Generative AI Projects — An Ongoing Exploration in Images & Video

Individual Project

Project Type
My Role

Creative Direction, Prompt Engineering, Image Compositing

Location

Logan, UT

Year

2026

Software and techniques used

  • OpenArt

  • Figma Weave

  • Claude

  • Adobe Photoshop

  • Adobe Premiere

Project Description

A self-initiated generative AI project exploring scalable lifestyle content for e-commerce — spanning images and video. Using AI generation tools, I created virtual models and composited them with real product photography to produce editorial-quality lifestyle content across NBA team merchandise, apparel, and accessories, without traditional photo shoots. This is an ongoing collection with new directions added over time.

Generative AI Design

Target  Audience

E-commerce teams, creative directors, and brand marketers in sports and lifestyle retail looking to scale visual content production while maintaining editorial quality and brand consistency.

Design Process

Design Problem

The core challenge was making AI-generated models and scenes feel cohesive with real product photography. Mismatched lighting, unnatural fabric drape, and inconsistent brand aesthetics required iterative prompt refinement and manual compositing to produce results that read as intentional and on-brand — across both static images and video formats.

I began by studying existing lifestyle editorial references to define a visual language — color palette, model styling, and scene composition. AI-generated models were prompted iteratively, then refined in Photoshop using masking and inpainting to integrate real product images. Video explorations followed a similar process, extending static compositions into motion. Each output was evaluated against brand tone and visual consistency before being added to the collection.

The Town

A focused product showcase built around the Golden State Warriors City Edition jersey. The goal was to highlight garment details — stitching, colorway, and construction — through AI-generated model photography and video, creating content that functions both as a product feature and a visual story.

Starting with the jersey as the hero product, AI models were generated to match the team's aesthetic and skin tone diversity. Lifestyle stills captured key product details from multiple angles, while video extended the shoot into motion — bringing the garment to life beyond static imagery.

Generated Stills

Generated Video

Cour Adjacent

A creative lifestyle exploration using NBA tee designs as a starting point. Rather than product-focused shots, the goal was to build an editorial mood — placing AI-generated models within color-blocked scenes that echo each team's identity, resulting in imagery that feels more campaign than catalog

Beginning with Celtics and Knicks tees, AI models were generated and styled to fit a cohesive visual direction. Scenes were built around each team's colorway — deep green and bold orange — to create a consistent yet team-specific lifestyle aesthetic across the collection.

Generated Stills