Supermoon puts voice AI into physical venues: theme parks, stadiums, hospitals, museums, city tours and national parks. It is designed for noisy environments and for wearables and ambient devices rather than a phone screen, with Meta Glasses SDK access, personalisation and memory, and customer data kept on-premises.
Noise is the constraint that makes this a different engineering problem. Voice assistants are built and benchmarked in quiet rooms, and a stadium concourse or a queue line is where they fail; targeting those environments explicitly is more useful than another assistant that works well at a desk.
Enterprise integrations, APIs and moderation tooling sit around it. No pricing is published. Wearable AI in public venues raises consent questions about the people nearby who did not opt into anything, hardware dependence ties the deployment to device availability, and voice guidance in a venue is a safety-adjacent function when it is giving directions in a crowd.
