Skip to content

Type a term, e.g. API, Shopware or SEO.

    ← All projects

    06 · AI integration · Fashion · Custom development

    AI product images instead of photo shoots

    LANAKILA, the Leipzig sports brand of Sylvia Michalk, runs an online store with an extensive product range. Every new collection required expensive photo shoots — time-consuming, costly, barely scalable.

    Project for LANAKILA, the Leipzig sports brand of vona owner Sylvia Michalk · project from 2025

    • Custom development
    • Nano Banana Pro
    • PHP middleware
    • MariaDB
    • Online store integration
    • Social media
    0 shoots
    needed for new product images
    according to LANAKILA
    Weeks → hours
    from collection briefing to finished product image
    according to LANAKILA
    1 pipeline
    for online store and social media campaigns

    01 · Problem

    LANAKILA runs an online store with a very large number of products and regularly releases new collections. Every collection required professional photo shoots — with photographers, models, locations and extensive post-production. That costs not only a substantial budget but also weeks of lead time. New products could only go online after a significant delay.

    02 · Solution

    vona built a custom AI pipeline — without N8N, fully tailor-made. At the core of the solution is a PHP middleware that generates high-quality product images with Nano Banana Pro, based on reference images of the products and dynamically managed system and user prompts (stored and maintained in MariaDB). The prompts can be varied by collection, season or campaign — without touching the code. The generated images go through an approval workflow right inside the middleware: once approved, they are automatically assigned to the corresponding products in the online store via an interface and are immediately available for social media campaigns.

    03 · Result

    LANAKILA no longer needs elaborate, expensive photo shoots for every new collection. New product images are ready in hours instead of weeks. Through the middleware's approval workflow, images can be assigned to products directly in the online store — with no manual steps in between. The same images can be used for social media campaigns at the same time. The result: the cost of shoots disappears, new products go online much faster, and LANAKILA keeps full control over image style and prompt strategy without developer effort.

    04 · Workflow

    1. Reference images: of the products
    2. Prompts: MariaDB: system and user prompts, adjustable without touching the code
    3. PHP middleware: no n8n, tailor-made
    4. Nano Banana Pro: generates product images
    5. Approval: workflow in the middleware
    6. Online store: assigned to products automatically
    7. Social media: ready for campaigns right away
    Reference images of the products and dynamically managed prompts go into a custom-built PHP middleware that generates the images with Nano Banana Pro. The prompts can be varied by collection, season or campaign without touching the code. Once approved in the middleware, the images land in the online store automatically and are immediately ready for social media campaigns.

    Next project

    Your project here?

    Tell us about your problem. The first conversation is of course free of charge and without obligation.

    Request a project →

    Newsletter

    News from the vona workshop.

    AI tools we actually use, our weekly AI review & insights from our projects – short, practical and without spam.