AI & Generative Workflows

I am currently exploring the possibilities of generative AI from a production and pipeline perspective, with a particular focus on image generation and its integration into established VFX and CG workflows.

Rather than treating AI image generation as a standalone creative tool, I am interested in making it controllable, repeatable and production-ready.

My current R&D work is centered around ComfyUI and the development of my own AI production pipeline. The goal is to build a structured environment in which AI-generated imagery can be integrated with existing 3D workflows, while maintaining as much control as possible over composition, perspective, lighting and other production-critical parameters.

My background in VFX, technical direction and pipeline development strongly influences this approach. Instead of simply experimenting with prompts, I focus on understanding how the underlying systems work, how different types of data can be used as guidance, and how these technologies can be connected to existing production processes.

This includes exploring workflows that combine 3D scene data, depth, cameras, image references, conditioning and generative models to achieve more predictable results.

The long-term goal is to develop a flexible AI production environment that can support real-world applications such as visual development, product visualization, automotive imagery, look development and VFX production.