kadence · 7 hours ago
Research Scientist - Diffusion
San Francisco Bay Area
Full-time
Hybrid
Senior LevelKadence is an early-stage, well-funded deep-tech company building next-generation quantum computing hardware coupled with advanced AI algorithms. They are seeking an AI Research Scientist to lead foundational research in diffusion models and quantum algorithms, focusing on enhancing training and inference methods and collaborating with hardware researchers.
Hiring Manager
Aaron Chan
Responsibilities
Design and implement methods to accelerate training and inference for diffusion and probabilistic generative models
Explore mappings from classical acceleration techniques (e.g. ODE / solver-based methods such as DPM-Solver) to quantum algorithms
Build and benchmark “drop-in” quantum-inspired or quantum-native replacements for components of diffusion pipelines
Investigate how quantum noise can be treated as a feature rather than solely a limitation
Study noise distributions in diffusion processes and how quantum noise differs from classical noise
Identify application domains where quantum sampling or noise properties may provide unique advantages
Develop inference optimization strategies that generalize across GPUs, TPUs, and emerging hardware
Build and maintain robust research codebases for diffusion models
Evaluate approaches with clear quantitative metrics (speed, quality, cost, scaling behavior)
Work closely with quantum hardware and algorithms researchers across sites
Translate between AI/diffusion requirements and quantum hardware constraints
Help define abstractions and interfaces that expose new hardware capabilities to AI practitioners
Qualification
Required
Deep experience in diffusion / probabilistic generative models or quantum algorithms, with strong interest in the other
Proven research ability, demonstrated through publications, preprints, technical reports, or equivalent output
Strong coding skills (e.g. Python; PyTorch, JAX, or TensorFlow) and comfort with modern ML stacks
Ability to independently: Comfort moving between theory and practice (math, algorithms, and production-quality research code)
Experience working with modern accelerators (GPUs required; TPUs or other stacks a plus)
Preferred
Hands-on work with diffusion models (image, video, text, or scientific domains)
Experience with ODE/SDE-based samplers or inference acceleration techniques
Background in quantum circuits, quantum error models, or NISQ vs. fault-tolerant regimes
Experience on inference, optimization, or efficiency teams for large models
Prior collaboration with quantum ML or quantum information research groups
Company
kadence
Kadence - We place elite AI talent at the intersection of science & engineering, from PHD to Executive.
2-10 employees
Funding
Current Stage
Early StageCompany data provided by crunchbase