Deep Learning Data Scientist II
Spectral MD Inc
Summary
We are seeking a motivated Data Scientist to research, develop, and enhance AI/ML algorithms for computer vision and healthcare applications. This role involves developing and improving deep learning models, maintaining high standards of data quality, optimizing models for efficient deployment, and translating advances in AI research into practical healthcare solutions. The Data Scientist will collaborate with clinical partners and internal stakeholders to understand clinical and business needs and develop appropriate technical solutions. The ideal candidate will be capable of independently managing research and development projects while contributing effectively within cross-functional teams.
Job description
Essential Duties and Responsibilities
- Research, develop, evaluate, and enhance machine learning and deep learning algorithms for computer vision applications, translating advances in AI research into practical healthcare solutions.
- Implement and optimize model deployment pipelines to enable efficient, scalable, and hardware-optimized inference.
- Apply model optimization techniques to improve computational efficiency and inference performance while maintaining model accuracy.
- Maintain data quality and integrity throughout the data lifecycle to support robust algorithm development, training, and evaluation.
- Collaborate with clinical partners and internal stakeholders to understand clinical and business needs and translate them into appropriate AI/ML solutions.
- Independently manage assigned research and development projects from problem formulation through implementation, evaluation, and delivery.
- Ensure reproducibility of analyses and experiments through clear documentation, version-controlled workflows, and effective knowledge sharing.
- Stay current with advances in AI/ML, computer vision, and related technologies and evaluate their applicability to ongoing research and development.
- Communicate technical methods, results, and recommendations effectively to both technical and non-technical stakeholders.
Qualifications
- M.S. or Ph.D. in Computer Science, Electrical Engineering, Biomedical Engineering, Statistics, Applied Mathematics, or a related quantitative field.
- Minimum of two years of relevant research or professional experience in computer vision.
- For candidates with a master’s degree, the required experience must be professional experience obtained after completion of the degree.
Knowledge, Skills, and Abilities
- Strong programming skills in Python and experience developing maintainable, reproducible code for machine learning applications.
- Strong foundation in state-of-the-art deep learning architectures (e.g., vision transformers) for computer vision tasks including image classification, object detection, and semantic segmentation.
- Proficiency with deep learning frameworks and libraries, including PyTorch and TensorFlow.
- Strong understanding of AI/ML evaluation metrics, benchmarking methodologies, and best practices for rigorous computer vision model validation.
- Solid understanding of model optimization techniques, including pruning, quantization, mixed-precision inference, and other strategies for improving computational efficiency and inference performance.
- Working knowledge of model deployment and inference tools, such as ONNX, LibTorch, and TensorRT.
- Self-driven and capable of independently navigating ambiguous research problems, working with incomplete data, and delivering high-quality results.
- Strong written and verbal communication skills, including the ability to prepare technical documentation, reports, and scientific manuscripts.
Preferred Qualifications
- Experience working with multispectral and/or hyperspectral imaging data.
- Experience with generative AI models for medical image synthesis and simulation.
- Experience applying AI/ML methodologies to healthcare or biomedical applications, with demonstrated contributions to successful research, clinical, or product outcomes.
Requirements
- M.S. or Ph.D. in Computer Science, Electrical Engineering, Biomedical Engineering, Statistics, Applied Mathematics, or a related quantitative field.
- Minimum of two years of relevant research or professional experience in computer vision.
- For candidates with a master’s degree, the required experience must be professional experience obtained after completion of the degree.
- Strong programming skills in Python and experience developing maintainable, reproducible code for machine learning applications.
- Strong foundation in state-of-the-art deep learning architectures (e.g., vision transformers) for computer vision tasks including image classification, object detection, and semantic segmentation.
- Proficiency with deep learning frameworks and libraries, including PyTorch and TensorFlow.
- Strong understanding of AI/ML evaluation metrics, benchmarking methodologies, and best practices for rigorous computer vision model validation.
- Solid understanding of model optimization techniques, including pruning, quantization, mixed-precision inference, and other strategies for improving computational efficiency and inference performance.
- Working knowledge of model deployment and inference tools, such as ONNX, LibTorch, and TensorRT.
- Self-driven and capable of independently navigating ambiguous research problems, working with incomplete data, and delivering high-quality results.
- Strong written and verbal communication skills, including the ability to prepare technical documentation, reports, and scientific manuscripts.
Company overview
Spectral MD Inc, now operating as Spectral AI, is a pioneering predictive AI company focused on revolutionizing wound care through advanced medical diagnostics. Founded in 2009 and headquartered in Dallas, Texas, the company specializes in developing artificial intelligence-driven solutions for faster and more accurate treatment decisions in wound management. Their flagship product, the DeepView Wound Imaging System, is an FDA Breakthrough Designated platform that combines multispectral imaging with proprietary AI algorithms to provide clinicians with immediate, objective assessments of wound healing potential. The technology captures visible and infrared images of burns, chronic wounds, and other skin conditions, offering physiological information not visible to the human eye. Spectral MD has received significant government support, including over $125 million in non-dilutive contracts from U.S. federal agencies, particularly for burn wound assessment applications. The company's AI model is trained on a proprietary database of 263 billion clinically validated data points, creating significant barriers to entry for competitors. With applications in burn care, diabetic foot ulcers, and other wound types, Spectral MD aims to transform clinical decision-making, improve patient outcomes, and reduce healthcare costs through its innovative predictive diagnostics technology.