One source, a complete set of outputs.

Product photos. A full set from a single photo, even a phone snap in bad light: a packshot on a plain background, lifestyle frames, a detail and an in-hand shot. The biggest risk in AI product photography is that the product stops being yours along the way: the text on the label turns into gibberish, the cap changes shape, a seam disappears. In my work, the label, the shape and the proportions are the one thing that never moves.

Ad creatives. One campaign in versions for Instagram, Facebook, Pinterest and website banners. I lay out every size separately, so the headline always stays readable and the product stays clearly visible, instead of cropping one image to fit every format. Headline and copy included.

Fashion. Studio, street and editorial frames and a grid of looks, with one model across the whole set. In fashion the product is the garment, so the cut, the fabric and the detail stay the same in every frame.

Good fit when you have the product but no studio, no photographer and no three weeks to spare, or you're launching a collection before every size is even sewn.

Orchestrated end to end in Claude Code, on my own production system.

I quote after the brief. Everything I deliver is AI-generated and I state that in writing.