Senior Staff / Staff Research Scientist

Dyania Health•Jersey City, NJ

4h•Remote

About The Position

Dyania Health is a venture-backed company founded in 2019 which has developed Synapsis AI, an end-to-end system that combines a medically post-trained LLM with a physician-driven algorithmic reasoning engine to understand and assess clinical characteristics in electronic medical records. Synapsis AI is designed for installation within the healthcare system's computing environment or on healthcare system private clouds to automate manual chart review of EMRs without removing any data from the healthcare system. Synapsis AI completes pre-screening on both unstructured and structured patient data, deploying medical logic with temporal sensitivity to dynamically match changing patient characteristics to complex clinical trial criteria within the exact window when a given EMR may qualify for study protocol criteria. We are seeking a Staff or Senior Staff Research Scientist to lead the design, development, and implementation of advanced AI/ML and LLM-based models for clinical data understanding. This is a senior, hands-on research role. You will build and customize models, design architectures, and drive ML innovation from concept through production. This is not a data science or analytics role—we are looking for someone who deeply understands how models work internally and can independently solve complex, open-ended problems. Leveling is based on experience: Staff Research Scientist: 7+ years of relevant experience Senior Staff Research Scientist: 10+ years, including technical leadership responsibilities What We’re Looking For A senior researcher who builds models, not dashboards A hands-on technologist who understands model architecture, training, and optimization Someone who can define strategy, lead initiatives, and still write production-quality code A strong independent problem solver comfortable with ambiguity and research-driven development A technical leader who can mentor others while delivering individually

Requirements

7+ years (Staff) or 10+ years (Senior Staff) of experience in AI/ML research or ML engineering
Graduate degree in Computer Science, Engineering, or a related technical field
Strong hands-on experience building, training, and deploying ML and AI models, including LLMs
Deep understanding of model fundamentals and architecture design
Proven ability to solve problems end-to-end, from research through implementation
Advanced proficiency in Python for model development and training
Experience customizing models rather than relying solely on pre-built solutions
Ability to independently research, evaluate, and propose technical approaches

Nice To Haves

Publication record in top-tier AI, NLP, or ML conferences
Experience working with EMR or other complex clinical datasets
Experience with multi-modal AI systems
Prior senior or lead-level technical experience
Familiarity with regulatory considerations for clinical AI systems
Experience contributing to long-term technical strategy or system architecture
PhD is welcome but not required

Responsibilities

Design, build, train, and deploy AI/ML and LLM-based models
Own the full model lifecycle: problem definition, research, prototyping, training, evaluation, and deployment
Apply strong understanding of ML and AI fundamentals to real-world biomedical problems
Develop and optimize models using Python and modern ML frameworks
Research emerging techniques and propose innovative solutions aligned with product and business needs
Independently analyze complex, ambiguous problems and determine appropriate modeling approaches
Evaluate and recommend technical solutions and system architectures
Translate research concepts into scalable, production-ready systems
Contribute to long-term ML and AI strategy and technical vision
Lead cross-project research and development initiatives
Provide technical mentorship and guidance to junior researchers and engineers
Establish best practices for experimentation, model development, and deployment
Influence technical roadmaps and architectural decisions