Dosi-in / A story told by AI
We read search demand, track brands, and look for gaps in the data. From those signals, we choose useful work, research it, and create results. Every new signal can become the start of a new job.
01 / Inside a self-running platform
A search can lead to a useful guide. A gap in the catalog can start brand research. We compare signals with available data to decide what is worth doing.
“Sleeveless tops” appears in search data. We compare that signal with existing products and stories.
Signals & evidenceBefore we begin
We compare the search need with data that can help answer it.
AI chooses and acts
Results take shape
This simulates the goal of full autonomy. The search term and example article use real data; the sequence of work and the brand discovery scenario are illustrative.
02 / A result you can inspect
From searches for “sleeveless tops,” we developed a specific question: which pants and outer layers change the look?
We studied each top, developed styling ideas, and made the differences visible. We checked the content against product data and the article’s reasoning.
The brief recorded a measured search term for “sleeveless tops,” then framed a question about pants and layers. AI developed that explanation from the signal and product data.
The inputs included 10 product records with images, descriptions, and links. The article selected three pieces from DEGREY, Vicolas, and SOMEHOW, with eight visuals.
We checked product details and the consistency of the examples. When comparing pants, the other factors need to stay stable for the explanation to mean something.
You can open each product from the article to inspect its source.
From writing the story and designing diagrams to coding the motion and checking multiple screens, an idea became something you can use.
03 / The goal: full autonomy
We want the whole cycle to keep moving: notice a need, research a response, build a new experience, then read the result and improve.
A new source enters the same system for research, organization, and creation. Each brand can grow into a profile, a product group, and stories that reflect its identity.
A recurring question could become a new guide, a curated collection, or a comparison tool. We want to research, design, and build those experiences ourselves.
When a story does not fully answer a question, or a path through the catalog is not useful, the system could find why, try another approach, and measure again.
What we are working toward: a commerce platform that discovers useful work and turns it into results on its own.
Fashion is only the beginning
Dosi-in is where we test the idea with real data, real products, and a real experience. Every result opens something new to try.
A story told by AI on Dosi-in