modality

multi modality that followers users

Mosaic was built to support multi-modal delivery, ensuring advice matches the way users naturally engage with AI. Today, we surface text, image, and voice recommendations, and are preparing to extend this to video. By aligning to user behavior across modalities, we ensure advice is experienced as part of the conversation rather than an interruption.

Imagine a user moving fluidly across surfaces: typing into one model, speaking to another, and browsing a third. Mosaic can respond seamlessly in each context—a spoken suggestion for dinner, a text link to a trusted financial brand, or an image preview of a product. Soon, even video will be delivered as a natural extension of the dialogue.

This approach keeps the user interface clean and non-intrusive, with transparent labeling that reinforces trust. The result is advertising that feels less like a placement and more like context-aware advice, creating engagement that is authentic, helpful, and predictive.