Amazon [1] Alexa is already capable of identifying different voices to give personalized responses to different users in the household, thanks to the added support[2] for voice profiles[3] two years ago. Now, those same personalization capabilities will be offered to Alexa Skill developers, Amazon has announced.
Alongside Amazon’s big rollout of new consumer devices[4] on Wednesday, the company also introduced a new “skill personalization” feature for the Alexa Skills Kit, that lets developers tap into the voice profiles that customers create through the Alexa companion app or from their device.
This expanded capability lets developers make skills that are able to remember a user’s custom settings, address their preferences when using the skill, and just generally recognize the different household members who are speaking at the time, among other things.
To work, Alexa will send a directed identifier — a generated string of characters and numbers — to the skill in question, if the customer has a voice profile set up. Every time the customer returns to that skill, the same identifier is share. This identifier doesn’t include any personally identifiable information, Amazon says, and is different for each voice profile for each skill the customer users.
Skill developers can then leverage this information to generate personalized greetings or responses based on the customers’ likes, dislikes, and interests.
If the customer doesn’t want to use skill personalization even though they configured a voice profile, they can opt out of the feature in the Alexa app.
Personalization could be a particular advantage to Alexa skills like games, where users may want to save their progress, or to music or podcasts/audio programming skills, where taste preferences come into play.
However, Alexa’s process...