Visual production with generative AI

Photoreal image series that don’t look like stock. I direct the casting, the skin, the light and the setting until the audience sees itself in the picture — the same calls you’d make on a real shoot, just executed with whichever model the campaign needs. I’ve been working with generative AI since 2024 and watched it go from plastic to photoreal up close, so I know where each model breaks and how to fix it. Keeping a face, the light and the color consistent across a whole series happens piece by piece, through art direction.

Deliverables: Social media series, campaign key visuals, product catalogs, portraits and products in context.

Video for paid media and social, from the first idea to the final cut. Character and color continuity between shots gets solved in production, not patched in the edit. Every campaign ships in the formats its platform actually uses — vertical, square and horizontal — all from one art direction, not cropped versions thrown together at the end.

Deliverables: Spots for Meta Ads and Google Ads, Reels and vertical content, product video, brand pieces.

Landing pages and campaign sites built with AI agents, following a pipeline I developed myself: spec, greybox, automated tests, art direction and generative assets, in that order. What you get is versioned, tested code, locked-down credentials, server-side validation and an architecture that can grow. And because page speed moves your Quality Score, every technical decision is made with the cost of the click in mind.

Deliverables: Conversion landing pages, campaign sites, Quality Score optimization for Meta Ads and Google Ads.