Discovering the potential of multi-assistant architecture in a smart home system

In the field of smart home automation, there is currently a wide array of diverse solutions, with relatively small number of users. The idea has emerged to consolidate smaller brands within a unified environment, by branding only the applications, while using the same IoT platform in the background. However, this causes issues with the branding of smart home voice skills, since the APIs created by voice assistant providers are tailored for the use with the single brand per platform. Therefore, we propose the method to use multiple different voice assistant skills on the same cloud. This initiative has brought forth several advantages, such as the adaptation of skills for different manufacturers or types of devices, and the introduction of a freemium model where scenes can be independently acquired within a new skill. Beyond business decisions, this solution has played a role in enhancing load balancing and performance. Specifically, API calls per skill now involve fewer devices, resulting in a noticeable reduction in processing requirements.