Member of Technical Staff, Causality

About the role

This role advances causal inference and survival analysis to strengthen predictions of treatment response. The position extends research impact into clinical deployment and supports physician decision tools. You will own the end to end design of causal models that quantify how treatments affect outcomes under realistic conditions. A core part of your work will be translating complex statistical ideas into robust software components clinicians can trust. You will partner with medical experts to ensure that methodological choices reflect real clinical workflows and constraints. The role demands rigorous experimentation, careful validation, and clear documentation of assumptions. You will shape research agendas and influence how causality is applied across the product line. Your contributions will directly affect how clinical teams interpret evidence and make decisions.

Key facts

What you'll do

Design and implement causal inference methods for treatment effect modeling to strengthen predictions of patient outcomes using observational and trial data.

Translate machine learning research into production-ready code that embeds causal findings into clinical workflows and decision support tools.

Construct model evaluation frameworks that assess performance, robustness, and calibration on clinical datasets with complex censoring and missingness.

Communicate research insights through co-authored papers, conference abstracts, and technical reports aimed at scientific and clinical audiences.

Align technical methods with clinical requirements by collaborating with physicians, domain experts, and engineers under real world constraints.

Mentor junior team members to improve project execution, code quality, and research capabilities across the causality group.

Champion best practices in experiment design, causal identification, and sensitivity analysis to ensure credible treatment effect estimates.

Integrate modern deep learning techniques with survival analysis and causal discovery tools to handle high dimensional and heterogeneous clinical data.

Drive the creation of reproducible analysis pipelines and documentation that support regulatory considerations and clinical governance.

Continuously monitor deployed models, analyze failure modes, and iterate on methods to maintain safety and performance in production environments.

Requirements

A PhD in causality, statistics, or machine learning is required for complex clinical prediction work and to ensure deep methodological rigor.

A deep understanding of causal inference methods and concepts is necessary to identify valid treatment effect estimates and avoid common biases.

Hands on experience with observational and randomized trial data is mandatory for handling real world clinical evidence and interpreting results responsibly.

Preference is given to candidates with publications in top venues such as ICML, ICLR, NeurIPS, CVPR, or leading causality and statistics journals that demonstrate impact.

A thorough grasp of core machine learning concepts is required to ensure alignment with clinical objectives and to select appropriate modeling strategies.

Solid foundations in statistics, linear algebra, probability, and machine learning are required for rigorous model development and interpretation.

Strong Python and PyTorch skills are required for implementing, debugging, and optimizing complex models that operate on sensitive clinical data.

Deep learning experience is expected to manage high dimensional clinical data effectively and to leverage modern representation learning.

Bonus domain experience in survival analysis, multi modal learning, domain adaptation, model interpretability, and computational pathology is valued for tackling challenging problems.

Prior experience with medical data supports improved prediction accuracy and safer decision making in healthcare contexts.

Nice to have

Publications that directly address causal inference, survival analysis, or healthcare applications in top conferences and journals.

Experience contributing to open source data science or machine learning libraries used in production settings.

Familiarity with regulatory and compliance considerations for clinical decision support systems.

Background in computational pathology, medical imaging, or multi modal learning alongside structured electronic health records.

Demonstrated ability to communicate technical results to non technical stakeholders and to translate research into actionable guidance.

Practical notes

The role is based at the New York HQ with full time engagement and standard working hours.

Strong work ethic and rigorous prioritization are required to meet research and delivery standards in a fast paced environment.

Typical interview steps for data roles commonly include a SQL or coding exercise, a statistics question, and a case study focused on causal or survival problems.

Candidates may be asked to design a metric, interpret an experiment, build a small model, or analyze a real dataset under time constraints.

Some interviews may include a take home analysis that mirrors real world tradeoffs in healthcare data.

Expect questions about past projects, methodological choices, and the business or clinical impact of your work.

Interviewers often evaluate how you communicate uncertainty, limitations, and tradeoffs, not only the mathematical correctness of your solutions.

Bringing a clean write up of a past analysis, with code, visualizations, and clear conclusions, is well received and can demonstrate your rigor.

Good to know

Roles in this field focus on methods that uncover cause effect relationships in patient data to improve predictions and decisions.

Common tools include Python, PyTorch, and specialized frameworks for survival analysis, causal discovery, and model interpretability.

Research outputs often appear in top machine learning and medical imaging venues, reflecting both methodological novelty and clinical relevance.

Work typically spans clinical domain knowledge, data engineering, and algorithm design, requiring comfort across multiple abstractions.

Teams are small and multidisciplinary, integrating AI pioneers, statisticians, and clinical experts to ensure solutions are both scientifically sound and practical.

Career growth

Data careers grow toward senior analyst, staff data scientist, or data engineering lead, with increasing ownership of complex systems.

Many professionals specialize further in machine learning, analytics, infrastructure, or regulatory and compliance roles as they advance.

Cross functional work with product and engineering teams becomes more important at senior levels, especially when deploying models into clinical settings.

The field changes quickly, so continuous learning is part of the job and staying current is essential for long term success.

Professionals who can translate numbers into decisions tend to advance fastest and take on greater strategic impact.