SwitchBot Adds LLM-Powered Assistant for Smart Home Automation

SwitchBot, a provider of AI-enabled embodied home robotics systems, has launched kata, an LLM-powered assistant that lets users manage connected devices and create smart home automations using natural-language voice or text commands. Instead of switching between device pages, searching through manuals, or learning automation logic, users can open the SwitchBot App and tell kata what they want to do.

Built into the SwitchBot app, kata can access device names and status information to carry out supported actions. The assistant can also interpret requests such as a room being too hot and ask follow-up questions when information is missing. Supporting text and voice input, kata maintains context within an active session and can ask users to confirm when multiple similar devices or missing parameters are detected.

SwitchBot is extending the AI interface beyond device control. Kata can identify products using optical character recognition, provide installation guidance and consult manuals, FAQs and videos when troubleshooting. It can also retain context as users move toward additional support.

The release illustrates how generative AI interfaces can abstract increasingly complicated IoT configuration. Instead of requiring users to understand individual device settings and automation logic, an LLM can translate intent into actions across connected systems. That approach also makes contextual accuracy and confirmation important, particularly when assistants are authorized to change physical device states rather than simply provide information. Kata is available in SwitchBot app version 9.29 or later.

Posted by Pure AI Editors on 08/10/2026