Staff Applied Scientist, 3D - Firefly Foundry

Start free →

Or apply yourself on the official page →

About the role

The Opportunity Firefly Foundry is Adobe’s enterprise managed-service offering for custom multi-modal generative AI — deep-tuned image, video, and 3D models built on each customer’s IP, paired with creative production workflows and a media-intelligence layer, and deployed across new and existing Adobe surfaces. The business has gained significant traction in Media & Entertainment, marketing, and consumer retail, and is expanding rapidly into adjacent verticals. We are especially looking for candidates experienced in large-scale, industry-level pre-training and mid-training of multi-modality generative models with the focus on dynamic, physically accurate 3D worlds. This role has a direct impact on the quality of Adobe’s generation models, supporting next-generation creative workflows for millions of users. As a Staff Applied Scientist at Adobe, you will join a world-class team of applied researchers and engineers building the future of digital experiences — innovating across the full training stack, collaborating across data, modeling, and product, and shipping your work to customers worldwide. What You’ll Do • Build novel generative models and pipelines to support mesh generation, physically based material generation and texture generation, scene reconstruction, lighting estimation, animation, and simulation for mediate & entertainment use cases as well as high-fidelity product imagery,  • Build agentic workflows that use AI systems to create, manipulate, repair, annotate, evaluate, and refine 3D and 4D content. • Develop large-scale synthetic data pipelines that involve game and rendering engines, digital environments, 3D geometry, physically-based materials, lighting, animation, and simulation. • Develop, implement, and rigorously evaluate advanced techniques for conditional generation and editing — with emphasis on improving instruction compliance, controllability, and visual quality in pre-trained models. • Collaborate cross-functionally with researchers, engineers, and product teams to translate multimodal innovations into scalable downstream applications within Adobe products. • Develop and maintain robust evaluation pipelines to assess generative models across quality, efficiency, robustness, and safety metrics. • Translate research concepts and published work into production-ready implementations using Python and modern ML frameworks.   Required Qualifications • Ph.D. or MS in Computer Science, Machine Learning, Computer Graphics or a related field, with 8+ years industry experience. • Extensive practical experience in large-scale generative AI training with a focus on image and video generation and editing. • Experience with diffusion models, transformers, or other state-of-the-art generative architectures. • Strong working knowledge of 3D content systems, including meshes, materials, textures, scene structure, cameras, lighting, coordinate systems, skeletal animation, and motion workflows. • Strong coding and prototyping skills in Python, PyTorch, and ML infrastructure tools. • Excellent communication skills and ability to collaborate across cross-functional teams. • Experience working with product teams on technology transfers. • Strong publication record in Computer Science, AI/ML, or related fields.   What You’ll Bring • Experience on large-scale generative model training • Experience on synthetic data generation • Experience of working with large-scale datasets • Experience of working with product teams on technology transfers #FireflyGenAI About Adobe Adobe empowers everyone to create through innovative platforms and tools that unleash creativity, productivity and personalized customer experiences. Adobe’s industry-leading offerings including Adobe Acrobat Studio, Adobe Express, Adobe Fir

View the official posting →