The night clerk's most familiar questions are rarely mysteries. What is the Wi-Fi password? When is checkout? Can someone bring up two towels? The hard part is not knowing the answer. It is receiving the request, finding the reservation, checking the property's rules, telling the right colleague and closing the loop while a line forms in the lobby. Akia has built a company around that unglamorous chain of events.
Founded in 2018 by former Facebook engineer Evan Chen and a team with backgrounds in machine learning and consumer messaging, the San Francisco company started with hotel texting. Its product now looks more like an operating layer for the guest lifecycle. Messages arrive by text, WhatsApp or web chat; phone calls can be answered by an AI voice; reservation data comes from the hotel's property-management system; and a request can lead to a reply, a digital form, a staff task, a door code or an offer to buy a later checkout.
That sequence matters. The hospitality software market has plenty of chatbots that can retrieve an answer. Akia wants its agent to finish the job. In the company's vocabulary, the Knowledge Base holds what a property knows - its parking rules, room types, pet policy and preferred recommendations. Skills describe how the agent should behave - who to notify, what action to take and when to hand a conversation to a person.
The hotel app that is barely an app
One of Akia's sharper product choices is to accept that most travelers do not want hotel software. A guest may stay for two nights and never return. Asking that person to visit an app store, create an account and remember a password turns a simple check-in into homework. Akia's Mini Apps open from a link inside the same message thread the guest is already reading. They can collect a registration card, verify an ID, authorize a payment, display a guidebook, sell an upgrade or deliver arrival details without a download or login.
For hotel employees, the appeal is less screen switching. A unified inbox keeps conversations together. Automated journeys schedule messages before arrival, during the stay and after checkout. Digital check-in moves paperwork out of the lobby. Surveys can surface a problem while the guest is still on the property, when recovery is possible, rather than after a one-star review appears.
“Consumers are tired of poorly made chatbots, waiting on hold for call centers, or downloading apps for single use.”Evan Chen, founder and CEO
A labor problem disguised as a messaging problem
Akia's timing was consequential. The company launched commercially in 2019, then watched travel collapse in 2020. Hotels that remained open suddenly needed contactless check-in and communication that did not funnel every guest through a crowded desk. In a funding announcement that year, Akia said 72 percent of guests at participating properties engaged with staff by text; 84 percent of those guests used text to check in. Contactless service moved from amenity to operating requirement.
The crisis faded; the labor constraint did not. Hotels still operate around the clock, often with small teams and seasonal demand. A person at the desk might be checking in a family, answering a call and coordinating housekeeping at once. Akia's basic economic argument is that software should absorb predictable volume. The valuable employee is then available for the complaint, safety issue or peculiar request where judgment and empathy matter.
That framing avoids the clumsiest promise in service automation: replacing hospitality with a robot. Akia instead sells capacity. Revival Hotels, which uses technology to support eight properties in an “Invisible Hospitality” model, reported in an Akia case study that 70 percent of guest communication was automated. It also reported variable operating costs down by more than half and guest satisfaction 24 percent above its comparison set. Those are company-published customer results, not universal forecasts, but they show the operating thesis in practice.
Revival Hotels, reported after deployment
The ordinary chatbot
Finds a canned answer, sends it and waits for the next question.
Akia's ambition
Reads property context, follows a procedure, acts in connected tools and escalates the exceptions.
From the stay to the next stay
The product has also crept into revenue work. A hotel can offer an early arrival, late departure, room extension or amenity at the moment it is relevant. Cashless tipping appears in a conversation rather than on a lonely QR card. The Marketing Suite, announced in 2025, adds email and SMS campaigns built on live property-management data. Instead of exporting stale spreadsheets, an operator can segment past guests and attribute direct bookings to a campaign.
This places Akia at the intersection of several categories that hotels once bought separately: guest messaging, digital check-in, CRM, reputation management, upselling, staff collaboration and voice. Hotel Van Zandt said it replaced three tools with one, while its upsells covered the Akia subscription. Consolidation is a persuasive sales story because each extra vendor creates another login, integration and bill. It is also a demanding product promise. Breadth only helps if every workflow is reliable.
Akia connects with property systems including Oracle Opera, Mews, Cloudbeds, Streamline, Stayntouch, Guesty and WebRezPro, plus smart-lock and operations tools. These integrations form the moat more plausibly than the conversational interface alone. A polished answer to “Can I check out late?” is easy to imitate. Knowing whether the room is available, collecting the correct fee, changing the stay and notifying housekeeping is not.
A crowded new front desk
Competition comes from several directions. Canary Technologies and Duve cover broad digital guest journeys. HiJiffy and Runnr.ai emphasize AI conversation. Revinate is strong in hotel CRM and marketing. Whistle sits inside Cloudbeds, while Operto and Guesty overlap in vacation-rental operations. Hotels can also assemble a stack from their PMS, a contact center, survey software and staff messaging.
Akia's difference is the attempt to make one hospitality-trained agent the thread through all of it. The agent can begin in supervised mode, drafting messages for approval, before a property allows it to answer specific question types directly. Staff corrections can become suggestions for the reference library. Complaints, refunds and sensitive matters can remain human. That gradual trust model is more credible than a single automation switch.
- Buys itHotel owners, general managers, operations leaders, marketers and vacation-rental managers.
- Uses itFront-desk teams, central service centers and guests on text, WhatsApp, web chat or phone.
- Pays for itThe property or management group through a quote-based business software subscription.
- ReplacesPieces of messaging, check-in, upsell, survey, marketing and staff workflow software.
The company has raised a supplied total of $12.1 million. Its disclosed rounds include a $6 million Series A led by Altos Ventures in January 2023, with GSR Ventures participating, and a further $5 million round in September 2024. That is modest beside some hotel-technology peers, which makes focus important. Akia may touch marketing, voice and operations, but hospitality remains the organizing constraint: the reservation is the context, the stay is the timeline and service is the standard.
The agent moves behind the desk
The newest products make the direction clearer. Ask Akia lets employees question the same system that speaks with guests. A new hire can ask about a policy, send a guest message, create a task or issue a door code without digging through a binder. AI Voice brings calls into the operating loop, with the company saying it can answer questions, make a reservation, collect a payment and push the result into the PMS.
This is a larger ambition than better texting. Akia wants to become a system of action for a property: the layer that receives intent in natural language and translates it into hotel work. The danger is familiar. An inaccurate answer about breakfast is irritating; an incorrect door code or unauthorized refund is operationally serious. The company's separation of knowledge, skills, supervision and escalation is therefore not product taxonomy. It is the safety architecture on which the proposition depends.
For operators, the practical test will be measured one shift at a time. Does the inbox get quieter without guests feeling ignored? Do corrections become durable knowledge? Can a manager add a new policy across several properties without retraining every employee? Those questions are less theatrical than a talking robot, but they decide whether automation becomes infrastructure or one more tab.
A successful stay is full of tiny moments that should feel obvious to the guest and coordinated to the operator. The towel arrives. The key works. The room is ready. The message sounds like the property, not a generic bot. Akia's wager is that AI earns its place in hospitality by making those moments less visible as work. The front desk does not disappear. It becomes a conversation, backed by software that knows when to answer, when to act and when to find a person.