About the role
The Opportunity Adobe is seeking Applied Scientists at both early-career and senior levels to develop AI innovations from research concepts through product validation and technology transfer. You will identify promising research directions; design, train, post-train, and rigorously evaluate multimodal architectures, with particular emphasis on image and video editing; and contribute to technical strategy. Depending on experience, you will contribute to or lead ambiguous, high-impact projects while remaining deeply hands-on. You will strengthen the broader research and engineering team through technical judgment, experimentation, collaboration, and, at senior levels, mentorship. This role requires hands-on model development and applied research. Experience limited to prompt engineering or model-API integration is insufficient. What You’ll Do Identify research advances with the potential to create differentiated Adobe product capabilities. Own or contribute to applied research from problem formulation and experimentation through prototype implementation and validation. Curate and analyze multimodal training data, including sampling, filtering, deduplication, synthetic-data generation, and quality measurement. Design and optimize multimodal architectures, including vision encoders, modality projectors, and cross-modal attention. Build rigorous evaluation pipelines covering model quality, robustness, safety, latency, scalability, and product usefulness. Diagnose model failures through ablation studies, error analysis, and data- and architecture-level experimentation. Minimum
Qualifications Master’s or Ph.D. in computer science, machine learning, artificial intelligence, or a related field, or equivalent practical experience. Hands-on experience developing transformer, diffusion, or multimodal architectures through direct training or adaptation of model parameters. Strong Python and PyTorch skills, including experience developing, debugging, and improving research-quality model code. Experience with end-to-end model development, including data preparation, training, evaluation, and failure analysis. Experience with post-training techniques such as supervised fine-tuning, preference optimization, reinforcement learning, alignment. Ability to communicate technical decisions, experimental findings, and measurable impact effectively. Candidates will be evaluated on demonstrated technical depth and impact rather than years of experience. Preferred
