Three-Step AI Commercial Workflow (Assets → Setup → Generations)

A production framework for building an AI-generated commercial in three phases: assets (build and organize all reference images/elements), setup (configure the project — folders, elements, continuity references), and generations (run and iteratively direct the actual shots). The claim is that the same three-phase pipeline transfers to any product category — sneakers, perfume, cars — by swapping only the assets, not the process. Functions as the umbrella framework for more specific techniques used within each phase: Hierarchical Project Folder Structure for AI Generation Projects and Higgsfield @-Element Asset System belong to assets/setup, while Multi-Model Specialization and Compositing, Actor-Directing Prompting Method (Takes and Incremental Notes), and Blocking Lock via Position Reference belong to generations.

Case Studies

Arcti (2025): a real client project for a newly launched Finnish canned still/sparkling water brand. The client supplied only two rendered product files and a mood board — everything else was built out from that starting point: reference gathering via Pinterest Reference Sourcing for AI Image Generation, prompt-writing via Custom-Trained GPT for Image-to-Prompt Conversion, and generation on Higgsfield ('Cadream'/Nanobanana Pro) at 4K resolution.