Artizan — AI Art Marketplace

The problem

What a piece of art is worth is not a published number. It is inferred from what comparable work sold for, and those results are scattered across auction houses that each publish on their own site, in their own format, on their own schedule.

So the person who can price a work confidently is the person who already knows where to look. Everyone else is guessing, or asking someone who is not disinterested.

What the app does

It collects auction results from more than 17 houses into one searchable place, puts an AI-generated valuation next to a work, and carries an editorial feed pulled from a third-party art publication so the market context sits beside the numbers.

On the other side of it, sellers list work and run auctions from inside the app.

What I owned

I worked on Artizan as a React Native developer at Inseed, on the client rather than the aggregation behind it.

My part was the entry side of the product: the add-auction flow and most of the app's other forms. This was not an architecturally complicated app, and it would be dishonest to dress it up as one — the aggregation and valuation were the hard engineering, and they were not mine.

The work that was mine had its own difficulty, and it is one that gets underrated. Listing a work for auction is a long form. A phone is a poor place to fill in a long form: there is no space to see what you have already answered, and abandoning halfway is easy and costs the marketplace a listing. Getting that flow to be finishable on a small screen was most of the job.