After providing our clients with stay related mobile solutions, it was inevitable that they would approach us to provide an NLP endpoint. We didn’t see this as a big departure from what we already did with workflow management. The problem to solve was to use the NLP engine offered by Amazon to understand what the user wanted then send that into our workflow management system.
We created a skill that would understand a set of intents that were common in the hospitality industry and create a specific request for staff. For example - extra towels or breakfast hours. If the NLP engine didn’t understand the intent then we would use speech to text to send the request directly to staff.
Creating the skill was half the battle. Deploying 150 Alexa devices in a hotel is another story. The simple task of identifying the room that a request was initiated and whether the device was up and running where the two fundamental goals. The devices are great in-home devices but managing a fleet of them in one building was new, even for Amazon. We were able to successfully use an Amazon service to remotely manage and configure all devices.
The project proved to be a success for us and our client as we were able to delivery a voice solution for the brand and hotel. Along the way we also discovered a consumer’s expectation of a voice device was different in a home vs in a hotel. The use case is different and personalizing a voice device can be very difficult with current technologies. It was a fun project and I’ll fondly remember being able to wake 150 Alexa devices simultaneously in our conference room :)
Click here for an article on the project.