Data Scientist, Machine Learning in Epidemiology and Patient Data Products

Staff Data Scientist, Machine Learning in Epidemiology and Patient Data Products

Lexington, Massachusetts, United States;
Remote;
San Francisco, California, United States

Valo Health is a human-centric, AI-enabled biotechnology company working to make new drugs for patients faster. The company's Opal Computational Platform transforms drug discovery and development through a unique combination of real-world data, AI, human translational models and predictive chemistry.

Our talented team of biologists, chemists and engineers, armed with advanced AI/ML tools, work together to break down traditional R&D silos and accelerate the speed and scale of drug discovery and development.

Valo is committed to hiring diverse talent, prioritizing growth and development, fostering an inclusive environment, and creating opportunities to bring together a group of different experiences, backgrounds, and voices to work together. We embrace new ways of learning, solve complex problems and welcome diverse perspectives that can help us advance patient-centric innovation.

Valo is headquartered in Lexington, MA, with additional offices in New York, NY and Tel Aviv, Israel.

As a Staff Data Scientist, Machine Learning in Epidemiology and Patient Data Products, you will be a core member on a team of data scientists building a powerful computational platform for advancing the discovery and development of new medicines. In this role, you will develop machine learning tools for patient data and drive their adoption across teams, under the guidance of epidemiology and biology program leads.

Successful candidates will work with a diverse group of scientists and domain experts, in ways that cut across traditional industry boundaries in an innovative startup environment.

Your primary areas of responsibility will be:

• As a senior member of our team, you will lead the development of machine learning (ML) methods and analyses of patient data with diverse stakeholders. For example, integrate clinical insights into supervised and unsupervised learning approaches and generate patient profiles.
• Perform project-specific hands-on analysis and modeling of high-dimensional longitudinal real-world data, spanning electronic medical records (EHRs), clinical notes, sequencing data, and multi-omics, using modern data science tools in cloud environments.
• Contribute to the design, implementation, and evaluation of innovative machine learning approaches for patient data to provide novel clinical insights.
• Be comfortable with scientific uncertainty and embrace curiosity and creative solutions. Many of the challenges we tackle don't have known solutions or established pathways.
• Use your technical knowledge and intuition to articulate and break down large problems into solvable pieces. There are a lot of problems to solve; you'll need to prioritize which of these are critical-path today from those that can wait.
• Be a dynamic and active team member, championing shared coding standards, participating in code reviews, and providing regular updates on your work and input into the work of your colleagues.

• MS, MPH, or PhD in health data science, biostatistics, or a related quantitative field, with 5 years of experience developing and applying ML methods, including at least 3 years working directly with real-world patient data. Experience in a biopharmaceutical, epidemiological or biostatistical setting is a plus.
• Extensive experience developing and implementing machine learning solutions in healthcare databases, including EHRs, administrative claims, and patient registries. Familiarity with U.S. and global medical coding ontologies and data models (ICD, ATC, LOINC, SNOMED, CPT, HCPCS, OMOP, etc.). Confident working with highly sparse and high-dimensional data. Experience processing and mining clinical notes is a plus.
• Extensive experience building, maintaining, and operationalizing ML pipelines, and translating model outputs into meaningful insights for diverse audiences.
• Broad proficiency across core ML paradigms (e.g., supervised, unsupervised, semi-supervised) and experience with linear and logistic regression,…