About the TeamThe Surgical Data Science Center (SDSC) develops AI-powered tools that analyze surgical video to improve surgical techniques. Our platform processes thousands of surgical procedures, extracting tool-tracking metrics, surgical phase detection, and clinical efficiency scores that help surgeons understand and improve their technique. We work at the intersection of machine learning, clinical research, and software engineering.Role SummaryWe are looking for a Data Scientist to lead analytical workstreams across our surgical analytics platform. You will design and execute studies that validate AI-derived surgical metrics against clinical outcomes, develop composite scoring methodologies, and build data pipelines that scale across procedure types and clinical sites. You will be a key bridge between our ML engineering team and clinical collaborators, translating model outputs into clinically meaningful tools.This role requires independent judgment about statistical methodology, comfort working with messy real-world clinical data, and the ability to communicate complex findings to both technical and clinical audiences.ResponsibilitiesStudy design and execution: Design and run clinical validation studies — correlating AI-derived metrics with surgical outcomes (e.g., complications, resection extent, procedure duration)Scoring methodology: Develop and refine composite scoring algorithms (PCA-weighted, Bayesian, or other approaches) that summarize multi-dimensional surgical performance into interpretable scoresStatistical modeling: Apply appropriate statistical methods (logistic regression, mixed effects, survival analysis, dimensionality reduction) to clinical datasets with clustered, sparse, and heterogeneous dataData pipeline development: Build and maintain Python pipelines that extract, transform, and analyze data from MongoDB, PostgreSQL, and S3 at scale (hundreds to thousands of procedures)Data quality and integrity: Design and implement data validation checks, investigate discrepancies across data sources, and ensure reproducibility of analysesClinical collaboration: Work directly with surgeons and clinical researchers to define metrics, interpret results, and refine tools based on clinical feedbackReporting and communication: Produce analysis reports, methodology documentation, and presentations for internal teams, clinical partners, and external stakeholdersQualificationsRequiredMaster's degree (or equivalent experience) in statistics, biostatistics, data science, computer science, or a related quantitative field2+ years of experience in applied data science or quantitative researchStrong Python skills for data analysis and pipeline development (pandas, NumPy, SciPy, scikit-learn)Solid understanding of statistical methods: regression, hypothesis testing, dimensionality reduction (PCA/factor analysis), bootstrap inferenceExperience with SQL databases (PostgreSQL preferred) and NoSQL databases (MongoDB)Ability to work independently on ambiguous problems — scoping analyses, choosing methods, and communicating trade-offsStrong written communication — ability to produce clear reports for both technical and non-technical audiencesExperience with Git and collaborative software development practicesPreferredExperience with healthcare, clinical, or biomedical dataFamiliarity with Bayesian methods or mixed-effects modelsExperience with cloud infrastructure (AWS — S3, SageMaker, or similar)Experience building interactive dashboards or data visualization toolsFamiliarity with surgical workflow, medical devices, or clinical methodologyAbout us: The Surgical Data Science Collective (SDSC) is a nonprofit on a mission to unlock the power of surgical data. We bring together surgeons, scientists, and engineers to turn surgical videos into searchable, data-rich tools. Using AI, we help uncover insights that improve technique, sharpen decision-making, and elevate patient care. From smarter metrics to secure video libraries, we give surgical teams the tools to ask better questions—and find better answers. Because when surgeons get better, patients do too.
