Liiper Sport began as an exploration of how an AI influencer could become the central figure of a complete fitness campaign rather than remain a single generated portrait. The project combines character consistency, futuristic sportswear, brand direction and video editing around Alexa, a virtual model developed for a premium “Gold Standard” visual concept.
The starting point was a simple question: how could a virtual character communicate the energy, confidence and discipline associated with a modern sports brand? I wanted Alexa to feel aspirational but still believable as the face of a campaign. The objective was not merely to create an attractive AI image. It was to build a visual system that could remain coherent across portraits, promotional frames and motion.
The “Gold Standard” concept became the creative anchor. Gold suggests achievement, ambition and premium quality, making it a natural fit for a performance-focused campaign. Darker technical environments and controlled highlights were then used to give the gold elements more impact without allowing the background to compete with the character.
Character consistency was one of the most important parts of the process. Alexa needed to remain recognisable while her pose, expression, framing and environment changed. For each image, I evaluated whether the facial identity, attitude and styling still belonged to the same character.
This required a process of visual selection rather than relying on the first generated result. Images were compared according to a small set of creative criteria:
The strongest images were those in which all these elements supported the same story. This selection stage was essential because consistency is created as much through curation as through generation.
The campaign was designed around a controlled visual vocabulary. Gold and black establish the premium fitness tone, while the lighting adds a cinematic quality. The environments suggest a near-future training world without becoming so complex that they distract from Alexa and the apparel.
I treated each image as a potential advertising frame. That meant leaving enough visual space for movement, transitions and possible campaign copy. Wide compositions were useful for establishing the world, while closer portraits brought attention back to the character and clothing details.
Material definition was also important. Reflective surfaces, fabric texture and edge lighting help the sportswear feel intentional rather than incidental. The goal was to make Alexa and the clothing appear to belong to the same designed environment.
Although the campaign began with still imagery, the final sequence was considered throughout the image-development process. A strong standalone image does not always work well in an edit. Video requires variation in framing, direction and visual intensity so that one shot can lead naturally into the next.
I therefore selected a combination of establishing shots, medium compositions and close details. The sequence needed enough contrast to maintain interest, but every frame still had to share the same character, palette and campaign tone. This balance between variety and continuity shaped the final selection.
The video-editing stage transformed the individual assets into a campaign. I organised the selected shots by visual scale and energy, using wider frames to introduce the world and closer images to create emphasis. The rhythm of the cuts was designed to make Alexa feel like the lead of a sports advertisement rather than a sequence of unrelated AI images.
Transitions were kept in service of the concept. Their purpose was to connect movement, shape and light between frames, not to draw attention away from the character. Color continuity also helped unify assets created from different compositions. By maintaining the gold-led palette and a consistent level of contrast, the edit gained a more deliberate and professional finish.
The final pass focused on pacing. Some frames needed more time to communicate the clothing and environment, while others worked better as brief accents. This process gave the sequence a clear progression: introduce the character, establish the brand world, build energy and conclude with a strong campaign identity.
The main lesson from Liiper Sport is that AI-generated content becomes much stronger when it is directed as a complete production. Generating images is only one stage. The concept, character rules, styling, selection, sequencing and edit are what turn those images into a coherent brand experience.
The project also reinforced the importance of planning for the final format. When the goal is video, image creation should anticipate movement and editing from the beginning. Compositions need to provide visual variety, and the character must remain stable enough for the audience to believe that every shot belongs to the same campaign.
Most importantly, the technology did not replace creative direction. It expanded the range of possibilities available during development, while the final quality still depended on deliberate choices about what to keep, what to remove and how to present the work.
Liiper Sport presents Alexa as more than a virtual model. She becomes the central identity of a futuristic fitness campaign built across still images and motion. The project demonstrates how AI character creation, art direction and post-production can work together to produce branded content with a consistent visual voice.
Explore more character development in the Alexa AI Influencer showcase, or view another product-focused experiment in the Conceptual Sport Shoes showcase.
You cannot copy content of this page
English 
Portuguese 
