Shopify's Sidekick separates showing an answer from finishing every lookup behind it. Its streaming design lets the model begin responding with placeholders that fill as supporting content arrives. The interface receives the main answer and those later results through one stream, placing each result in its designated area.
A small buffer also holds incomplete formatting, such as a link whose address has not finished arriving, until it can render correctly. Together, these mechanisms let users see progress without watching unfinished formatting turn into links or waiting for every external lookup before any response appears.
Terms in this piece · Glossary
streaming — Sending a model's response token by token as it is generated, so the reader sees text immediately instead of waiting for the whole answer.
token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.
