Next Yacht Group

Next Yacht Group

AI Production · 07 2026

A social launch campaign announcing the in house production of hulls and metal components for the group's future superyachts. Materials as the only subject, with the vessel teased only in the final seconds.

NamedSport Hero Film

NamedSport Hero Film

AI Production · 05 2026

A fully AI generated hero film for NamedSport, built around eight athletes and four product ranges. Designed as a modular system: one master film, four standalone cuts, four aspect ratios.

Alia TVC

Alia TVC

AI Production · 03 2026

A luxury inspired TVC promoting the recycling of electrical appliances by highlighting the precious metals hidden inside them.

Autostrade per l'Italia

Autostrade per l'Italia

AI Production · 01 2026

A DOOH campaign across Milan announcing the company as a partner of the Milano Cortina Winter Games. Two subjects, a figure skater and a downhill skier, in an illustrative style built to sit apart from photographic realism.

Air Action Vigorsol

Air Action Vigorsol

AI Production · 09 2025

A social campaign for a tarot themed online activation, built around an original character: a Vigorsol fortune teller yeti in an ice cave. From a single static concept to four films, each with 6, 10 and 15 second cutdowns.

✕
AI Production · 05 2026

NamedSport Hero Film

The brief asked for eight athletes across four disciplines, from occasional gym goers to committed endurance runners, each in a distinct setting. That spread of talent and locations was hard to justify on a conventional shoot, so we proposed moving the production to AI and the client backed it as an experiment. We ran it with the same logic as a live action job: virtual casting to lock faces and locations, then mastershots and environments to build a storyboard precise enough to work as the drive images for generation, keeping consistency across characters, light and product from one block to the next. When the packaging was rebranded mid production, we generated the affected shots with blank product references and rebuilt the packshots later through targeted generation instead of starting over. The voiceover stayed human, recorded with a professional speaker and used as the spine of the edit, while post carried the kinetic type, the motion graphics and a grain and distortion pass to pull the generated footage back towards a real camera.

  • Virtual casting and storyboard
  • Storyboard driven generation
  • Packshot regeneration after rebrand
  • Kinetic type and motion graphics
  • Modular delivery, 16 cuts