All articles

Data acquirement for language assistants: Where do Alexa and co. get their information from?

The hype about artificial assistants and automated services is undeniable. And terms such as AI or chatbots are at the center of a popular discussion. The biggest advantage of the new technologies is the automated communication. It provides guests with a 24/7 service while also reducing the workload of employees.

Advertisements show flawless communication between man and machine. Alexa turns the light out, OK Google turns up the radiator, plays music or buys groceries, which of course are delivered to your front door within just a couple of days. Intelligent systems that take on the workload and automatize routines – doesn’t that sound terrific? Technology has made major improvements over the last few years. Language operated communication has already been integrated into our everyday lives. But how does it work? Where does the data that Alexa and co. use come from?

Behind every chatbot and every digital assistant hides a mountain of programming, which of course only becomes obvious once one takes a closer look at it. Amazon Echo, Google Home or Apple’s Homepod are stand-alone solutions that depend on the programming of their ecosystem. Thus, they are only as smart as the programming that they are based on. The obvious solution is to reduce the workload and improve customer service. Unfortunately, simply acquiring the hardware will not work!

While Alexa and co. are capable of answering most questions by searching for content online or on already integrated services (e.g. weather apps), they struggle with more specific questions. Those capabilities need to be “trained”, i.e. programmed.

Establishing a digital assistant – but how?

First of all, you need to decide what your digital assistant should be used for. Is it meant to support the booking of accommodations as a website chatbot? Or as a digital assistant in your hotel lobby that answers around the clock questions about opening hours, events or hiking tracks? Digital assistants, whose language skills are complemented by a screen, are particularly practical. This way, advertisers have the chance to additionally display information visually. Moreover, a guest that requires information about a nearby hiking track also gets to look at the route and all basic information.

Again the following principle applies: the better the data that builds the basis of a conversation with a digital assistant, the more satisfying the answer. Data quality as well as its permanent maintenance and examination are extremely important.

Conclusion: Behind every digital assistant hides a complex system that depends on a consequent analysis of conversations and an ongoing editing of basic content. Both sides profit from this: the users that effortlessly get helpful answers and the operators that experience a new level of guest contentment and loyalty.

Are you considering to establish a digital assistant for your customers? It would be our pleasure to assist you with the realization!