360's Nano Work Launches Two-City, Thousand-Store AI Onboarding Plan

Speaking at the event, Zhou said competition in the AI industry is shifting from model parameters to actual delivery. Large models determine whether a system can understand a task, while agents determine whether it can get the job done, he said, adding that whether a product is good ultimately depends on whether it can complete real work in real scenarios. He predicted that next year may be a "first year of AI delivery" that tests product strength.

A June survey by CCID Think Tank found that among nearly 190 million micro and small business operators, more than 90 percent have learned about AI tools, but fewer than 30 percent use them continuously and only 6 percent have standardized AI processes. The think tank identified not knowing how to use the tools, finding them hard to use, and finding them unaffordable as common obstacles.

The plan sets Beijing and Shanghai as starting points and will advance in stages, with a long-term goal of putting AI to work in more than a thousand operating physical stores. Nano Work will follow what it calls a "few deep, batch support" approach: breaking down the needs of a single store into reusable workflows, then copying them to more stores through unified templates. The 798 stop is meant to explore how store experience can be converted into AI workflows, while the Lujiazui stop will test whether that path fits the coffee industry. Running the two cities together is intended to verify whether the model can be replicated across formats and stores.

Zhou said Nano Work is not simply delivering tools to stores but aims to help operators complete AI implementation in practice. Before the plan started, a forward deployed engineer (FDE) team visited merchants in the 798 park to map their operating processes, judge which steps are suited to AI, and build customized AI upgrade plans for individual merchants. The company said the goal is to find a replicable path for AI adoption in physical commerce that can be extended to more stores.

A Nano Work representative said the plan will gradually build up task lists, scenario templates and service methods for physical operations, so that practice leaves behind replicable methods and results can be evaluated with evidence. The "thousand-store onboarding" target is not limited to restaurants or coffee shops, the representative said, but covers a broader range of micro and small physical businesses, with the aim of bringing AI into the daily work of physical commerce.