ML Engineer

Role overview Develop and evaluate generative models for architectural visualization, helping turn promising techniques into reliable product capabilities. You’ll assess image and video models, adapt them for specialized design tasks, and work with product and infrastructure colleagues to bring improvements into production.

Responsibilities - Evaluate image and video models on architectural tasks and recommend which are suitable for product use. - Train or fine-tune adapters such as LoRA for different materials, styles and drawing types. - Design generation pipelines and prompting strategies in collaboration with the product team. - Build evaluation datasets and metrics that show whether model quality has improved. - Work with infrastructure engineers to deploy models into production services.

Requirements - Practical experience training or fine-tuning diffusion models. - Strong PyTorch skills and familiarity with Diffusers and the Hugging Face ecosystem. - Ability to evaluate model quality using practical measures beyond training loss. - Experience moving a model from an experimental notebook into a service. - Clear written communication.

Nice to have - Familiarity with ControlNet-style conditioning, image-to-3D or video models. - Background in architecture or design.

Benefits and work setup - Full-time, remote role on an AI research team, with a direct line to technical leadership.