Assignment description
Our customer is seeking a ComfyUI Workflow Architect to design, optimize, and scale advanced generative media pipelines.
You will architect diffusion workflows for animation production, integrate custom conditioning systems, and ensure consistency and coherence across multi-shot generations.
The role involves deep collaboration with engineering and production teams to develop bespoke model components and next-generation video synthesis systems that power large-scale AI-driven content creation.

Some of your day-to-day work tasks:
– Design and optimize production-scale diffusion model pipelines: Image-to-video synthesis architectures
– Style transfer, LoRA adaptation, and identity-preserving generation
– Multi-stage conditioning systems (depth, pose, optical flow)
– High-resolution upsampling and noise reduction pipelines
– Develop custom model components: fine-tune ControlNets, train domain-specific LoRAs, implement custom attention mechanisms and conditioning layers
– Collaborate with production engineers to ensure visual consistency through latent space control, embedding management, and multi-shot coherence strategies
– Integrate ML pipelines with training infrastructure, model versioning systems, and asset management tools

About the team
You will join a small, multidisciplinary tech team that blends engineering, creativity, and practical production insight.
The group includes one of the company’s co-founders, a full-stack developer based in Ukraine, and a specialist from the VFX industry.
This mix reflects a deliberate strategy to combine strong software development skills with a deep understanding of animation and visual storytelling.
The team operates at the intersection of ML research, pipeline engineering, and creative media production.

Must haves
– Strong expertise with ComfyUI’s model architecture and node-based inference graphs (required)
– Strong foundation in computer vision and diffusion models: DDPM, DDIM, latent diffusion, conditioning mechanism
– Proficiency with ControlNet, LoRA, and custom conditioning architectures
– Expert in PyTorch/CUDA optimization: memory management, kernel fusion, distributed inference
– Strong communicator able to bridge ML engineering and production requirements

Other requirements
– Self directed and proactive: The culture values individuals who take initiative, work independently, and drive their own tasks forward.
– AI tooling fluency: A forward-leaning attitude toward using AI tools in development is expected, aligned with the company’s AI-first production workflow.
– Strong communication and pragmatic problem-solving: Clear communication, structured thinking, and practical decision-making are essential, especially in cross-functional collaborations.
– Commitment to diversity: The company values a diverse team and aims to broaden representation beyond typical AI-startup demographics.
– Experience with custom ComfyUI nodes or PyTorch operators is a plus
– Familiarity with distributed training, dataset curation, or automated evaluation pipelines
– Knowledge of VFX workflows, real-time rendering, or asset management systems is beneficial

About the customer
Our customer is building the future of animation and digital IP creation.
By combining world-class storytelling with cutting-edge AI technologies, they produce high-quality animated content at scale for YouTube, TikTok, streaming platforms, and global licensing partners.
Their production ecosystem merges creative vision with advanced machine learning, enabling rapid content generation and large-scale media innovation.

Industries
Entertainment, Media and publishing