3D Pose & Reference Workflows
Cinamon · AI Lead · 2026 — Built browser-based posing and reference-image workflows for controllable AI animation inputs.
Product problem
AI animation needs controlled visual anchors before generation begins. Creators need practical ways to pose rigged characters, translate image references into usable poses, and carry those references into a repeatable image and video workflow.
Role and ownership
As AI Lead at Cinamon, I developed product workflows for browser-based character posing and 3D reference creation, contributed image-to-pose functionality, and connected the resulting assets to CineV’s creation experience.
What I shipped
- Built browser-based workflows for posing rigged characters and preparing controllable scene references.
- Developed an image-to-pose path to reduce the manual setup required for character staging.
- Connected 3D pose authoring to reference-image generation and reusable project assets.
- Integrated the workflow with CineV’s broader storyboard and AI-enhancement path.
Engineering depth
Interactive 3D authoring, rig-aware character handling, image-to-pose processing, reference-asset workflows, browser and service integration, and production-oriented validation.
Public outcome and evidence
The work supports CineV’s public 3D-first direction model, where creators stage characters and shots before applying AI enhancement. This provides more controllable inputs than treating each generated shot as an isolated prompt.