Questions, answered straight.

Event apps own registration, badges and the agenda — and stop at the venue door. Aikebom knows your event just as deeply — the programme, the venue, live changes — and keeps going where the app stops: the whole stay, the city, the trip, by voice, in six languages. It runs alongside your event app — or standalone, if you don’t have one. Complementary, not competing.

No — keep them. Aikebom is a companion layer, not a migration: your app keeps doing registration, badges and the agenda, and Aikebom adds the layer that knows your programme, venue and live changes — and plans the whole visit around them: the trip, the city, voice, six languages and your sponsor layer. Your programme imports from the tools you already run, so nothing gets ripped out. And if you don’t have an app at all, Aikebom is the whole attendee-facing layer on its own.

No — by design. Aikebom is not a ticketing or registration platform; yours keeps doing that. In companion mode the concierge points attendees to your official channels for tickets and registration, and handles the rest itself — your programme and venue, live updates, the whole stay, the city and your sponsors.

The concierge is built to look venues, addresses and opening hours up in live data — Places, ratings and web search — rather than answer from memory, and it is instructed never to invent a place or complete an address it can’t confirm.

No attendee or guest accounts. A session in the browser is anonymous unless a link the guest arrives on identifies them — a verified link you issued, or your own reference for them on the link — and at a property the verified link is the default: the concierge opens only for your own guests, who arrive on a link that tells it at most their booking reference, name, room or seat, and a guest tier if you use one, so there is no passer-by data to collect in the first place. We don’t store voice audio; a call’s transcript joins the conversation and deletes with it, and what the voice providers receive follows their own terms. What happens next depends on the product: an event has one published cut-off date, after which all attendee data is purged automatically. A property never ends, so each guest’s conversation data deletes on its own clock instead — about a week after they stop using the concierge, and never more than 45 days after their last message. Booking details from your property’s system follow the stay instead; our privacy policy has both. Where your PMS reports a guest’s checkout, their access ends shortly after it, so that inactivity clock starts. Usage statistics and short notes survive both; our privacy policy lists them. Either way it’s us who enforce it, not a setting an organiser can quietly widen — that’s part of the commitment. Privacy is built into the schema, not bolted on.

Weeks, not months — and the platform itself deploys in days; loading your content well is what takes the rest. We ground the concierge in your city and brand it; you manage sponsors and partners, announcements and stand QRs yourself from an admin panel. No engineering on your side.

English, Portuguese, Spanish, French, German and Italian today — and the translation is agentic, meaning the whole conversation works in the attendee’s language, not just the buttons. The words that show up verbatim — your welcome message, your example questions, your guide’s tagline, your reception’s contact details — can be written in more than one of those languages too: write in your language, machine-translate, review every string, publish — and in each language you publish, guests read your own words, just in theirs.

Yes — since July 2026 the concierge can live on WhatsApp. Guests message the property’s number and the same concierge answers there: grounded recommendations, staff-confirmed requests, a photograph of what they asked about — sent as its own image message after the reply, where you have published one — and a personal day plan that arrives as a message with its share link — in the guest’s own language, switching the moment they do. Same engine and the same privacy limits as chat and voice; it’s switched on per property. Since September 2026 it also speaks first — on WhatsApp and by email — in two moments: a welcome when a guest checks in, in the language on their booking where you serve it, and a message when your team marks their request done or acknowledged, in the language the guest has been writing in. A STOP on WhatsApp ends those messages on that channel, and request notices we email carry a one-click unsubscribe link. Each is yours to switch per property; the welcome is off until you turn it on, and reaches WhatsApp only where you confirm your guests have agreed to be messaged there. Morning briefings and schedule alerts on WhatsApp are still to come.

Yes — colours, name, tone and voice, and the guides themselves. The personas your concierge can take are yours to specify, each with its own portrait, backstory and way of talking, and a matching voice in every language you serve. At an event, attendees pick the one they want when they arrive. At a property you choose instead: a verified guest’s link opens the conversation already in your guide’s voice, greeting them by name, and where you offer more than one the others sit as a small choice under that first message — tap one and they take over from there. The concierge is unmistakably yours, not ours.

We set your property or event up for you and hand you a no-code admin panel to run it. Setup is white-glove — that is how you go live now.

Both — and they’re two products. Aikebom Concierge AI runs always-on for a hotel or resort: 24/7, all year, for every guest who checks in, sold as a subscription per property. Aikebom Events AI is scoped to one event — anchored on its programme and dates, then purged when it’s over, sold as a per-event license. Same engine underneath; what differs is duration, framing and how it’s priced. A hotel with meeting space typically runs both.

Within a conversation, yes: preferences mentioned once shape every suggestion after — no forms. Across the event, your dashboard shows what attendees actually ask, and every morning a digest flags what the concierge couldn’t answer, grouped by topic with a suggested fix. Fill the gap or publish an announcement and the very next answer improves — live, with no redeploy. And it learns for you, not at your expense: every event is its own tenant in our application, which is built to scope every query to it — and once your event’s published cut-off date has passed, attendees’ conversations, profiles and plans are deleted automatically; what remains is what our privacy policy lists — usage statistics and short notes such as these digest topics and examples, with no fixed deletion date. Our platform administrators can see, across clients, the topics attendees could not get answered and how often each came up, with the events and properties concerned and a download, and use that to decide what public local facts to add to the shared layer every event and property in a place draws on. They see your example questions and suggested fixes only one client at a time: when supporting you, on your own dashboard and reports, as your team sees them.

Yes. Aikebom Concierge AI connects to your property-management system through a vendor-neutral webhook contract: your PMS — or your integrator — pushes bookings, check-ins, room moves and checkouts, and the concierge follows. A guest moved to a new room gets answers about the new room without telling anyone twice. And check-in doesn’t just mint each guest a personal concierge link — it delivers it: an email to the address on the booking, in its language where you serve it, opening straight into a conversation that already knows their name, their room and their dates, plus a short note on WhatsApp where you confirm you collect the consent. The welcome is off until you turn it on. If your system can call a URL, it can talk to Aikebom.

Yes — upload the in-room guide, menus, spa brochures or appliance manuals, or point it at your official pages, and the concierge answers from them with the source attached: “per the in-room guide, page 12”. Your facts always outrank the open internet, and your live announcements outrank both. Photographs and diagrams work the same way: upload them, check the description it drafts, publish — and the concierge shows the right one inside the answer, describing it in words too.

Every property and event can already run its own MCP endpoint — the standard interface AI assistants use to query live services — with partnerships disclosed. It is off unless we switch it on for you, and it requires a key by default. An integration’s key reaches your guest-facing information — your facts, programme, announcements, knowledge and exhibitors, and, where switched on, your partners and their offers — and the public lookups you allow, never guest data. And since July 2026, hotel properties can go one step further: a checked-in guest gets a personal, stay-scoped connection, so their own AI answers “when is my spa booking?” from their actual reservation. Of guest data, it reaches only that guest’s own stay, plan and requests — personal, revocable, and it expires with the stay. Where writing is switched on for the property, the guest’s assistant can act as well, in two ways that differ: it can raise a request with your staff — logged, attributed, and confirmed by your team before anything happens — or it can add a stop to the plan the guest has saved, which is the guest’s own, so that stop is added without anyone on your team being asked.

Yes, and it’s a separate channel from the one your guests use. Alongside the MCP endpoint, every property and event can publish an agent card — the emerging agent-to-agent (A2A) standard — at a fixed address on your own domain, readable by any peer agent without a key. It answers what a buyer asks before booking: where you are, your meeting and function spaces with their capacities, the packages currently running, your accolades, who to call in sales — plus your official partners and their valid offers, with each partnership disclosed. You write that pack yourself, and it’s switched on per client, off by default. Two limits we state plainly: it’s promotional only, so guest announcements and your internal programme are never served there; and it makes you addressable rather than listed — getting into the agent directories is a separate step we take by hand, with your consent.

Not from the desk. The board and the drawer show the request the concierge wrote up, its status and its age, never the conversation. That is by design.

No. The concierge raises a request and never promises what your team hasn’t approved. Your team acts from the desk. Where your property has switched notices on and has a way to reach the guest, the guest hears on WhatsApp or by email when the request is taken and when it is done; when a guest can’t be reached, the desk shows it.

Yes for requests, the Studio, announcements and most of Insights. Arrivals, rooms in house, room moves and the check-in welcome need your PMS connected, through a vendor-neutral contract your integrator can wire.

No. It shows requests, response times, languages and partner reach. It does not estimate money it can’t count.

Not by itself. It never asks a phone for its location. Your team sees only what a guest puts in a request, such as a pickup point. It answers when a guest asks, wherever they happen to be.

As soon as the guest has their link: in a message before arrival, at check-in, or from the desk. A link tied to a booking stops working after checkout (plus any grace the property sets); a link from the desk lasts the days the desk chose, and a phone that opened it can stay signed in after that until its chat is deleted. A guest’s conversations are deleted on the property’s retention schedule.