A proposal from inSylos for VyVa™ — where every clip locks your product and your model's likeness, across every generation.
Most UGC video tools generate clips that look different every time — the product packaging shifts, the model's face changes, the lighting drifts. For a brand like VyVa, where precision and trust are the product, that's not acceptable.
A pipeline designed so no single generation call decides what your brand looks like.
An LLM analyzes the campaign brief and produces a structured shot plan — a JSON breakdown of each scene: duration, setting, mood, framing. The "director" layer, ensuring clips tell a coherent story, not a disconnected set of AI generations.
The core of the system. A detailed style prompt describing exact visual treatment — lighting, color grade, camera style — is prepended to every clip's generation call. A reference image is passed into the video model's image-to-video mode, so the product and model appearance are anchored, not reimagined, each time.
Individual 6–8 second clips generate in parallel via Google Veo (primary) and Kling (fallback, noted for strong reference-consistency). A background job queue handles many clips simultaneously, making multi-clip outputs practical within a reasonable wait time.
ffmpeg stitches clips together, applies a consistent color grade so every clip reads as one production, and mixes in audio — delivering a final cut that feels authored, not assembled.
A small reference set — multiple product angles — feeds the consistency mechanism. Combined with LoRA-style visual referencing, this keeps VyVa's packaging, logo placement, and product geometry accurate across every generation. Not approximately correct — accurate.
The same reference-image anchoring used for the product holds the model's face and features consistent across dozens of generated clips. Same face, same hair, same skin tone — clip after clip.
Scene planning and generation layers are largely reusable infrastructure. The genuinely difficult, VyVa-specific work is the consistency layer — building and testing the right reference-image pipeline so product and model don't drift across dozens of generations.
This includes experimentation across models (Veo, Kling, Runway) for best likeness and product fidelity, plus an automatic review/rejection step that regenerates any clip drifting too far from the reference.
Custom packages designed around VyVa's campaign volume, model requirements, and product portfolio.
A reference-anchored UGC video campaign for one VyVa product, including scene planning, generation, and assembly.