I built a rotation checker for product photos, then a whole campaign tool on top of it
Two products came out of one problem. Sellers kept uploading product photos that looked fine on their phone and came out sideways on the marketplace, because the camera's rotation tag and the pixels disagreed. So I built UprightAI. It reads the pixels and tells you the exact rotation needed, rather than trusting the tag. Once the rotation worked, the next question was what sellers do with the…
Two products emerged from a single issue plaguing sellers. Photos uploaded by sellers looked perfect on their phones but ended up sideways on the marketplace due to a disagreement between the camera's rotation tag and the pixel data. To address this, UprightAI was developed. This tool scrutinizes the pixels and provides the exact rotation required, rather than relying on the tag's accuracy.
With the rotation problem solved, the next question arose: what should sellers do with the photo after that? Most then required studio shots, lifestyle scenes, a short video, and platform-specific copy for each product. This led to the creation of BlueVeta. It operates on a pay-per-pack model, with no subscription, starting at $9.99. Each purchase comes with a 30-minute refund guarantee. The website is blueveta.com.
The construction of BlueVeta involved multiple Claude Code sessions, each focusing on a specific aspect of the tool. The clear division of ownership between these sessions has proven effective. However, one challenge remains: image models occasionally redraw label text incorrectly. Consequently, each generated image must undergo a verification process against the actual package image. For online sellers of physical products, the question posed is: what would make you trust a generated product image enough to list it?
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
