precisión
Quick answer: How do I control what my AI says to guests?
With AI policies: plain-English rules that describe how your business handles each situation - arrivals, pets, refunds, noise, verification. Before every reply, EnsoAI reads the policy that covers the guest's situation and answers with your rules and your voice, whether it's running as CoPilot or AutoPilot. If a policy says a human should take it, or nothing covers it, the conversation is handed to your team through AI Escalations with full context in the Unified Inbox. And you don't have to write them all yourself: EnsoAI studies how your team already answers and drafts policies unique to your business, which you approve before they go live.
AutoPilot no hace suposiciones. Lo comprueba.
Cuando un huésped pregunta «¿Puedo llevar a mi perro?» o «¿Hay aparcamiento para dos coches?», AutoPilot busca la política exacta de ese anuncio específico antes de responder. No una respuesta general, sino la respuesta correcta para esa propiedad.
Si una política no existe, AutoPilot no la inventa. Marca la laguna para que pueda rellenarla. Con el tiempo, su biblioteca de políticas cubrirá todo lo que los huéspedes suelen preguntar, y cada respuesta será precisa, siempre.
Las políticas se pueden asignar de forma global en toda su cartera de propiedades o limitarse a anuncios individuales. Una sola actualización actualiza inmediatamente cada respuesta afectada.
A situation arises
"Can we check in at noon?"
"Is our dog okay to bring?"
"Neighbors are complaining about noise"
policy check
Your policy covers it → it answers
Instantly, with your rules, your fees, your tone.
It's beyond policy → it hands off
The AI doesn't wing it. Your team gets the conversation with a note, a tag and a notification. How AI Escalations work →
PROCESO
Escrito una vez. Referenciado en cada mensaje relevante.
No code, no decision trees. A policy is a few plain-English sentences about how you handle a situation. EnsoAI recognizes the situation in any phrasing, in any language, and applies your rule.
A policy, written in plain English
"We welcome up to two dogs under 40lbs with a $75 pet fee. No cats. If a guest booked with a pet, confirm the fee was paid and share the dog park two blocks away."
Arrival & Departure
Check-in, verification, late check-out
Pets
Fees, size limits, pet-friendly properties
Neighbors & Noise
Quiet hours, complaints, manager alerts
Refunds & Cancellations
What the AI may offer, what goes to a manager
Apply
The same playbook, in every conversation
EnsoAI combines your policy with the guest's actual booking: dates, listing, purchases, and writes a natural reply.
Learn
Your AI studies how your team already answers
Your playbook already exists, scattered across thousands of past conversations. EnsoAI finds the patterns in how your team handles situations and drafts policies from them, unique to your business. You review, edit and approve each one.
Nothing goes live without your approval.
Suggested policy - drafted from 214 conversations
Accuracy
Hospitality judgment, tuned by your rules
Recognizing that "any chance we could drop bags early?" is an arrival question takes real hospitality understanding. Your policies supply the rules; this is the judgment underneath.
50M+
hospitality conversations behind its judgment
60,000
listings shaped the policy templates it starts from
Frontier
AI models read your policies on every reply
"After updating my parking policies, EnsoAI started answering parking questions even better than I would myself."
Martin Milette
Owner, Luxe Property Managements
FAQ
Questions about AI Policies
What are AI policies?
AI policies are plain-English rules that tell EnsoAI how a business handles specific guest situations: arrivals, pets, refunds, noise, verification. Before every reply, the AI consults the policy library and answers the way the operator would, or hands the conversation to the team when the rules say a human should take it.
How does the AI decide whether to answer or hand off?
Every guest message is checked against the policies. If a policy covers the situation, EnsoAI answers instantly with the operator's rules and voice. If the policy says the situation needs a human, or no policy covers it confidently, it hands off: the team is notified with context and the guest is told what happens next.
Do I have to write every policy myself?
No. EnsoAI studies how the team already answers guests and drafts policies from those conversations. Each one is reviewed, edited and approved before it goes live. Any policy can also be written or edited directly, in plain English.
Can a policy apply to just one listing?
Yes. Policies can apply to all listings or be scoped to specific properties, so the lakehouse can allow pets while the downtown condo does not, and each property answers by its own rules.
What happens when no policy covers a situation?
The AI does not guess. The conversation is handed to the team through AI Escalations: the guest gets a holding reply the operator has approved, a note explains what came up, and the right person is notified. Unhandled situations are also how new policies get suggested.
How are policies different from canned responses?
A canned response is one fixed answer to one predicted question. A policy is how an operator handles a situation: the AI applies it to any phrasing, combines it with booking details like dates and purchases, and writes a natural reply. One arrival policy covers hundreds of differently worded questions.
