AI Product Photography + Short-Form Video — Bridal Social Content (Paid Test, Ongoing Potential)

Overview We're a D2C bridal brand with a library of high-quality in-studio product photography. We want to turn those existing studio images into scroll-stopping social content — both stills and short subtle-motion videos — by placing our gowns into beautiful, believable scenes using AI, without altering the dress itself. We do not want the results to look like AI was used. This posting is a small paid test project. It's the first step toward a much larger ongoing engagement: a full AI-driven "photoshoot" where we drop our gowns into a range of styled scenes at scale. We're hiring the test to find the right long-term partner, so strong performance here leads directly to bigger, recurring work. The single most important requirement: The gown must stay identical. Beading, lace pattern, neckline, silhouette, fabric drape, and color have to be preserved exactly from our source photo. We are changing the scene, lighting, and atmosphere around the dress — never the dress. If your work tends to "reinterpret" the subject, this isn't the right fit. Please only apply if you can show before/after pairs where the product is unchanged. Aesthetic direction: Refined and editorial. Clean, premium, light-driven. For video, we want subtle living-photo motion: fabric breathing, light shifting, gentle model movement — not big camera moves or anything that looks "AI-animated." Loop-friendly and elegant. Paid test deliverable: We'll provide 2 in-studio product photos. For each gown, deliver: 3 finished social stills — the gown placed into a styled scene, preserved exactly, with believable lighting, shadows, and ground contact. Put the model in three different spots in the scene. Delivered in platform-ready ratios (9:16 and 4:5 or 1:1). 1 short subtle-motion video (approx. 5–10 seconds, 9:16) in the living-photo style described above. So the full test is 6 stills + 2 videos across 2 gowns. Please also include a short note (3–4 sentences) describing the tools and workflow you used — what handles subject preservation, scene generation, and the video step. What we'll provide: 2 high-resolution in-studio product photos Brand/aesthetic references Target output specs and aspect ratios What makes a strong applicant: A portfolio with input→output pairs showing a real product preserved while the background/scene changed (fashion, apparel, or ecommerce especially) Fluency with a serious, repeatable pipeline — e.g. ComfyUI workflows, FLUX / Seedream or comparable models, inpainting/masking, relighting, and a current image-to-video tool Experience producing cohesive sets (not just one-off images) Comfort with commercial/ecommerce delivery and platform-ready formats. Bonus: The ability to build a reusable workflow we could eventually run in-house. Please answer in your application Share 2–3 examples where you kept a product identical while changing the scene around it. (Input + output, ideally.) What's your tool stack for this kind of work? Walk us through your pipeline briefly. How do you keep a set of images visually consistent (lighting, mood, style) across a collection? What's your approach to the subtle-motion video step, and which tool(s) do you use? Budget & structure We will do a paid test for the deliverable above. Let me know your fee for this paid test. We're shortlisting a small number of applicants. The strongest performer moves into the larger ongoing project (full scene-based gown shoots at volume), which we'll scope and budget separately.