooming-list change can affect Housekeeping, Front Office and Transportation simultaneously.

These are long-horizon hotel tasks.

They require the agent to move across departments, systems and time.

That’s why we are building Zoya around these kinds of scenarios.

The goal isn’t simply:

“Ask AI a question and get an answer.”

The goal is:

“Give AI a complicated hotel responsibility and let it work through the situation with you.”

It should investigate.

Reconcile information.

Identify risks.

Build a plan.

Assign actions.

Monitor what changes.

And come back to the GM when something needs attention.

What happens when the data isn’t perfect?

The hotel world is not perfectly integrated.

There will be missing information.

Conflicting records.

Late updates.

Things that exist in someone’s head but haven’t made it into a system.

That’s normal.

Zoya should not hide that uncertainty.

If it cannot confidently determine whether a particular operational bottleneck will occur, it should say so.

Perhaps:

Option 1: Add staffing capacity — lower operational risk, higher cost.

Option 2: Reallocate existing staff — lower cost, but creates pressure elsewhere.

Option 3: Maintain current staffing and introduce an escalation plan — lowest cost, higher operational risk.

Then the GM decides.

That’s the relationship we envision.

Zoya prepares the paths. The GM chooses the direction.

Imagine the GM briefing on Friday morning.

Instead of spending hours collecting updates from six department heads, the GM has a concise briefing:

Group readiness: 87%

Highest risks:

1. Housekeeping turnover on Saturday
2. Breakfast capacity on Sunday
3. Four unresolved Engineering issues

Recommended actions:

• Reallocate two Housekeeping team members
• Add breakfast service capacity
• Resolve priority Engineering issues before Friday evening

Department owners:

• Housekeeping → action A
• F&B → action B
• Engineering → action C
• Front Office → action D

GM decisions required:

• Approve temporary staffing adjustment
• Confirm VIP handling approach

The GM still talks to the department heads.

Still walks the hotel.

Still makes the decisions.

But now the GM starts the conversation prepared.

That matters.

Because when the hotel is busy, the GM’s most valuable resource isn’t another dashboard.

It is attention.

Attention for guests.

Attention for staff.

Attention for the unexpected problem that cannot be predicted by any system.

This is the kind of hospitality AI we believe in.

Large groups should be exciting for a hotel.

They should represent revenue, reputation and an opportunity for the team to deliver something memorable.

They shouldn’t automatically mean chaos behind the scenes.

If AI can identify the pressure points five days before they become problems, coordinate the work across departments and give the GM clear choices, then it is doing something genuinely useful.

Not replacing the Front Office Manager.

Not replacing Housekeeping.

Not replacing the F&B team.

Not replacing Engineering.

Helping all of them work together better.

That is the operating lens we are taking with Zoya.

We want hotel teams to have an AI buddy they can talk to about complicated situations — one that understands the hotel’s context, thinks beyond a single department and helps them decide what to do next.

Because the best outcome isn’t an AI that runs the hotel.

It is a hotel team that has more time, less unnecessary coordination and fewer surprises.

What do you think?

If you were a GM preparing for a large group arrival, what would you want an AI agent like Zoya to analyze, and what decisions would you always want to keep in human hands?

I’d genuinely like to hear from hotel professionals. Write your opinion to me. What do you think about this approach to AI in hotel operations?